Biomorphometric Studies in Benign Prostatic Hyperplasia

Home , Androstanediol

BIOMORPHOMETRIC STUDIES IN BENIGN PROSTATIC HYPERPLASIA

Thesis submitted for the degree of

DOCTOR OF MEDICINE

in the Faculty of Medicine

of the University of London

Paul David Miller

Departments of Urology and Biochemistry

St. Bartholomew’s Hospital Medical College ProQuest Number: U063954

INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted.

In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. uest.

ProQuest U063954

Published by ProQuest LLC(2016). Copyright of the Dissertation is held by the Author.

ProQuest LLC 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 48106-1346 8

.101 IN iH NTi.H ivL’s - r w

JOHN HUNTER F.R.C.S.

From a portrait hanging in the Great Hall

St.Bartholomew's Hospital ABSTRACT

The ageing population structure, with concomitant increase in the prevalence of the pathologies of age and the mor bidity associated with prostatectomy, require that we look to alternative forms of treatment for benign prostatic hyperplasia (BPH). Although prostatic growth is dependent on androgens, especially dihydrotestosterone, the effect of steroids is probably mediated by other regulators. Androgens and oestrogens are believed to provide the hormonal environ ment that is essential for other biochemical factors to induce hyperplasia and perhaps malignancy. The aim of this study was to investigate one of the newly discovered relevant biochemical "factors" (epidermal growth factor). This growth factor is found in high con centrations in prostatic fluid and its relationship with the hormonal environment, steroid receptors and the pre sence of prostatic pathology have been examined.

Benign prostatic hyperplasia and carcinoma are known to occur predominantly in two different regions of the gland. Variations in growth factor binding and steroid receptors within the gland have therefore been studied, as well as differences between patient groups. Use of new computer ised morphometric techniques allowed the measurement, and more accurate comparison of the amounts of stromal an# glandular tissue in different samples of prostatic tissue. These were correlated with the biochemical data from the same samples.

The results show increased epidermal growth factor bind ing in BPH and a close association between steroid recep tors and growth factor receptors in the abnormal part of the prostate. Although this suggests up regulation of the growth factor receptors in BPH, receptor binding is also found to be closely related to tissue morphometry.

These findings should help understanding of growth regu lation in the prostate and the pathogenesis of BPH. Ulti mately it is hoped that this will also help identify those groups of patients who will respond to pharmacological treatment as opposed to those who will require prostatect omy for relief of their symptoms.

4 Acknowledgements

Human tissue for this study was obtained from patients under the care of Mr.R.S.Kirby, Mr. H.N.Whitfield and Mr. W.F.Hendry at St.Bartholomew's Hospital and Homerton Hospital. I am grateful to them for allowing me to study their patients. All the work in this study was carried out in the Departments of Urology, Biochemistry and Histopa- thology at St.Bartholomew's Hospital, London. The tissue biochemical assays and immunohistochemistry was performed by the author in the Department of Biochemistry at St.Bartholomew's Hospital Medical College after expert tuition by Mr.S.Barker and Mr.J.Puddefoot. I would like to thank them for their guidance and patience throughout this project. The computer morphometric analysis was performed by the author in the Department of Histopathology. The computer program for this analysis was written in conjunction with

Mr.C.Sowter and Prof.G.Slavin to whom I am greatly indebt ed.

The work at St.Bartholomew's Hospital was performed under the guidance of Mr.R.S.Kirby MA, MD, FRCS (Urol) and Prof.G.Vinson Ph.D whom I would particularly like to thank for their help and encouragement. I received valuable help and advice during this project from Dr. Colby Eaton, Dr. Maureen Harper and Prof. K. Griffiths at the Tenovus Institute for Cancer Research, Cardiff. 1 must also thank the staff of the Department of Medical

Illustration and the Medical college library. Miss J.Bryan

collected many of the specimens from the operating theatre

for which I am also most grateful.

This work would not have been possible without the finan cial support of The Special Trustees of the Medical Col lege of St.Bartholomew's Hospital. Ethical permission to study patients was obtained from the Ethical Committee of St.Bartholomew's Hospital. All patients gave informed verbal and written consent. Mr. J. Higgins F.R.C.S. kindly gave me permission to reproduce figures 1 (iv) (v) and (vi), which are from his original work. CONTENTS PAGE

FRONTISPIECE 2 ABSTRACT 3

ACKNOWLEDGEMENTS 5

CONTENTS 7

LIST OF FIGURES 9 LIST OF TABLES 12

LIST OF APPENDICES 14 CHAPTER ONE - REVIEW OF THE LITERATURE 15 I Introduction 18 II Epidemiology and Aetiology 20 III Anatomy 27 IV Endocrinology and Biochemistry 48 V Morphometry 88 CHAPTER TWO - AIMS OF THE STUDY 98

I Background 99 II Experimental strategy 100 CHAPTER THREE - MATERIALS AND METHODS 102

I Tissue collection 105 II Chemicals 109 III Apparatus 113 IV Preparation of tissue for receptor assay 113 V Binding studies 115 VI Serum hormone concentrations 121 VII Immunohistochemistry 124 VIII Morphometric techniques 126

CHAPTER FOUR - VALIDATION OF THE METHODOLOGY 137

I Introduction 139 II EGFR Assay 139 III Steroid Receptors 145 CHAPTER FIVE - BIOCHEMICAL RESULTS 148

I Introduction 150 II EGFR binding studies 151 III Steroid receptor binding studies 167 IV Regional distribution 175 VI Serum hormone levels 184

CHAPTER SIX - MORPHOMETRIC RESULTS 189 I Introduction 191 II Morphometry in BPH and normal prostate 193 III Comparison with pathologist 198 IV Biochemical and morphometric results 200

CHAPTER SEVEN - GENERAL DISCUSSION 208 APPENDIX 224

BIBLIOGRAPHY 238

PRESENTATIONS/PUBLICATIONS 298

8 LIST OF FIGURES Page

1. (i) Percentage increase of BPH with age 18

1. (ii) Prevalence of BPH pathology at autopsy 21 1. (iii) Transverse slice of normal human 34

prostate

1. (iv) Coronal section of whole normal 35 prostate (x2.5)

1. (v) High power (x230) of central zone 36 prostatic tissue

1. (vi) High power (x230) of peripheral zone 37

prostatic tissue

1. (vii) Diagram of longitudinal zonal prostatic 45 anatomy in BPH with corresponding transrectal ultrasound scan

1. (viii) Diagram of transverse zonal prostatic 46

anatomy in BPH with corresponding

transrectal utrasound scan

1. (ix) Diagram of the "two step" model for the 61

mechanism of action of steroid hormones 1. (x) Diagram of the mechanism of action of 63

steroid hormones based on the "one step"

model

1. (xi) Diagram of the EOF cell surface receptor 81 1. (xii) In situ hybridisation of EOF receptor 85

9 3. (1) The Kretz Combison transrectal ultrasound 106 probe

3. (la) The transrectal probe in the rectum 107

3. (ii) Longitudinal transrectal ultrasound of 108

the prostate during biopsy

3. (iii) The IBAS image array processor 127 3. (iv) Video camera and microscope 128

3. (V) Monochrome Image 130

3. (vi) Binary Image 131 3. (vii) Binary Image inverted and dilated 132

3. (viii) Acinar contours of image 133 3. (ix) Epithelial component of image 133

3. (X) Epithelial component in red superimposed 134 on original image

3. (xi) Epithelial component and acinar lumen 135

4. (i) Duration of incubation 140 4. (ii) i25i gQp concentration 141 4. (iii) Influence of dilution 142 4. (iv) Comparison SA v SDA oestrogen receptors 147

5. (i) Saturation curve and Scatchard plot of 155 125i-EGF binding to BPH tissue

5. (ii) EGFR SA V SDA in BPH TURP specimens 157

5. (iii) EGFR binding, normal v BPH tissue 159 5. (iv) Immunohistochemistry - BPH - control 163

5. (V) Immunohistochemistry - BPH - EGF antibody 164

10 5. (vi) Immunohistochemistry (HPx320) - BPH - EGF 165 antibody

5. (vii) EGFR V ER binding BPH specimens 173

5. (viii) Regional distribution EGFR binding 178

5. (ix) Regional distribution ER binding 179

5. (X) Regional distribution PgR binding 181 5. (xi) EGFR V ER binding inner zone 182

5. (xii) EGFR V ER binding outer zone 183 5. (xiii) Steroid hormone and SHBG controls v BPH 185 5. (xiv) T/E2 ratio controls v BPH 187

6. (i) Morphometry Normal v BPH tissue 197 6. (ii) Epithelium % v EGFR binding 202 6. (iii) Glands v EGFR binding (BPH) 202 6. (iv) G/S ratio v EGFR binding (BPH) 203

6. (V) % Stroma v ER binding (BPH) 205

6. (vi) G/S ratio v ER binding (BPH) 205

6 . (vii) G/S ratio v T/E2 ratio (BPH) 207

7. (i) Correlation of % increase in flow rate 221 and smooth muscle

11 LIST OF TABLES

4. i. Reproducibility of EGFR assay - intraassay 144

variation

4. ii. Reproducibility of EGFR assay - inter assay 144 variation

5. i. EGFR assay BPH tissue - SDA 153

5. ii. EGFR SDA v Scatchard analysis (SA) 157 5. iii. EGFR assay normal tissue - SDA 159 5. iv. Immunohistochemistry v SDA - BPH 162

5. V . Immunohistochemistry v SDA - normal 162 5. vi. ER - BPH tissue 168

5. vii. ER - normal tissue 169 5. viii. PgR - BPH tissue 170 5. ix. PgR - Normal tissue 171

5. X . EGFR V Steroid receptors 173

5. xi. Regional distribution EGFR 177 5. xii. Regional distribution ER 179 5. xiii. Regional distribution PgR 180 5. xiv. EGFR and steroid receptor distribution 183

5. XV. Serum hormone levels BPH and controls 184

5. xvi. Serum hormone levels, EGFR and ER binding 186

in BPH

6. i. Sample of morphometric data and subsequent 192

calculation for each patient

12 6. ii. Morphometry data - normal post pubertal 194 prostatic tissue

6. iii. Morphometry data - normal prostatic tissue 194 6. iv. Morphometry data - BPH tissue 195

6. V. G/S ratio and "glandularity" 199

6. vi. Morphometry and biochemical analysis - normal 200

prostatic tissue 6. vii. Morphometry and EGFR - BPH 201

6. viii. Morphometry and steroid receptors - BPH 204 6. ix. Morphometry and steroid hormones - BPH 207

13 LIST OF APPENDICES

I Data from validation of methodology 224

II EGFR results raw data 227

III Steroid receptor results raw data 231

IV Regional distribution results 233

V Steroid Hormones - controls 237

14 CHAPTER ONE

REVIEW OF THE LITERATURE

15 Review of the literature PAGE

I INTRODUCTION 18

II EPIDEMIOLOGY OF BENIGN PROSTATIC HYPERPLASIA 20 1. Demographic factors 20 2. Racial and other risk factors 23 3. Relationship between BPH and malignancy 25 4. Summary 26

III ANATOMY 27

1. Introduction 27

2. Comparative anatomy 28 3. Human foetal and developmental anatomy 30 4. Normal human adult anatomy 32

5. Pathological anatomy 39

6. Ultrasonic anatomy 41

IV ENDOCRINOLOGY / BIOCHEMISTRY 48 1. Historical background 48

2. Steroid hormones in the pathogenesis of BPH 51 3. Mechanisms of hormone action 59 4. Other hormones regulating prostatic growth 70

5. Stromal-epithelial interaction 73

6. Growth factors 77

16 V MORPHOMETRY 1. Introduction 88 2. Morphometric techniques 88

3. Image analysis 91

4. Morphometry in the prostate 95

17 I INTRODUCTION

Benign prostatic hyperplasia (BPH) is an abnormal, benign growth of the prostate which commonly develops with age in men. Enlargement of the gland gradually occurs and classi cal stromal nodules are formed, together with glandular hyperplasia. As the gland enlarges it often begins to cause mechanical obstruction to the outflow of urine from the bladder during voiding. This leads to the typical symptoms of poor stream, hesitancy and terminal dribbling. In addition, the obstruction may cause secondary bladder

instability which accounts for the irritative symptoms (frequency, nocturia and urgency) of this condition. Pathological changes frequently begin in the fourth de cade, but significant enlargement with concomitant symp toms does not develop in most cases until after 50 years of age (Swyer 1944, Franks 1954) (Fig.l (i)).

Fig.l (i) Percentage of age group with BPH 100 r ■ ■ Mieroaeople BPH

I ) Maereaeople BPH

80 rm rm ciimcai bph

0 20 40 60 80 100 Host Age (Years)

18 Epidemiological studies have attempted to define the dimen sions of the clinical problem that surgeons may face In the future, as well as trying to Identify aetlologlcal risk factors. Ageing and the presence of Intact testes are necessary features In the development of BPH, but the specific biochemical and/or hormonal factors that provoke this condition are unknown. Better understanding of the biochemistry and endocrino logy of the normal prostate and of the changes which occur In BPH may provide clues to the aetiology and pathogenesis of BPH. This In turn could lead to a more rational ap proach to treatment. Recently the role of Interactions between stroma and epithelium In maintaining physiological homeostasis and In neoplasia has been recognised. There fore morphometric studies make a crucial contribution to Interpreting the biochemical results. Hence the focus of this thesis Is on exploring the relationship between disturbance of biochemical and hormonal balance In BPH on the one hand and morphometric changes In the prostate gland structure on the other.

This review of the literature will first outline the extent of the clinical problem of BPH. This will be fol lowed by a description of the anatomical, biochemical and morphometric background relating to current research Into the pathogenesis and possible medical treatment of BPH.

19 II EPIDEMIOLOGY OF BENIGN PROSTATIC HYPERPLASIA

Although benign prostatic hyperplasia (BPH) is the most common benign proliferative abnormality in men, knowledge of the epidemiology of BPH remains fragmentary and re quires more research. This is due mainly to difficulties of collecting reliable epidemiological data and variations in the diagnostic criteria used for BPH (Meyhoff and Hald

1978). Some data are available from the Hospital In-pa tient Enquiry, but these are based on episodes of admis sion and, as an unknown number of in-patient episodes may result from a single case of BPH, they are difficult to interpret (Berry and Jones 1982). Not all BPH is sympto matic and it has been shown that symptoms correlate poorly with prostatic size (Turner-Warwick et al. 1973). This and the lack of standard diagnostic criteria must be borne in mind when considering data from clinical case series.

1. Demographic Factors

Calculation of incidence and prevalence rates depend on the identification of a source population (population at risk) in order to provide a denominator, which is usually difficult to determine. Indeed up to now the only popula tion-based study is that carried out by the Veterans Administration (Glynn et al. 1985).

Some of the early studies were based on autopsy series which have the advantage of overcoming variations in

20 diagnostic criteria but, as subjects are not necessarily representative of the general population, they cannot be used to derive prevalence rates. An early Austrian study

(Moore 1935) was based on a collection of 675 prostates from every male autopsy conducted in two Vienna hospitals

In this study the prevalence of "benign enlargement"

(a term not defined) rose from 0% at age 20-29 years to 75% in men aged 80 and upwards. In Britain, Franks (1954) performed detailed histological studies on autopsy specimens. He also found increased prevalence in each age category from 0% at 20-29 years to 85% in men over 80 years old. Data collated from several autopsy series was published by Berry et al. (1984) (Fig. 1 (ii)).

Fig. 1 (ii) ( From Berry et al. 1984)

PREVELANCE OF BPH PATHOLOGY WITH AGE IN 1075 HUMAN AUTOPSIES Percent of Prostates 100

1-10 11-20 31-40 41-50 51-6021-30 61-70 71-60 27 35 105 94 221

21 In a series of nearly 7000 Insurance medicals, also by their nature a selected population, 20% of men in the sixth decade and 43% of men in the eighth decade had palpable enlargement of the prostate (Lytton et al. 1968). A recent British study (Garraway et al. 1991) suggests that the prevalence of BPH is higher than has been repor ted previously in clinical retrospective and autopsy studies. They report a prevalence rising to 430 per 1000 men aged 60-69 years using symptom score, flow rate and transrectal ultrasound in 705 men. All these studies have their limitations, nonetheless it can be concluded that BPH is a very common condition in males over 50 years of age and, in those over 80, histolo gical criteria can be fulfilled in over 80% of subjects. Rotkin (1976) observed that the frequency of symptomatic BPH in American men of all ages, with palpably enlarged glands was 50-70% in most series. Two studies have attempted to determine time trends in the evolution of BPH. The most recent study by the Veter ans Administration is one of the few that have followed a cohort for a prolonged period (Glynn et al. 1985). Apply ing clear criteria for clinical diagnosis and monitoring surgical treatment they followed 2036 volunteers for up to 25 years. From life-table analysis of these data it was estimated that the lifetime probability of requiring surgical treatment for BPH was 29%. The need for operative relief of obstruction seems a more useful measure of

22 significant BPH than any clinical assessment of prostatlc size or histological evidence of hyperplasia. This figure however Is higher than that calculated retrospectively by Lytton et al. (1968), who estimated that a 40 year old man had a 10% chance of requiring surgery before he was 80.

2. Racial and other risk factors

There are reports that prevalence of BPH Is higher amongst blacks than In white men and of a relatively low frequency In Orientals. Many of these studies are epldemlologlcally flawed (Rotkin 1983), but the evidence of racial differ ences cannot be Ignored. Kambal (1977) reported a high prevalence of BPH In North ern Sudan (where Intermarriage with arab migrants occurs) but no case was found In the "pure African" population of Southern Sudan. Significantly higher levels of BPH have been shown In blacks than In whites In New Orleans (Derbes et al. 1937) and Alabama (Morgan and Spruell 1983), but the reports from Africa are largely anecdotal and lack a secure epidemiological basis. In Britain, Richardson (1965) In a study relating social class to risk of BPH, showed a higher morbidity In the upper classes In Scotland, and Ashley (1966) reported an excess of deaths from BPH In areas with a high Indigenous Welsh population.

23 Age is by far the strongest risk factor associated with prevalence and probably the incidence of BPH. Many believe that the age association reflects age-related hormonal changes, and there is now some evidence to support this view (Zumoff et al. 1982).

The Veterans study (Glynn et al. 1985) reported higher rates of BPH among Jewish men and lower rates, particular ly of clinically significant BPH, among smokers. However the former association may reflect socio-economic status and the latter could reflect a reduced likelihood of surgery among smokers because of their increased anaesthe tic risk or other confounding variables. A link with hypertension reported by Bourke and Griffin (1966) has not been verified by subsequent studies (Glynn et al. 1985, Morrison 1978). An association between BPH and diabetes (Bourke and Griffin 1968) and cirrhosis of the liver (Robson 1964) may also be related to hormonal imbalance. A relative increase in plasma oestrogens may provide the link, due to de creased metabolism of oestradiol in the liver in diabetes and cirrhosis. These endocrinological factors could be epidemiologically relevant, as genetic variation between populations may operate on hormonal function and correlate with risk.

24 3. Relationship Between BPH and Malignancy

In view of the high prevalence of BPH, its presence in approximately 20% of all cases of invasive carcinoma does not necessarily imply that it is a pre-malignant condi tion. There has been considerable controversy about this relationship, Armenian et al. (1974) argued on histologi cal grounds that BPH is an early stage in the development of prostatic cancer, but Greenwald et al. (1974) found no evidence to support this conclusion. Relative risks for development of carcinoma in patients with BPH are quoted as being between 3 and 5. This may merely reflect the increased chances of finding one disease if the other is present and under investigation or treatment (Catalona and

Scott 1978).

At the present time conflicting evidence makes it impos sible to conclude that BPH is a precursor condition for the development of prostatic cancer. Very recent ultra sound evidence (F.Lee, 1990, personal communication) suggests that these are separate conditions and also that carcinomas with different malignant potential arise in different areas of the gland which underlines the hetero geneous nature of this organ.

25 4. Summary

Genetic, hormonal and environmental factors are all apparently associated with the development of BPH. Epide miological studies however, do not reveal any specific risk factors for the development of BPH or its evolution. In particular no factors amenable to preventive interven tion have been identified. It is certain that this condi tion requires more epidemiological investigation. In 1900 only 25% of the U.S. population survived to age 65, now 70% reach this age. The proportion surviving to over 80 is 30% and increasing (Brody 1984). In a few years projected figures suggest that 50% of the male population will survive more than 80 years (Brody 1985). This group have an 88% prevalence of BPH (Berry et al. 1984), so there is undoubtedly going to be a marked increase in the absolute number of patients with clinically relevant BPH, requiring effective treatment. The implications of this increase for health service provision and the personal and economic costs are considerable.

26 Ill ANATOMY

1. Introduction

Comparative anatomy, embryonic development and adult morphology form interlocking bodies of evidence relevant to human organogenesis and pathogenesis. It is necessary to consider all of these in order to begin to understand the function and pathology of the different areas of the prostate gland.

The heterogeneous composition of the prostate, with different glandular regions and regions of fibromuscular tissue tightly fused within the prostatic capsule, make dissection difficult and this has led to several misinter pretations of the anatomy. Based on embryological stu dies, Lowsley (1912) first described the prostatic lobes

(middle, lateral and anterior), but since then several differing descriptions of these prostatic lobes have been published (Gil-Vernet 1953, Tisell and Salander 1975).

Whether or not separate morphological "lobes" exist, there are undoubtedly areas within the normal gland which are stucturally quite different. For instance the histological appearances of the central portion of the gland differ markedly from those of the peripheral part.

The histological distinction between zones is clinically important because the peripheral zone is almost exclusive ly (80%) the site of origin of carcinoma (McNeal 1969).

27 Also the study of the anatomical and histological hetero geneity within the gland, which almost certainly reflects important biological differences, may help an understand ing of the pathogenesis and pathological anatomy of prostatic neoplasia.

2. Comparative Anatomy

All male mammals possess one or more accessory reproduc tive glands, but the only one found in all orders of mammals is the prostate (Price and Williams-Ashman 1961). However, there is considerable variation in the compara tive anatomy both between species and within species and it is difficult to identify homologous structures in man. Hunter (1786) identified further difficulties when he recorded the change in size of the prostate during the rutting season, this observation has since been made in many other species (Price 1963). The variations in the prostatic anatomy seen in various mammals can be divided into a) a diffuse gland lining the

urethra (eg. sheep, goat), b) a discrete "body" (eg.dog,man), which may be lobular (eg. rat) c) a combina tion of the two (eg. bull). The use of the term "lobe" has caused some confusion as it may refer either to grossly separate elements of the organ or to a part of the gland which is histologically distinct or differs in its re sponse to hormones.

28 The work of Hill (1953-1961) is invaluable in the study of the comparative anatomy of our phylogenetically closest relations, the primates. The basic pattern of the primate prostate is two anatomically distinct bodies, a cranial gland and a caudal gland, the main bulk of both being dorsal to the urethra. The vas deferens and seminal vesi cle ducts run through a cleft between these two glands, and open separately or as a common ejaculatory duct onto the verumontanum. The caudal prostate tends to be larger and envelops the cranial gland, thus lying more posterior ly. The ducts of the caudal gland enter the urethra below the ejaculatory ducts, whereas the ducts of the cranial gland enter above the verumontanum. In some species there is a suggestion of a "middle" lobe just cranial to the verumontanum, in addition to the main dorsal component of the prostate. There are many variations on this structure but with this basic model we can draw some homologies. The concept of dual morphology can be supported if we consider that the cranial prostate is homologous to the central zone described by McNeal (1981) which lies cranial to the ejaculatory ducts. This can be extended to include the coagulating gland of the rodents as being homologous to the primate cranial prostate. Van Wagenen (1936) demon strated that the secretion from the cranial lobe of the monkey (Maccaca mulatta) causes coagulation of the fluid from the seminal vesicles. The microscopic differences within the prostate in man emphasised by McNeal have been

29 shown to exist in the cranial and caudal prostates of the monkey (Blacklock and Bouskill 1977). Both groups found that the glands of the caudal prostate tend to have smal ler acini and more delicate stroma.

3. Human Foetal and Developmental Anatomy

Androgens trigger the budding of ducts from the urethral lining into the adjacent stroma (Siiteri and Wilson 1974). This process begins in the twelth week of intrauterine life when the embryonic testes begin to produce testoster one. Subsequent duct development was studied in detail by Lowsley (1912) who described five lobes, (posterior, anterior, middle and two lateral) drained by five separate groups of ducts. However, Franks (1954) and others (Le Duc 1939, McNeal 1972) have since shown that there is little evidence for a lobular structure in later development and they stated that "Lowsleys lobes" do not exist in the adult prostate. McNeal (1968) proposed an alternative embryological basis for the adult structure. He observed that the central zone ducts arose with the ejaculatory ducts and followed their course superiorly through the prostate. He suggested that this implied derivation from the Wolffian ducts rather than from the urogenital sinus. Closer examination shows that as the central zone ducts branch they enclose the ejaculatory ducts both anteriorly and posteriorly and in the adult the compact stroma of

30 these ducts blends with the stroma of the ejaculatory ducts. The stroma of the rest of the prostate tends to be finer and more loosely woven. Furthermore the mucosal pattern of the acini, with epithelial infolding is similar to that of the seminal vesicles and ejaculatory ducts and individual cell characteristics are similar (McNeal 1968).

At birth there is hyperplasia and occasionally squamous metaplasia of the prostatic duct epithelium especially in the zone surrounding the Wolffian ducts; possibly due to maternal oestrogens in the blood (Swyer 1944). These changes subside six to seven weeks after birth and suse- quently there is little change in the prostate up to the age of nine. It is interesting that Blacklock (1982) has demonstrated that sections of pre-pubertal prostates showing earlier acinar development in the central zone and incomplete differentiation in the peripheral zone. This may be due to greater androgen control of the peripheral zone.

From the age of nine there is progressive hyperplasia of the ductal epithelium and formation of side buds, and the gland grows slowly and continuously up to puberty. During puberty the gland doubles in size due to follicular devel opment from end buds and modification of ductal branches. Growth is associated with a condensation of stroma which decreases relative to the glandular tissue (Swyer 1944), and is probably due to increased testosterone production.

Up to the third decade glandular epithelium grows by

31 Irregular multiplication of the epithelial Infoldlngs Into the lumen of the follicles.

4. Normal Human Adult Anatomy

Histological studies have shown that In the adult pros tate there Is no lobar pattern, but rather there are two concentric zones of glandular tissue which partially surround the prostatlc urethra (Le Duc 1939, Franks 1954, Ferguson and Gibson 1956). Differential histology In the mature gland has been described by Moore (1936), but he concentrated on the pleomorphlsm of cells In different acini rather than on the architecture of the gland. All of these early studies were anatomically Incomplete because they were limited to transverse sections through the middle of the gland and did not Include either the more proximal or distal tissues, nor did they study other planes of section. In contrast, the now classic studies of McNeal (1968) used serial blocks throughout the entire gland In multiple planes of section. From his findings In over 500 adult specimens he was able to construct a three dimensional model of the gland (McNeal 1981).

In essence he described four basic anatomical regions, the central and transitional zones, the peripheral zone and the anterior fIbromuscular stroma. In order to under stand their relationship It Is Important to appreciate that the prostatlc urethra Is not a straight tube since at

32 its midpoint it undergoes a sharp anterior kink of approx imately thirty five degrees where it is joined by the ejaculatory ducts at the level of the verumontanum.

The central zone ducts arise close to the ejaculatory duct orifices and follow them proximally, branching out later ally towards the base of the prostate. Thus these ducts occupy the majority of the gland in the angle between the ejaculatory ducts and the proximal urethra, and constitute approximately 25% of the glandular prostate. This zone is surrounded by the peripheral zone which also encloses the distal prostatic urethra. The peripheral zone ducts arise separately in the recesses lateral to the base of the verumontanum, and develop laterally into the mesenchyme posterior to the distal urethral segment. Their branches fan out laterally and proximally to cover a wide area and make up 75% of the glandular tissue. The ducts of both these zones grow into the same block of mesenchyme, and the two are therefore continuous. The most proximal branches of the peripheral zone lie almost in contact with the most lateral of the central zone, but despite the fusion of the two zones at their borders, an anatomically distinct boundary line still remains visible between them in the normal adult gland (Fig.l (iii)). This clear anatomical delineation between the two zones is highlighted by the histological differences between the two zones which suggest differences in biological func tion.

33 Fig.l (iii)

m 10

Fig.l (iii) Transverse slice of normal human prostate age

48 years divided macroscopically to show urethra (U), verumontanum (V), and central (C ) and peripheral zones of the gland.

The different histological features within the normal prostate have recently been studied in a group of young transplant donors (J.Higgins, personal communication 1991)

(Fig.l (iv).

34 ; Fig.l (iv)

Fig.l (iv) Coronal section of whole normal prostate age 19

years (x 2.5). Massons trichrome stain shows capsule and external sphincter.

In the peripheral zone, the acini are more regular and the

fibromuscular stroma is thinner (Fig.l (v)). In the cen

tral zone the stroma is more abundant and the secretory

acini larger (Fig.l (vi)). Also the lining epithelium of

the acini differ (McNeal 1981); in the peripheral zone

there is a simple columnar epithelium with distinct bor

ders and small nuclei, whereas in the central zone the

cells are of variable size, with less distinct cell mem

branes and more crowded larger nuclei.

35 Fig. 1 (v)

% 4

Fig.l (v) High power (x230) of tissue taken from the

central zone of the normal prostate. This shows large acini with complex infoldings. The fibromuscular trabecu

lae are considerably thicker than in the peripheral

prostatic tissue and the acini are larger.

36 Fig.l (v i )

Fig.l (vi) High power (x230) of tissue taken from the peripheral zone of the normal prostate. Small regular

acini and thin fibromuscular trabeculae are demonstrated

37 No major ducts arise from the urethral segment proximal to the verumontanum and this has been called the pre prostatic region. Surrounding this segment, a cylindrical smooth muscle sphincter extends down and is continuous with the bladder neck muscle. Within this cylinder, in the immediate periurethral stroma, there is a group of glands

(the periurethral glands) which appear incompletely deve loped and extend upwards towards the bladder neck. In order to attain full glandular development the periure thral ducts would either have to penetrate the cylindrical barrier or the muscle would have to be incomplete. It is only at the lower end of the prostatic urethra, near the base of the verumontanum where the sphincter is incom plete, that a small group of ducts is found which are able to branch out laterally towards the central zone and thus achieve glandular development. Though there is more duct branching and acinar proliferation than in the other periurethral ducts, it is still only a small area and

constitutes only 5% of the glandular prostate. This area is macroscopically indistinguishable from neighbouring zones but is known as the transition zone and is thought

to be the site of origin of early benign prostatic hyper plasia (McNeal 1978).

The last main anatomical region is the anterior fibro

muscular stroma . This area is non-glandular and of little pathological significance and makes up about one third of

38 the weight of the gland. It is continuous with the detru sor muscle of the bladder neck and inseparably fused to the gland, producing the anterior convexity.

5. Pathological Anatomy

There is now general support for the concept that benign and malignant diseases of the prostate are independent entities and typically originate in different areas of the gland (McNeal 1969). Morgagni (1769) first observed as early as the 18th century that BPH was due to hyperplasia of the inner group of glands, but the location of these glands was not clearly defined until the end of the nine teenth century. In 1894, Jores described groups of submu cosal glands on the trigone, vesical neck and the urethral walls just below the bladder neck. These enlarged with age and median lobe enlargements arose from these glands. In addition, Albarran and Motz (1902) reported periurethral glands embedded in the inner longitudinal smooth muscle layer of the urethral wall between the verumontanum and bladder neck.

Since these observations there has been controversy concerning the exact site of origin of BPH. Moore (1944) believed that BPH arose in the lateral and middle lobes of the true prostate (central portion) as defined by Lowsley

(1912) in his embryological studies. By contrast, Deming and Neumann (1939) and Le Duc (1939) described BPH arising

39 in and limited to the immediate periurethral tissue, or to

a selected group of neighbouring "inner" prostatic glands

(Franks 1954). It is still not clear whether or not the

hyperplastic glandular tissue involved arises from the

local periurethral ducts (Le Duc 1939 and McNeal 1972) or whether it is generated from branches of the main prostatic ducts (Deming and Neumann 1939).

The exact anatomical landmarks and boundaries of this region were incompletely defined until the work of McNeal

(1977), when he subjected the tissues within and lateral to the proximal prostatic urethra to systematic examina tion. Coronal sections were taken through the upper seg ment of the urethra and, in planes parallel to this, tracings were made of the anatomical boundaries and nod ules. As a result he was able to identify the precise area

in which early nodules appeared. He concluded that most nodules arose in the tissue that lies just proximal to the verumontanum, close to the muscle of the distal end of the bladder neck sphincter. Some nodules also seem to develop in the periurethral glands, but these are small and do not form the main mass of BPH tissue as described in previous studies (Le Duc 1939, Pradham and Chandra 1975).

This small area along the long axis of the urethra may well be the site of origin of BPH, but the further evolu tion of BPH is unclear and the subject of some specula tion. Further nodule genesis seems to occur in a focal and random manner, but is always confined to the transitional

40 and central zones of the gland, never appearing to spread out to involve the peripheral zone where the majority of the glandular tissue is situated. In many men over 70, enlargement and glandular proliferation takes place within the nodules already in existence.

Both stromal and epithelial hyperplasia may occur, alone or together, and it is impossible to say that one occurs before the other (Franks 1954). Further consideration will be given to this aspect when discussing stromal- epithelial interactions and the morphometry of BPH in more detail.

The pathological anatomy of BPH highlights the regional differences within the gland, with early diffuse growth of the transition zone occurring without change in its archi tecture, followed by massive enlargement of the nodules to occupy the centre of the gland.

6. Ultrasonic anatomy

Prostatic ultrasound started with an attempt at transrectal ultrasonography, which is now recognised as the most suitable method of imaging the prostate routinely. In

1955 Wild and Reid presented their screw type transrectal scanner, this was planned primarily for diagnosing tumours of the rectum and large bowel.

The principle was identical with that of modern transrectal ultrasound, but only a poor image was obtained because

41 of the inferior state of electronic technology at that time.

Several urologists followed up this idea (Takahashi and

Ouchi 1963, Gotoh and Nishi 1965), reporting on the A-mode appearances of the prostate via the transrectal route, but this could not reveal the distribution of tissues in various planes. The first radial scanning transrectal probe was developed by Takahashi and Ouchi in 1964 and although the picture quality was poor they obtained hori- z onta1 tomograms.

Watanabe et al. (1968 and 1971) were the first to use an ultrasonic probe with an oscillating disc in clinical practice. Their images were improved by examining the patients seated on a chair thus removing interference from air bubbles. Over the next ten years real time ultrasonic images steadily improved as a result of electronic advances and the first linear scanner able to provide longitudinal scans was introduced by Sekine et al. (1982).

This is still the state of the art scanner, although additional attachments for needle guidance, allowing accurate sampling of the prostate have been added (Saitoh et al. 1981).

It is now possible with transrectal ultrasound (TRUS) imaging to depict the major divisions and structures of the prostate. Ultrasonically the organ can be divided into an anterior component consisting of urethra and its muscu- lostromal complex and a posterior component comprising the

42 glandular tissue and ejaculatory apparatus. The centrally placed urethra can be divided into two equal segments as it angles forward at the proximal end of the verumontanum which provides a good reference point.

It is possible to distinguish the central and transition al zones from the peripheral zone both in the normal and the majority of diseased glands. However it is not pos sible to clearly distinguish the transitional and central zones in established BPH and biopsy them separately. In order to clarify the areas which have been biopsied the transitional and central zones, where BPH is confined, are combined and called the inner zone whereas the peripheral gland which is relatively spared is called the outer zone (Figs.l (vii) and (viii). The peripheral (outer) zone, located posteriorly and laterally makes up most of the apex of the gland in the normal patient. It appears sonographically as a homogeneous, "isoechoic" pattern, due to the fine glandular and ductal tissues present here. This is taken as the baseline of isoechogenicity to which other areas are compared.

In the normal gland the transitional zone is a very small area and indistinguishable from the central zone sonographically. The central zone has a more heterogeneous, coarser echo pattern with a contrasting mix of hyper and hypo echoic areas compared to the peripheral zone. In early BPH the transitional zone begins to enlarge and the sonographic appearances of the central and transitional

43 (inner) zones can be compared to the peripheral (outer) zone. These sonographic appearances are variable depending on which tissue component predominates. It may be hypoechoic with respect to the peripheral zone if the stromal elements dominate, isoechoic like the peripheral zone, or relatively hyperechoic if the glandular tissue dominates

(Weiss 1989).

In the majority of patients the boundary between the peripheral (outer) zone and the remainder of the gland becomes more clearly demarcated with progress of hyperpla sia, particularly if the surgical capsule is outlined by a ring of calcified corpora amylacea. However in other, often smaller glands, the two zones appear more hypoechoic and merge imperceptibly into each other.

The ejaculatory ducts are often imaged as paired cystic structures at the base of the gland, and their associated muscles may occasionally be seen as a hypoechoic halo around the ducts. On sagittal scans the ducts are seen running from the seminal vesicles towards the verumonta num. The pre-prostatic urethra proximal to the verumonta num can be seen angling towards the bladder neck, with the post-prostatic urethra distally. The true prostatic cap sule is seen sonographically as a bright echo structure laterally while posteriorly it separates the gland from the fascia and the rectum. The seminal vesicles lying cephalad to the prostate are usually hyperechoic relative to the peripheral zone and symmetrical in size and shape.

44 Fig.l (vii)

Anterior fibromuscular / stroma

Inner zone

Outer zone

Seminal vesicle

Fig,1 (vii) Diagram of the longitudinal zonal prostatic

anatomy (BPH) and the corresponding transrectal ultrasound

image. The peripheral (outer) zone and central (inner)

zones, seminal vesicle (SV) and bladder (HI) can be seen.

45 Fig.l (viii)

Anterior fibromuscular Inner zone C stroma

P Q

Outer zone

Fig.l (viii) Diagram of the tranverse zonal anatomy of the prostate through line A in Fig.l (vii) and the correspond ing tranverse transrectal ultrasound image.

46 The ultrasound appearances of prostate cancer initially were the subject of some controversy. However recent studies, utilizing newer technology and anatomical infor mation, have shown that prostate cancer, rather than being hyperechoic as originally described, is actually hypoe choic or echopenic. Correlation of pathological abnormali ties of the prostate with sonographic patterns was carried out by Egender et al. (1984) who studied fresh cadaver prostates. They concluded that cancer was echopenic, thus supporting their in vivo experience which had shown that over 70% of patients with prostatic cancer had echopenic lesions.

Dahnert et al. (1986), showed that BPH had a variable and irregular hypoechoic pattern in the central zone of the gland, whereas prostate cancer showed asymmetry with a tumour bulge posteriorly, and in 76% of cases, malignant lesions were hypoechoic. There was no case where cancer was hyperechoic. In Lee's series (1987) biopsy of relat ively hyperechoic lesions all demonstrated BPH.

Ultrasound has become an increasingly important imaging modality for the visualization of many internal soft tissue structures and the size, location and structure of the prostate make it one of the most suitable organs for the diagnostic application of ultrasound. It is small enough for ultrasound to penetrate and its proximity to the rectum allows insertion of a probe; moreover the clear definition of its internal anatomy allows accurate biopsy of its different regions and tumours.

47 IV ENDOCRINOLOGY / BIOCHEMISTRY

1.Historical Background

(1) Steroid Hormones In growth control

Steroid hormones are Important physiological factors In controlling cell division and proliferation, and they promote long-term responses such as the maintenance of growth and function In their target tissues. John Hunter

(1786) was the first to recognise the Importance of the testes In the maintenance of the growth of the prostate long before the concept of steroid hormones had been postulated. The association of endocrine function with abnormal or neoplastic growth was first recognised by Sir Astley Cooper (1836) In his book "Lectures on the Princi ples and Practice of Surgery". He noticed that there was a correlation between the growth of the human breast, and the phase of the menstrual cycle. In 1896, Sir George Beatson, a Glasgow surgeon, described two patients with Inoperable breast carcinoma, whose tumours significantly regressed following bilateral salplngo-oophorectomy. This may have been the earliest demonstration of an endocrine responsive tumour. However this finding preceded the Isolation of oestrogen from ovarian tissue and consequent ly the explanation for Beatson*s observation was not realised at the time.

48 Further advances In understanding of endocrine dependence of tumours were made following the isolation of steroid hormones. Doisy et al. (1929), isolated oestrone from the urine of pregnant women, and MacCorquodale et al. (1936) isolated a crystalline oestrogenic hormone from sow's ovaries. Laccasagne (1932), at the Institut Curie in Paris suggested a possible correlation between gonadal hormones and the development of a neoplasm. He showed that the injection of oestrone into three males of a particular strain of mice provoked a mammary cancer in each case. It is now clear that hormones which regulate the growth and function of hormone dependent tissues, have similar ef fects on cancers arising from these tissues; namely breast (Muhlbock 1972), prostate (Huggins 1967) and endometrium (Swenerton 1982). There is little evidence to support hormones as primary mediators of neoplasia resulting from a hormonal imbalance alone, but the way they may influence this process was summarised by Berenblum (1978). They may have a preparative effect, rendering the cells sensitive to other stimuli ; they may have a permissive effect on the mitogenic process itself; or they may have a condi tioning effect on the induced cellular proliferation. However, a prolonged phase of abnormally rapid growth may occur in the presence of chronic endocrine imbalance and this may eventually lead to the promotion of carcinogen esis (Furth 1975).

49 (ii) Steroid hormone action on the prostate

John Hunter (1786) contrasted the normal prostate of animals "large and fleshy", to that of castrated animals

"small, flabby, tough, ligamentous and little secretion".

Although lacking a specific diagnosis. White (1895) trea ted 111 men with bladder outflow obstruction by castra tion, and demonstrated rapid atrophy of the enlarged prostate gland. Cabot (1896) in a study with longer follow up reported improvement in symptoms in 84 % of men with

BPH following orchidectomy. A Russian sect, the Skoptzs, who were castrated at 35, were reported to have small prostates and no symptoms of BPH (Zuckerman 193 6).

That castration or absence of testicular function could have a preventative effect on the development of BPH was also demonstrated by Moore (1944). He studied the pros tates of 28 patients who were older than 45, but were either eunuchs, eunuchoid or had pituitary infantilism and had lost their secondary sexual characteristics before the age of 40. He serially sectioned their prostates but there was no evidence of BPH in any of this group.

However a further advance in the realisation that sex hormones could influence the prostate and the growth of other normal and neoplastic tissues, resulted from the studies of Huggins and Clark (194 0) . They demonstrated that orchidectomy or the administration of oestrogens to reduce circulating androgens in dogs, caused reduction in size of the normal and hyperplastic prostate gland and

50 also Inhibited the secretion of prostatic fluid. The same year Huggins and Stevens (1940) described 3 men with BPH who were treated by castration. One had a marked reduction in prostatic size and all three were shown to have epithe lial atrophy on subsequent biopsy. This evidence supported the view that the prostatic epithelium at least, was under the control of the testes. They concluded that since the normal target organ is dependent upon hormonal support for growth and metabolic activity then cancer derived from it may be similarly dependent. In a now classic paper (Hug gins and Hodges 1941) it was demonstrated that castration or oestrogen injection caused regression of prostate cancer and so began a new era in endocrine manipulation of tumours.

2. Studies of Steroid Hormones in the Pathogenesis of BPH (i) Plasma levels of steroid hormones

Many groups have investigated blood hormone levels in BPH patients and normal men, in an attempt to identify differ ences that might lead to conclusions regarding hormonal involvement in the aetiology of BPH. This has yielded very confusing data, and reports of alterations in hormone levels with ageing in normal men have also been inconsis tent. Harman and Tsitouras (1980) state that there is no change in plasma testosterone with ageing. However the attenuation of the early morning rise in testosterone is

51 lost with age (Bremner et al. 1983) and measurement of 24 hour mean plasma concentrations of testosterone shows a continuous decline with age (Zumoff et al. 1982).

In patients with BPH Hulka et al. (1987) found elevated serum testosterone levels but many others report no significant difference in comparison to normal controls (Hammond 1978, Bartsch et al. 1979, Drafta et al. 1982, Brochu and Belanger 1987). However several studies (Vermeulen et al. 1972, Pirke and Doerr 1975, Horton 1976, Rannikko and Aldercreutz 1983) report lower levels of testosterone in BPH. Although similar variations have been found for oestrogens, Zumoff et al. (1982) who measured 24 hour mean oestrone and oestradiol in normal men found no alteration with age. Thus they proposed that a relative rise in oestrogens may be involved in the pathogenesis of BPH, although two groups have also reported an absolute in crease in oestradiol with age (Rubens et al. 1974, Pirke and Doerr 1975).

This approach has yielded no general hypothesis explaining the role of steroid hormones in the pathogenesis of BPH and there is certainly little to suggest any primary endocrine abnormality which is reflected purely in a plasma hormone level.

52 fii) Tissue levels of steroid hormones

Several investigators have measured the tissue concentra tions of various steroid hormones in BPH tissue and compared these with "normal" prostate tissue. Siiteri and

Wilson (1970) found no significant difference in testosterone levels but noted a four-fold increase in dihydrotestosterone (DHT) levels in BPH tissue. Interest ingly they also found higher DHT levels in the central periurethral areas (where BPH characteristically commences), than in the outer regions of the same gland.

This may be important as Farnsworth and Brown (1963) found that DHT and androstenediol were the principal metabolites of radioactive testosterone incubated with slices of human

BPH. It therefore appears that testosterone is a prohor mone, and the active form of the androgen in the prostate is DHT, produced by the action of the enzyme 5 alpha reductase (Bruchovsky and Wilson 1968 a,b, Anderson and

Liao 1968). In contrast to most animals high levels of DHT formation persist in the dog and in man throughout life

(Wilson and Gloyna 1970). Since these are the only two species to spontaneously develop BPH, this highlights the likely importance of DHT in prostatic growth control and speculation on its role in BPH. Subsequently several groups also found increased local DHT levels in BPH

(Geller et al. 1976, Hammond 1978, Meikle 1978, Kreig et al. 1979). They also reported that depletion of androsta

53 nediols and androsterone in parallel with an increase in DHT. Moore et al. (1979) showed increased prostatic DHT concentrations in androgen induced prostatic hyperplasia in the dog.

Walsh et al. 1983, have questioned these findings, since they found a similar DHT content in normal peripheral and

BPH tissue obtained at open surgical procedures. A review of the literature reporting an elevated local DHT content in BPH reveals that in all series normal tissue was taken from autopsy specimens. Incubation of fresh prostatic tissue at 37°C even for a few hours, results in consider ably reduced DHT levels, and therefore in all these series the normal DHT levels were probably highly underestimated. This casts considerable doubt on the role of DHT as a major causative factor in the pathogenesis of BPH. In fact experiments in age-matched beagles, demonstrated no dif ference in DHT content between dogs with histologically normal prostates and those with spontaneous BPH (Ewing et al. 1983). This has led to the search for other factors which may sensitise this tissue and accelerate growth.

(iii) Androgen Metabolism

The most important step in androgen metabolism is the conversion of testosterone to dihydrotestosterone (DHT), which is catalysed by the NADPH dependent 5-alpha reduc tase enzyme in an irreversible reaction. DHT can be meta-

54 bollseâ further by 3-alpha or 3-beta hydroxysteroid dehy drogenase into the corresponding androstanediols and this step is known to be reversible. However the irreversible conversion of 3-beta androstanediol to the triol compound may play a role in the elimination of androgenic activity. The enzyme 17-hydroxy steroid dehydrogenase reversibly catalyses the interconversion between testosterone and androstenedione and DHT and androstanedione. Any change in any of these enzyme activities might cause a shift in the concentrations of several metabolites. In view of the now controversial reports of a rise in DHT in BPH tissue, studies have tended to concentrate on 5- alpha reductase (DHT formation) and 3-alpha/beta dehydro genase (DHT removal). Most of these studies indicate an increased activity of 5-alpha reductase (Morfin et al. 1978, Kreig et al. 1979, Bruchovsky and Lievovsky 1979, Isaacs et al. 1983), with two of these (Bruchovsky and Lievovsky 1979 and Isaacs et al. 1983) also reporting a decrease in 3-alpha/beta hydroxysteroid dehydrogenase and or 17-hydroxysteroid dehydrogenase. Such changes would lead to an increase in intracellular DHT. However these results must be treated with caution as they may be dependent on tissue heterogeneity and regional variation within the gland. Interaction between epithelium and stroma may be important in the pathogenesis of BPH and several studies (Cowan et al. 1979, Wilkin et al. 1980 and

Kreig et al. 1983) have reported higher 5-alpha reductase

55 concentrations in the stromal than in the epithelial fraction, although Habib et al. (1983) did not support this, and found that the capacity to metabolise testoster one was evenly distributed between stroma and epithelium.

These studies are influenced by the techniques used for the separation of tissue fractions and this may account for some of the differences. If there is an uneven distri bution of some of the enzymes between tissue fractions, then the tissue composition (i.e. epithelial/stromal ratio) could influence the metabolic capacity of the tis sue. In a large biopsy series (Schweikert et al. 1985), no correlation between 5-alpha reductase activity and epithe lial content could be found.

(iv) Role of Oestrogens in the Pathogenesis of BPH

Many researchers have speculated that oestrogens may have a stimulatory role in the development of BPH. Walsh and

Wilson (1976) demonstrated that oestrogens were essential for the development of BPH in the dog. Castrated dogs treated with androgens plus oestrogens developed large prostates, whereas those treated with androgen alone did not. De Klerk et al. (1979) showed that BPH could be experimentally induced with a combination of oestradiol and DHT. Many other authors (Thompson et al. 1979, Murphy et al. 1980 and Krieg et al. 1981) have also reported the preferential binding of oestrogens in the stroma of

56 BPH, and their activity in the development of hyperplasia. Only one study has indicated that BPH may be a result of an absolute increase in oestradiol in ageing men (Rannikko and Aldercreutz 1983). However an increase in the ratio of oestrogen to androgen, due to a fall in testosterone (Zumoff et al. 1982) or a rise in sex hormone binding globulin (SHBG) which reduces available androgen is a more likely stimulus for the development of BPH. Oestrogens may be produced locally in the prostate or peripherally (adipose/muscle tissue) by aromatisation of androstenedione (forming oestrone) or testosterone (forming oestra diol ). This conversion is catalysed by the cytochrome P450 dependent enzyme complex, aromatase. Certainly the experi ments of Schweikert (1979) using cultured BPH fibroblasts, and Kaburagi et al. (1987) and Stone et al. (1986), using whole tissue homogenates provide some evidence for local oestrogen production. In contrast other studies (Bartsch et al. 1987) have failed to show any endogenous aromatase activity in BPH tissue, by two different aromatase assays (oestrone formation and tritium release). This has recent ly been confirmed by J.Davies (1991 personal communica tion), who was also unable to find any increase in speci fic mRNA.

There are several reports of the beneficial effects of aromatase inhibitors in patients with BPH. These inhibi tors, such as 4-4 OH-4 androsten,3,17-dione and testolactone, block the peripheral conversion of testicular and

57 adrenal androgens. Tunn et al. (1985), studied the effects of treating patients with BPH with testolactone, a weak aromatase inhibitor and although this was not a randomised double blind study and the numbers were small they noted improvement in 60% of patients (7/13). More recently the same group (Schweikert and Tunn 1987) showed that prostatic size was significantly reduced (15-26%) in 50% of patients taking testolactone and that compared to a control group, symptoms of outflow obstruction were im proved. However these results must be treated with caution as the observed effect may be due to inhibition of 17-beta hydroxysteroid dehydrogenase rather than of aromatase. This would result in a decrease in the formation of oe stradiol from oestrone and of testosterone from androstenedione. Coffey (1988) recently confirmed that inhibition of endogenous aromatase activity was not an effective treatment for spontaneous BPH in the dog. Local aromatisation of androgens in the prostate remains a controversial issue, but does not exclude an important role for oestrogens in the regulation of prostatic growth. Quantitative morphometric analysis shows that stroma is the predominant tissue in BPH (Bartsch et al. 1979) and it is known that there is an increase in oestrogen receptors and tissue oestradiol in the stroma in BPH (Kreig et al.

1981). The importance of oestrogens is underlined by the significant reduction in prostatic size and metabolic activity which follows oestrogen withdrawal, although direct proof of this effect is still lacking.

58 Further studies using the specific cDNA probe for the aromatase mRNA would enable determination of any gene overexpression either due to increased transcription or gene amplification. This would help clarify whether or not aromatase inhibition was a worthwhile therapeutic man oeuvre in BPH.

3) Mechanisms of steroid hormone action

Steroid hormones circulating in the blood are bound predominantly to plasma proteins such as albumin, specific steroid binding globulins and alpha-feto protein; each of these carrier systems has a characteristic affinity and capacity. Testosterone, the major androgen in the circula tion, and oestrogens are transported largely bound to sex hormone binding globulin (SHBG) while progesterone circu lates predominantly bound to corticosteroid binding globu lin or transcortin. The delivery of hormone to specific target tissue is potentially a controlling mechanism and variation in the concentration of SHBG could be one of the factors that influence the biological effects of androgens on the prostatic cell. No control mechanism at the cell membrane has been found and testosterone and oestrogen are assumed to enter the cell in the unbound state by passive diffusion (Higgins and Gehring 1978), although some faci litation may occur (Baulieu 1978). A major advance in the understanding of sex steroid hormone action occurred when

59 Jensen and Jacobsen in 1962 at the University of Chicago synthesised tritiated oestradiol with a high specific activity and demonstrated its selective uptake and reten tion in target organs (pituitary, uterus and vagina) of female rats. The difference between target organ and non target organ is the ability to retain steroid (Peck et al. 1973).

( i) Receptors

Toft and Gorski (1966) identified an oestrogen binding protein in the cytoplasm of rat uterine cells and referred to this as an oestrogen receptor. In 1968 Gorski et al. and Jensen et al. independently suggested a two step model for the mechanism of interaction of a steroid hormone with its target cell (Fig.l (ix)). In this model the steroid enters the cytoplasmic compartment of the target cell, and binds to its specific receptor protein. These receptors are characterised by their specificity, their high affin ity and low binding capacity (Giannoupolous and Gorski

1971). They have also been demonstrated to be heat and pH labile (Toft and Gorski 1966, Myatt et al. 1979). The cytoplasmic steroid-receptor complex has to undergo acti vation which is a temperature sensitive process by which it acquires increased affinity for nuclear acceptor sites

(Buchi and Villee 1976). The steroid receptor site is then translocated to the cell nucleus (Shymala and Gorski 1969)

60 where it binds to specific nuclear sites, "acceptor" sites on the chromatin (Puca et al. 1974). The interaction of the activated cytoplasmic steroid-receptor complex with the nuclear acceptor sites results in the activation of

RNA polymerases and an increased rate of transcription of specific DNA segments. This includes specific messenger RNA (mRNA) (Borthwick and Smellie 1975) which is transpor

ted to the cytoplasm and results in synthesis of new cellular proteins (Edwards et al. 1980) and often in cell growth and division. Following translocation the depleted cytoplasmic receptor pool is replenished possibly by new synthesis (Dix and Jordan 1980), although receptor recyc ling probably also occurs (Clark and Peck 1976).

F i g . l ( i x )

SHBG

JT ACTIVAT! lECEPTOR COMPLEX ▲ FREE CYTOPLASMIC FREE RECEPTOR LIGAND TRANSLOCATION —)LACCEFTOF SITE CELL DNA SYNTHESIS DIVISION RNA

PROTEIN SYNTHESIS

DIAGRAM OF THE TWO STEP' MODEL FOR THE

MECHANISM OF ACTION OF STEROID HORMONES 61 The mechanism for androgen action based on the two step model for steroid receptors was reviewed extensively in a monograph by Mainwaring (1977), and was accepted as the mechanism by which androgens influenced prostatic cell biochemistry until recently. In the light of further studies the two step model for steroid receptor action has been reconsidered. King and Greene (1984) and Welshons et al. (1984) using immunocytochemical and biochemical techniques have been unable to identify a cytoplasmic receptor, and have proposed that the only intracellular receptor is located within the nucleus (one-step model)

(Fig.l.(x). Welshons et al. (1984) have suggested that the cytoplasmic receptor may be an artefact of tissue prepara tion.

In support of the one step model are the observations on the interaction of androgen with intact nuclear envelopes prepared from rat ventral prostate tissue (LePebvre et al.

1985) and the exclusive nuclear localisation of 5-alpha reductase in the human prostate (Houston et al. 1985,

Habib and Chisholm 1989). However there is no doubt that unoccupied receptor which is only weakly nucleus associa ted, is readily identified in cytosolic preparations.

There has now been a return of support for a cytoplasmic steroid hormone receptor and it is thought that the anti bodies used by Greene and co-workers may fail to recognise the 8s oligomeric complex of ER (Pratt et al. 1989, Jensen

1991).

62 Fig.l (x)

SHBG d)

TESTOSTERONE s a REDUCTASE

RECEPTOR

CELL DIVISION mRNA DNA SYNTHESIS

PROTEIN SYNTHESIS

DIAGRAM OF THE MECHANISM OF ANDROGEN ACTION BASED ON THE ’ONE STEP’ MODEL

(11) Androgen receptors In BPH

Despite the technical difficulties Involved there have been many attempts to derive a quantitative estimate of androgen receptors (AR) In prostatlc tissue. The majority of these have focussed on the amount of AR In prostatlc carcinoma In the hope that this may give a guide to the hormone sensitivity of the tumour. Unfortunately these results have been generally disappointing, probably due to tumour heterogeneity, although the presence of AR In BPH

63 has been shown by several groups (Van Aubel 1985). Early attempts to assay AR depended on measuring the specific binding of the tritiated natural androgen testosterone or dihydrotestosterone (DHT) (Mainwaring 1969 Steins et al. 1974). However in human tissues both these steroids bind to sex hormone binding globulin (SHBG) (Cowan et al. 1975, Mobbs et al. 1975). In addition it has been shown by Chamberlain et al. (1966) and by Bonne and Raynaud (1976) that DHT may be metabolised to androstane diols which have low affinity for androgen receptors. In 1975 Bonne and Raynaud claimed to have overcome these problems with the introduction of methyltrienolone (RlBBl), a synthetic androgen, which they claimed was not metabolised and did not bind to SHBG. Brown et al. (19B1) confirmed this resistance to metabolism. It was subsequ ently discovered that RlBBl bound avidly to progesterone receptors (PR) as well as AR (Asselin et al. 1976, Dube et al. 1976, Cowan et al. 1977). This affinity can be elimina ted by blocking RlBBl binding to PR with excess triamcino lone acetonide (Zava et al. 1979), but this in turn may inhibit RlBBl binding to AR (Shain et al. 19B0, Miller et al. 19B3). In an attempt to minimise these problems Steb- bings et al. (19B7) used a combined assay with tritiated DHT and RlBBl to overcome SHBG and PR binding using a

"hybrid ligand” technique.

Despite these techniques the methodological difficulties are still considerable and Ghanadian (19B2) in his review

64 also pointed out the labile nature of the DHT-receptor complex, especially with temperature change, the near

saturation of endogenous receptor with DHT and slow disso ciation of DHT from its receptor, making any exchange assay very difficult. The prostate is also rich in pro tease enzymes that can rapidly degrade many other enzymes and receptors and the prostatic connective tissue prevents disruption of functional cells.

Notwithstanding all these problems and the possibility that cytoplasmic receptors are the result of nuclear contaminants resulting from homogenate preparation (Ekman et al. 1982, Trachtenberg and Walsh 1982, Conor et al.

1984, Welshons et al. 1984), studies to characterise the androgen receptor in prostatic tissue have been carried out (Barrack and Coffey 1980). Nuclear androgen receptor binding sites exist as two measurable fractions, one

fraction extractable by high ionic strength (6M KOI) and the other resistant to salt treatment. Although some workers have shown a correlation between salt extractable nuclear androgen receptors and response to endocrine therapy in carcinoma of the prostate (Brendler et al.

1984) , there was considerable overlap in AR concentrations between responders and non responders.

The earlier work of Barrack and Coffey (1980) which demon

strated receptors in the salt insoluble, chromatin deple ted nuclear matrix, also implicated this nuclear subfrac tion in DNA and RNA synthesis and nuclear binding of

65 steroid hormones. Several workers have since characterised the matrix bound receptors and measured their concentra tions in BPH (Barrack et al. 1983, Conor et al. 1984). In studies by Walsh and his colleagues (Barrack et al. 1983), assays in all three subcellular compartments showed eleva ted levels of AR in both nuclear salt extractable and nuclear salt resistant fractions in BPH tissue (51% and

41% respectively). This supports the findings of other groups (Trachtenberg et al. 1981, Isaacs et al. 1983) who showed that total, salt extractable and salt resistant AR levels were higher in BPH than in normal tissue, but con flicts with the earliest reports that the AR content of the nucleus and cytoplasm was similar in normal, BPH and peripheral prostate (Bruchovsky and Wilson 1968a). Kirdani

(1984) using high performance liquid chromatography could find no difference in AR levels between BPH and carcinoma specimens and concluded that their measurement would be of no diagnostic value. However analysis of multiple biopsies from the same gland, indicate that considerable variation may occur within the prostate: coefficients of variation ranged from 30% to 109% (Van Aubel et al. 1985). Part of the heterogeneity is probably due to the nature of the tissue composition. The level of AR correlates with the percentage of epithelium, which can vary (Lahotnen et al.

1983), and autoradiographic studies have shown that AR are located predominantly if not exclusively in the epithelial nuclei of BPH tissue (Peters and Barrack 1987).

66 This has been confirmed by the most recent studies using a new monoclonal antibody to androgen receptors which also shows AR mainly in the nuclei of epithelial cells (Demura et al. 1989).

(iii) Oestrogen and Progesterone receptors in BPH

(a) Oestrogen receptors fER)

It now seems unlikely that the prostate peripherally converts significant amounts of androstenedione or testos terone to oestrone or oestradiol (Smith et al. 1982) . It is possible therefore that the bulk of the oestrogen is derived from circulating blood and preferentially retained by oestrogen receptors in prostatic tissue. Although the assay of ER in prostatic tissue has been controversial, again due to methodological difficulties, their presence has been established by various laboratories (De Voogt and

Dingjan 1978, Ekmann et al. 1979a, Karr et al. 1979,

Tilley et al. 1980, Pontes et al. 1982, Belis et al.

1983) . However the levels of ER in BPH cytosol are low

(usually less than 20 fmol/mg protein) (Kirdani et al.

1984) and when comparisons have been made both cytosolic and nuclear ER are lower in BPH than normal tissue (Trach tenberg et al. 1980, Ekmann et al. 1979b) and considerably lower than ER in canine prostate (Walsh 1984). By con trast, Auf and Ghanadian (1982) have demonstrated very high nuclear ER values in BPH tissue.

67 Measurement of these receptors may be affected by the amounts of different types of tissue within the prostate. Trachtenberg et al. 1980 found that ER were most numerous in the stromal rather than the epithelial cells within the gland. Kreig et al. (1981) reported that although androgen binding sites were evenly distributed in the cytosol of the epithelium and stroma from human BPH, ER were more frequently detected in stromal specimens analysed (42%) than in cytosol isolated from epithelial cells (5%).

(b) Progesterone Receptors

There is some evidence that progesterone may be important in the development of BPH. However again these results vary and are not always similar to other tissues, such as uterus and breast, where oestrogens are known to induce progesterone receptors and signal a functioning oestrogen receptor.

The role of progesterone is intriguing as it is a better substrate for 5-alpha reductase than testosterone itself, and many progesterone analogues such as cyproterone ace tate, are quite effective in blocking androgen binding to the receptor. Progesterone receptors (PgR) approach andro gen receptor concentrations in BPH, but are found only in the cytoplasm (Ekmann et al. 1982). In the lateral prostate of the rat, specific cytosol progesterone binding can be increased by treatment of the animals with

68 oestradiol and a combination of progesterone and oestra diol causes growth (Belis et al. 1984). The effects of progesterone are not mediated via the androgen receptor suggesting that the cytosolic progesterone receptor is capable of mediating a growth response to an agonist in the lateral prostate. Some studies (Bashilerahi et al.

1983) demonstrate the presence of PgR in most cases of BPH although they are less common in normal prostatic tissue.

However Belis et al. 1983, found that PgR were present in the cytosol of human BPH and normal prostate in similar concentrations. If the presence of PgR signals a function ing oestrogen receptor this fails to provide evidence for preferential ER stimulation in human BPH (Ekman et al.

1982) . It is possible that PgRs in BPH are a secondary phenomenon due to changes in androgen/oestrogen blood levels and it may be that progestational anti-androgens are more logical drugs to consider in control or reduction of BPH (Habib 1989).

69 4. Other factors regulating prostatlc growth

It is clear from the evidence discussed above that andro gens are not the sole regulators of prostatic growth.

Other factors act as mitogens acting either synergistical- ly with androgens or alone. This was elegantly demonstra ted in 1946 when Huggins and Russell reported more marked atrophy of the prostate after hypophysectomy and castra tion, than after castration alone. The pituitary related

factor has since been identified as prolactin and further evidence has accumulated to suggest that prolactin has a specific role in the control of prostatic function (Gray- hack et al. 1955, Grayhack and Lebowitz 1967). Lostroh and

Li (1957) reported a similar effect in the rat, and found that the administration of growth hormone (GH) or ACTH to hypophysectomised castrated rats stimulated prostatic growth whereas FSH, LH and prolactin did not. However, the simultaneous treatment of castrated rats with ACTH and prolactin produced greater prostatic growth than ACTH alone (Tullner 1963).

(i) Prolactin

The mechanism of prolactin action is unknown, although

Laurence and Landau (1965) reported impaired uptake of testosterone in hypophysectomised rats. It is possible that prolactin increases the cellular influx of androgen and enhances its availability (Farnsworth 1988). Prolactin

70 and its receptor have been confirmed immunocytochemically by specific staining (Harper et al. 1981, Witorsch 1981) and biochemically (Aragona and Friesen 1975, Leake et al. 1983). It has been suggested that androgens and prolactin act in synergy. This has been supported by the positive correlation between blood prolactin concentrations and androgen receptors (Odoma et al. 1985).

Other studies have observed the synergistic effects of prolactin and androgen on the growth of benign prostatic epithelial cells in culture (Yamanka et al. 1977). Even in cells maintained in steroid free medium, prolactin stimulates growth so this cannot merely be due to in creased androgen uptake (Syms et al. 1985). Since prolac tin increases testicular testosterone production by acting in synergy with LH (Hafiez et al. 1972), it also has an indirect effect on the prostate. Furthermore the increas ing oestradiol/testosterone ratio influences prolactin levels, due to the stimulatory action of oestrogen on prolactin release (Boyns et al. 1974). Finally there is good evidence that low to moderate levels of prolactin potentiate the trophic effects of testosterone on prostatic growth, but at high concentrations it becomes an inhibitor of growth (Sandberg 1981, Syms et al. 1985).

( ii) Growth Hormone

Although the specific role of growth hormone (LH) in the prostate has yet to be found, the presence of GH has been

71 detected lifimunocytochemlcally and the gland demonstrates a capacity to bind exogenous levels of the peptide (Sibley et al. 1984). In this study the GH bound primarily to the

stroma in both benign and malignant tissue although there was some evidence of binding in the epithelial cells. A similar study has been reported in the canine prostate (El Etreby and Mahrous 1979). Patients with metastatic carcinoma of the prostate have significantly raised plasma levels of growth hormone (British Prostate Group Study, 1979). This raises the possibility that growth hormones might be involved in the spread of cancer cells. The specific role of GH in the prostate has yet to be found, but growth hormone exerts its action upon all tissues via intermediate growth fac tors (insulin like growth factor or somatomedins). There are anecdotal reports that male acromegalic patients have larger prostate glands, but this may also depend on their ability to produce androgens since they can also have reduced gonadotrophin production, due to pressure from the pituitary tumour. The functional significance of GH binding sites in the prostate requires further investiga tion. GH stimulates the breakdown of collagen and proli

feration of fibroblasts (Deyl et al. 1976) and somatomedin is produced from fibroblasts in response to GH stimulation (Atkison et al. 1980, Clemmons et al. 1981). Also growth hormone binding sites have been characterised on mouse fibroblasts using iodinated GH (Murphy et al. 1983).

72 5. Stromal-Epithelial Interaction

The possible role of stromal-epithelial Interaction In the pathogenesis of BPH was first suggested as early as 1925 by Relschauer. His microscopic observations led him to believe that the Initial change In the formation of BPH was In the stroma, rather than In the epithelium, with the formation of flbromyomatous nodules. These nodules subse quently stimulated growth of epithelium and penetration of glandular elements from the periphery. This basic concept was reiterated by Le Duc (1939 ) and Demlng (1939 ) who analysed prostatlc histological specimens from men of various ages. Franks (1954) observed that from an anatomi cal viewpoint stromal and epithelial hyperplasia may occur alone or together and that It Is Impossible to say that one occurs before the other. He later showed that cytodlf- ferentiatlon and hormone sensitivity were dependent on association of epithelium with adjacent stroma (Franks and

Barton 1960) and that prostatlc epithelial cells required stroma for their growth (Franks et al. 1970). Separated epithelial cells could not be maintained In culture and did not synthesise DNA, although epithelial cells from explant cultures of undlssoclated prostatlc tissue can be maintained In vitro for many months. Further experimental evidence for an Inductive Influence of accessory sex organ stroma came with studies on embryonic mice (Cunha 1972).

73 It was soon proposed that the maintenance of adult morpho logy and function of organs composed of epithelium and stroma was dependent on a continued interaction between these tissues (Tarin 1972).

In contrast to previous studies McNeal (1978) did not find stromal nodular genesis to be the primary lesion in

BPH, but instead found glandular budding and branching toward a central focus. This process in the embryo is triggered by androgens and leads to budding of ducts from the urogenital sinus into the surrounding mesenchyme. A similar process occuring in adulthood led to the sugges tion that a "reawakening" of the inductive potential of prostatic stroma may occur in BPH (McNeal 1978). Sex hormones seem to have their main influence on the prostate after sexual maturity. During embryogenesis the normal growth and development of the gland is influenced partly by hormones but also by stromal-epithelial interactions (Cunha 1976, Cunha and Lung 1979). The stromal cells in BPH certainly have some of the morphological characteris tics of undifferentiated mesenchyme (Franks 1975), and there has been some speculation that possibly an altered endocrine environment with age leads to reactivation of the stroma. This may affect certain stromal-epithelial interactions which normally operate during the onset of prostatic development, resulting in BPH.

Several elegant experiments have provided good evidence that the mesenchyme of the urogenital sinus could be

74 responsible for inducing growth and cytodifferentiation of the prostatic epithelial cells (Cunha et al. 1983, Chung and Cunha 1983).

Bladder epithelial cells when placed alongside the mesen chyme of the urogenital sinus induces development of prostatic epithelial cells. The response of the epithelial cell and the amount formed are related to the mass of the mesenchymal tissue present. The epithelial tissue grows until the mesechyme is completely covered. The opposite, an excess of epithelial cells does not cause an excess growth of mesenchyme. This phenomenon has led Cunha and his colleagues to believe that the target cell for andro gen stimulation is the mesenchyme and not the epithelial cell (Cunha et al. 1983). Chung et al. (1984) demonstrated a large overgrowth of the adult mouse prostate gland with formation of new acini by implantation of foetal urogeni tal sinus into the gland. Although this supports McNeal's hypothesis (1978 ) it also raises the question of whether direct contact with the insoluble embryonic matrix or a soluble factor is responsible for inducing this over growth.

A new aspect was brought into this discussion by Muntzig

(1979) who from his animal work first suggested that the collagen matrix might limit growth of the epithelial cells. Marriotti and Mawhinney (1981) showed that the increase in collagen was counteracted by a breakdown of collagen through the action of the collagenolytic enzyme

75 secreted from the glandular cells. Treatment with cis- hydroxyproline, which inhibits collagen degradation, also

inhibits the proliferation and normal differentiation of

epithelial cells (Thornton et al. 1985). However, although the collagenous components of the stroma may be important

in supporting prostatic growth, no clear cause and effect has been demonstrated.

The idea of a soluble factor or factors being responsible

for inducing cellular proliferation has become increasing

ly popular since the isolation of a polypeptide growth

factor by Cohen (1962). Many adult tissues have growth

factors that can be demonstrated using in vitro fibroblast assay systems and in 1979 Jacobs et al. demonstrated a growth factor in crude extracts of normal, BPH and cancer ous human prostates. The recent studies by Chung et al.

(1988) have shown for the first time that tumorigenic processes could be activated across the cell membrane barrier through cell to cell interaction. Following inocu

lation of tumourigenic fibroblasts and non-tumourigenic prostatic epithelium into the subcutaneous space of rats, mixed adenocarcinomas and fibrosarcomas were induced which were androgen responsive. There is increasing evidence that chemical signalling, in the form of soluble growth

factors, is involved in cell to cell communication and phenotypic expression.

The concept that BPH results from an imbalance of sex steroids alone, fails to reconcile the clinical observa

76 tion that BPH is initiated during the third and fourth decades, whereas changes in the ratios of sex steroid concentrations does not occur in males until after the sixth decade (Wilson 1980). It seems more likely that a combination of systemic and local cellular factors such as a re-expression of embryonic growth potential (McNeal 1978), possibly mediated by growth factors, may be in volved in the induction of BPH with an abnormal endocrine environment permitting maintenance and proliferation of the disease once it has been induced.

6. Growth Factors

(i ) Introduction

In recent years the complex mechanisms which regulate vertebrate cell proliferation have gradually begun to be unravelled. This applies to normal cells, in which growth is tightly controlled, as well as to neoplastic cells, which may divide and proliferate in an uncontrolled fash ion. Molecular studies have led to the identification of a number of genes, all of which are functional in physiolog ical growth or embryogenesis in normal tissue, whose products are directly involved in regulating cell growth. Even more significant was the discovery that these same genes (cellular or proto-oncogenes), when overexpressed are capable of inducing a transformed phenotype.

77 The link between an oncogene and a cellular gene with a role in growth control, began with the isolation of a protein from mouse submaxillary glands by Cohen (1962).

This protein was found to accelerate incisor eruption and eyelid opening in newborn mice, and was called epidermal growth factor (EGF). EOF and its close structural homolo gue, transforming growth factor alpha (TGFa) are potent mitogens in many epithelial systems (Carpenter and Cohen 1979 ). These proteins are derived by proteolytic cleavage from large precursors and appear to mediate their actions, which are very similar, by interaction with the same cell surface receptor system (Todaro et al. 1980). Analysis of the EGF receptor revealed considerable sequence homology with the protein coded for by the v-erb-B oncogene of avian erythroblastosis virus (Downward et al. 1984). Many other polypeptide growth factors which stimulate the proliferation of fibroblasts or other cell types have since been described. Although the physiological role of growth factors is difficult to establish they are probably required for proliferation and differentiation of most cells during embryogenesis, growth, and wound healing. Normal growth control and subsequent disordered control involves multiple elements including genetic predisposi tion, mitogenesis in response to steroid hormones, growth factors and overexpressed growth factor receptors (Dickson et al. 1987, Cline et al. 1987). Growth control depends on the intimate relationship

78 between tissues, and this has been demonstrated in the prostate (Cunha et al. 1980). The processes of desmoplasia (stromal proliferation) and epithelial hyperplasia may be controlled by soluble mediators such as EGF and other growth factors acting in a paracrine fashion. Steroid hormones undoubtedly have an important influence on the development and proliferation of their target organs. More recently a great deal of evidence has arisen to suggest that the mitogenic effects of steroid hormones are partly mediated via peptide growth factors and their receptors (Lippman and Dickson 1989). So growth factors seem to play a crucial role in maintaining cellular communication and in the regulation of cell proliferation within tissues. Although many act as potent mitogens, they may also act as antagonists to cell proliferation. The concept of negative regulators is important as they provide a "shutdown" mechanism which counteracts proliferative stimuli when the optimal cell number is achieved. Although shutdown may be permanent, further proliferation may be invoked for cell replenishment after cell death or injury. Thus the posi tive and negative growth factors and their relationship to other stimuli such as steroid hormones provide a sophisti cated regulatory mechanism of growth control. Disturbance of this mechanism can lead to tissue growth, not neces sarily malignant if the process does not involve dediffer entiation or genomic abnormality.

79 (il) Epidermal Growth Factor

Human epidermal growth factor is a single chain polypep tide, comprising 53 amino acid residues with a relative molecular weight of 6045 daltons (Carpenter and Cohen 1979). Since it was first isolated 28 years ago (Cohen 1962), more than a dozen polypeptide growth factors which all stimulate cell proliferation have been described, but it is presently the best characterised epithelial cell mitogen. EGF is present in most body fluids but is found in particularly high concentrations in the colostrum (300 ng/ml), prostatic secretions (270 +/-30 ng/ml) and urine (30ng/ml). Its relatively low concentrations in plasma (< Ing/ml) suggests widespread production in glandular tissue. It is known to deliver its signal via specific transmembrane glycoprotein receptors (M.Wt. 170,000), which possess an intrinsic tyrosine kinase activity responsible for the transduction of signal to the cyto plasmic compartment (Carpenter 1987, Schlessinger 1988,

Hsuan et al. 1989). The EGF receptor protein has been purified (Yarden et al.

1985) and the complete amino acid sequence was deduced from the nucleotide sequence of cDNA clones (Ullrich et al. 1984). The mature receptor is composed of three major structural elements (Fig.l (xi)). The extracellular EGF binding domain is anchored in the plasma membrane by a

80 transmembrane region, which is linked to the cytoplasmic region containing the protein kinase domain.

Fig.1 (x i )

EXTERNAL DOMAIN EGF BINDING REGION

CYSTEINE RICH REGION

CELL MEMBRANE

CYTOPLASM

PROTEIN KINASE INTERNAL DOMAIN ACTIVITY

DIAGRAM OF EGF CELL SURFACE RECEPTOR

The primary event initiated by interaction of EGF with the receptor appears to be stimulation of the receptor tyro sine kinase activity associated with the intracellular domain of the receptor (Ushiro and Cohen 1980). This results in the phosphorylation and activation of a number of cellular proteins and in the phosphorylation of the EGF receptor itself (Margolis et al. 1985, King et al. 1988).

81 The mechanism by which binding of the EGF to the extra cellular domain of the EGF receptor stimulates the recep tor kinase activity in the cytoplasmic portion is not fully understood. The binding of EGF may induce a confor mational change in the extracellular domain and that in turn is transmitted through the membrane spanning region to the kinase domain which is then structurally altered and further activated. EGF has been found to induce recep tor aggregation and this itself may stimulate kinase activity (Yarden and Schlessinger 1987a). The binding of ligand is also thought to lead to an association with phospholipase-c and the production of inositol 1,4,5 triphosphate (IP3) (Schlessinger 1988), while protein kinase-c negatively modulates EGFR function (Cochet et al. 1984). At present the cellular substrates for the EGFR tyrosine kinase are not known, although it is capable of phosphorylating a number of intracellular proteins in vitro (Ghosh-Dastidar et al. 1984). The EGFR gene has recently been cloned and has consider able structural homology with other members of what is now regarded as a superfamily of membrane bound protein ki nases which include a series of related cellular oncogene products (Hunter and Cooper 1985, Yarden and Ullrich

1988).

82 (Hi) Epidermal Growth factor in the Prostate

There is now good evidence that EGF may be involved in the normal and abnormal growth of the prostate. EGF levels are high in prostatic fluid (Kavanaugh et al. 1983, Gregory et al. 1986) and seminal fluid (Elson et al. 1984). EGF has also been detected with immunocytochemical techniques in the cytoplasm of guinea pig prostatic epithelium (Shikata et al. 1984) and whole normal rat prostates (Maehama et al. 1986). A direct mitogenic effect of EGF with stimula tion of growth and proliferation of rat and human prostatic epithelial cells in vitro, has been reported

(McKeehan et al. 1984, Chaproniere et al. 1986). In common with other cell types (Carpenter and Cohen 1979), the activities of EGF in the human prostate are thought to be mediated through its specific receptor located on the

surface of the prostatic cell (Maddy et al. 1987). These receptors have been found by immunocytochemistry in the basal epithelial cells (Maddy et al. 1987), supporting the earlier suggestion that EGF may induce its biological effects both in vivo and organ culture by promoting pro

liferation of basal cell layers of epithelia of ectodermal

origin (Cohen and Taylor 1974). EGF receptors have been demonstrated in normal, hyperplastic and carcinomatous prostatic tissues (Maddy et al. 1987, Davies and Eaton

1989), however variations in receptor content are apparent between prostatic samples of different pathologies which

83 may account for differences in the activities of EGF-like growth factors in these tissues (Maddy et al. 1987).

In most biochemical studies, using radioligands binding to membrane preparations to assess receptor content, two classes of EGF receptor have been identified of high (Kd =

0.1-0.5 nM) and low (Kd = 2-5 nM) affinities. In epithe lia, high affinity binding has been associated with aggre gated, activated forms of these receptors, while low affinity binding has been assigned to monomeric less active forms (Yardin and Schlessinger 1987b, Boni-Schnetz- ler and Pilch 1987). EGF receptors have also been loca lised by immunohistochemistry in these tissues using a monoclonal antibody but this does not identify dimeric and monomeric forms separately. Similarly, distribution of receptor expression within tissues has been characterised by in situ hybridization (Fig.l (xii)). While some contro versy still surrounds the measurement of absolute concen trations of EGF receptors in prostatic tissues, normal,

BPH and prostatic carcinomas all contain high and low affinity receptors for EGF and all three tissue types are potentially responsive.

84 Fig.l (xll)

“ f

%KÏ»-:'À

EGF receptor localisation using in situ hybridization.

Receptors (dark silver grains) are visualised by autora diography after incubation with ^^P-labelled EGF receptor

cDNA. Receptors are predominantly located over regions of epithelium (E) as opposed to connective tissue (CT).

85 fiv) Epidermal Growth Factor in BPH

The precise role of EGF and related proteins in the aeti ology of BPH.is uncertain. Although one study has indicat ed that EGF is not the major mitogen in extracts of BPH tissue (Story et al. 1983), several other studies have demonstrated that the growth rates of epithelia derived from BPH (Kabalin et al. 1989) and from normal prostatic tissue (Eaton et al. 1988) are highly responsive to these proteins at low concentrations (1 ng/ml). BPH is a complex family of disorders involving the differentiated growth of both stromal and epithelial elements. It has been suggest ed that EGF receptors are not a major component of stromal cells and that EGF may not have a significant role in the development of stromal hyperplasia (Maddy et al. 1987).

Other growth factors belonging to the fibroblast growth factor (FGF) family have been implicated in this context

(Crabb et al. 1986, Story et al. 1987, Mydlo et al. 1988).

These factors stimulate the proliferation of both stromal and epithelial populations in vitro and are present at significant levels in prostatic tissues (Mydlo et al.

1988) . The importance of inductive interactions between stromal and epithelial populations has already been empha sised (Cunha et al. 1980). These processes may be mediated by the interplay of growth factor actions including those associated with EGF/TGFalpha.

Interestingly, a recent hypothesis (Tenniswood 1986)

86 includes the suggestion that androgens may be responsible for down regulating growth stimulatory factors derived from both stromal and epithelial cell types as the deve loping prostate progresses to normal adult status. This process may be controlled by the direct effects of andro gens or mediated by induced growth inhibitory factors. In experimental animals, castration results in a rise in prostatic EGF receptor levels suggesting that EGF pathways may be under some form of androgenic control in the adult gland and therefore involved in these processes (St.Arnaud et al. 1988). Failures, either temporal or absolute in these proposed feedback systems may lead to resumption of rapid growth similar to that of the embryonic prostate and could clear ly be central to the aetiology of BPH.

87 V MORPHOMETRY

1. Introduction

Morphometry or quantitative morphology is a means of quan tifying structural features. In biological terms this is a measure of the percentage of a tissue which is accounted for by a particular component. In the prostate this may include elements such as epithelium, lumen or stroma, but may be extended with electron microscopy to include subcellular compartments such as nuclei, mitochondria or endoplasmic reticulum. A quantitative knowledge of the make up of a tissue is essential in order to evaluate the significance of biochemical data obtained from that tis sue. The principles governing morphometric analysis, require that the structures to be quantitated have a homogeneous distribution within the tissue and are of similar size and shape from one part of the tissue to another. Furthermore the sampling of the tissue must be strictly randomised and sample preparation standardised (Weibel 1963, Weibel et al. 1969, Rohr et al. 1976).

2.Morphometric Techniques

The French geologist Delesse (1847) first suggested the principle behind estimation of volume occupied by various components, when analysing igneous rocks. This measurement

88 Is based on area measurement or planimetry of multiple random cross sections. He stated that the fractional volume occupied by a component was equal to the fractional area in cross section. This original method of calcula tion, involving planimetry, was replaced by point count ing, suggested as a more convenient method by another geologist Glagoleff (1933). This was later extended to biological specimens by Haug (1955) and Henning (1957). In histological specimens this procedure involves super imposing a point grid system over a microscopic section and counting is based on the number of points which coin cide with the component under investigation. When related to the total number of points projected over the entire section,the percentage of the component can be calculated. This method of achieving objective, quantitative values for a morphological structure is known as stereology, a term first coined by the International Society in 1961. The simple random sampling and subsequent statistical analysis that enables calculation of three dimensional structures from two dimensional cross sections, is de scribed by Weibel (1969). In order to count discrete tissue components, these stereological methods developed in geological and metal lurgical laboratories were applied to microscopical sec tions by Aheme and Dunnill ( 1966a,b, 1967, 1970). Initi ally they studied the morphometry of the human placenta

(Aheme and Dunnill 1966a,b) and later extended this to

89 muscle fibres (Aherne 1968) and the central nervous system (Aherne 1975). These techniques are slow and tedious and prone to observer error. Further variables could be intro duced by the type of grid chosen and its position on the

slide. The data produced was evaluated manually and pro duced results in a format which was difficult to inter pret. Nevertheless, these methods have until recently remained a valid and cost effective means of analysing tissue. A review (Hall and Fu 1985) raised interest in them again, since relatively simple computer programmes allow counting with rapid computation of results. However, with recent technological advances they are unlikely to remain the method of choice in the future. Instead, auto matic methods, employing computerised image analysis, could provide a means of accurate and rapid morphometric analysis of tissue sections (Cowan and Wann 1973, Sowter et al. 1987). However none of these morphometric techni ques applied to tissue sections is capable of measuring the absolute numbers of different cell types in the total organ (De Klerk and Coffey 1978). Only relative values can be provided due to the differential effects of fixation, dehydration and embedding of the sections on volume and shape of different tissue components (Bahr et al. 1957,

Peutilla et al. 1975).

90 3. Image Analysis

In life we are constantly analysing images and the combi nation of the human eye and brain is capable of performing very complex analyses far quicker than any computer. However we are unable to quantify what we see with any degree of accuracy or reproducibility. Image analysers are able to measure size, shape and optical density with accuracy and excellent reproducibility (Tucker 1979, Diamond et al. 1982, Heathfield et al. 1988). Simple image analysis has been carried out for many years without computers, using micrometer oculars or planimeters on projected images or photographs to determine areas (Aherne and Dunnill 1982). The computer began to be used to calcu late the parameters from traced objects on a digitised pad. Now TV or high resolution cameras input image data directly into the computer which functions as more than a simple calculator. The more advanced image analysers are capable of image processing to improve the quality of the image prior to analysis. This entails enhancing contrast within the image in order to highlight architectural features or to reduce shading. i ) Automatic methods

During the 1960's sophisticated, highly complex and expensive automatic computer based analysis systems began

91 to be used in microscopy (Lipkin 1966). An automatic analyser called the Quantimet, used in metallurgy was applied to biological research (Fisher 1967). Geological samples still provided the stimulus for the development of such machines, as rocks were more suitable to image analy sis than tissue sections, complicated by histological artefacts. A digitised tablet linked to a computer allows the out lines of objects to be drawn by hand on the tablet. The computer can then calculate the area and perimeter of each object from the coordinates of the outlined boundaries. This method is cheaper and easier to use than automatic machines, but less versatile. More complex and expensive automatic scanning systems are needed to measure parame ters such as cell density. These systems require skilled and experienced programmers and operators and were not available until the mid 1970Vs. Even now they are not established as a routine diagnostic aid and their use is generally confined to the larger research institutions.

(ii) Development of Image Analysis in Pathology

The component parts of an image analysis system are a computer, a microscope and a television camera. Therefore until the early 1950's, when computers capable of handling large amounts of alpha-numeric data were introduced, improved lenses had been manufactured by the Zeiss compa

92 ny, and the vidicon camera tube was first used. Integrated analysis was not possible. Initial expectations were high (Walton 1952) but it took a further twenty five years until these machines were capable of useful applications in pathology. Cole (1966) and Mawdsley-Thomas and Healey (1969) reported the first image analyser used with a microscope. However biological images are usually low contrast and these machines using interlaced lines with low band width did not reduce noise to acceptable limits for image analysis. Neurobiology groups were the first to use digitised tablets interfaced to dedicated microcompu ters, to measure nuclear/perikaryon ratios. Automated systems for differential cell counts also met with some success (Fisher 1971), but the impetus for the development of image analysis has been the desire to try to automate cytodiagnosis of cancer, particularly in exfoliative cytology. Trials of automated screening of cervical cancer smears were started and encouraging results were reported (Ploem et al. 1979). The measurement of nuclear size or area is one of the most important parameters which allows assignment of a given cell to one category or another

(Weid et al. 1982). Software based systems introduced a new flexibility needed to analyse biological images. In 1975 one of a new generation of machines was produced, the Magiscan.

Although this produced a digitised image, it was unable to store a complete image in its memory, and so image

93 enhancement and noise reduction could not be performed. However it was successfully used for several years, to apply morphometry to diagnostic pathology. In particular, this method was used to analyse changes in gut biopsies

(Slavin et al. 1980) and muscle pathology (Hanid et al.

1981). The current generation of image analysers is far more powerful and some can store up to 16 images. The Kontron IBAS II, used in this study, is a typical example of one of these machines. Now digital filtering techniques are able to enhance images from a complete digitised single scan TV image. Using this system a monochrome image is digitised to give a resolution of 512*512 pixels (picture elements), each pixel having a possible grey value 0 - 255. Although doubts were raised about complete automation (Exner 1981), this type of machine has proved fast and reliable and the contrast in this image, with 256 grey levels is more than sufficient for all biological applica tions (Bradbury 1983). The development of computer tech nology in the last decade has meant that complex analyses of biological images can now be performed with speed and accuracy, thus improving reproducibility and reducing observer error. The use of this technology applied to the prostate is demonstrated in the following chapters.

94 4.Morphometry in the prostate

Morphometric studies on the prostate up to the present time have used stereological methods applied mainly by two groups of investigators led by Bartsch in Austria and

Coffey in Baltimore, U.S.A.. Initial studies on the rat ventral prostate (Bartsch et al. 1975, DeKlerk et al.

1976, Rohr et al. 1976), showed a mainly glandular organ

(76%) of which 53% was accounted for by glandular luminar, and the rest (24%) was extracellular space. Soon a morpho metric model of the human prostate was developed (Bartsch et al. 1976,1977), which consisted of two major divi sions, the stromal part including connective tissue, blood vessels, nerves and smooth muscle fibres and the glandular part including the lumina of the acini and the glandular epithelial cells. From this model absolute stereological parameters were determined including the volume density of these components, length density and diameter of the glandular tubules. Distinct differences were shown between the morphometric parameters of the rat and dog compared to man. The human prostate contains more interstitial fibromuscular tissue with many smooth muscle cells contrasting with the more glandular prostates of the rat or dog

(Bartsch et al. 1976).

Furthermore quantitative variation in the glandular elements within the normal prostate was demonstrated

(Bartsch et al. 1979) . The glandular elements of the inner

95 part occupy 45% of its volume and this inceases to 55% in

the outer region of the gland, whereas the volumetric

amount of interstitial fibromuscular tissue is higher in

the inner part (55%) than the outer region of the same

gland.

The stereological data in BPH tissue in comparison with

normal tissue shows that there is a statistically signifi

cant increase in the volumetric amount of stromal tissue

and the glandular lumina. However with this method, no

significant difference was seen in the absolute volumes of

glandular tissue in BPH compared with normal prostate, or

in the surface density or height of the epithelial cells

(Bartsch et al. 1979) .

Attempts have been made to correlate stereological and

biochemical results to see if altered steroid metabolism

is reflected in the morphological architecture. However

endogenous steroid levels (oestrone, oestriol, oestradiol

and progesterone) did not significantly correlate with any

of the morphometric parameters measured (Bartsch et al.

1987) . Indeed there was no significant difference, in this

small series (n=10), in endogenous hormone levels between

normal and hyperplastic tissue. A statistically weak

correlation (p<0.5) was obtained between tissue DHT levels

and the volume density of the glandular tissue in BPH.

Receptor studies showed a weak correlation (p<0.5) between

cytosolic progesterone receptors and some of the morpho

metric data (volume density of the glandular lumen and

96 diameter of the glandular lumen and tubules). Although present in normal tissue, no cytosolic oestrogen receptor could be detected in BPH patients. This is unusual as the presence of the progesterone receptor is thought to signal a functioning oestrogen receptor, and the oestrogen recep tor has certainly been detected in other series (Tilley et al. 1980, Belis et al. 1983). The stereological analysis of the prostate has been used to study the differential effects of various drugs on BPH tissue. Coffey (1988) showed that there is a range in the glandular/stromal ratio in human BPH, from 0.07 in the more fibromuscular hyperplasias to 2.57 in the fibroadenomatous. The higher the ratio the more sensitive the prostate tissue is to androgen deprivation. This can be seen in the changes produced by cyproterone acetate (an anti androgen), which reduces the amount of glandular tissue, whereas tamoxifen

(anti oestrogen) appears to have no effect on the prostate

(Kreig et al. 1983).

However all these studies have so far failed to demon strate that overgrowth of any element is reflected in the endogenous hormone levels or any other biochemical parame ter of the whole prostate. In order to clarify the aetiol ogy of the disease other factors must be studied. In particular factors of the extracellular matrix, which may control the growth and function of the glandular epitheli um and may respond to the changing hormone environment in the ageing male, need to be studied.

97 CHAPTER TWO

BACKGROUND TO THE STUDY / EXPERIMENTAL STRATEGY

98 1.Background

The foregoing discussion clearly demonstrates the multifactorial aetiology of BPH, involving steroid hormones, their receptors, and locally derived tissue growth factors and their receptors. BPH is an entity independent from malignant disease of the prostate, arising in a different region of the gland (McNeal 1978,1984). Why these two areas of the same gland should respond differently despite being exposed to the same changes in the hormone environ ment with age, is still a mystery. Over the past three decades, extensive research has increased our understanding of the endocrine factors which regulate prostatic growth (Griffiths et al. 1979, 1987). However no primary endocrine abnormality in terms of circulating or tissue hormones has so far been detected which could account for either BPH or malignant change. There is therefore more interest in the biochemistry of target cells involved in uncontrolled growth of the var ious elements of prostatic tissue in BPH. The studies of Cunha et al. (1983) have emphasised the role of stromal- epithelial interaction, and both these elements are poten tial sites for steroid hormone action. The discovery of diffusible trophic growth factors (Spom and Roberts 1985), which may regulate growth processes, offers excit ing possibilities for this research. The need to quantify tissue elements and relate these to the biochemical find ings has never been more important. In the past this has

99 relied on labour intensive stereological methods, confined to dedicated laboratories. Now that space age technology has linked the computer to the microscope, a more user friendly and rapid system of investigation is a reality.

Likewise, in the past, studies on the biochemical and morphometric differences between the zones of the prostate have relied on tissue obtained by post-mortem, radical prostatectomy or organ donors (McNeal 1968, Bartsch et al.

1976, Habib et al. 1983). On occasions this has led to some misleading information (Geller et al. 1976, Hammond

1978, Kreig et al. 1979). Ultrasound guided prostatic biopsy offers the opportunity to sample tissue from dif ferent regions of the gland i n v i v o , and new assay techni ques are capable of producing reliable biochemical data on that tissue.

2.Experimental Strateav

The initial aims of this investigation were to establish reliable assays for the selected biochemical parameters.

It has been shown that epidermal growth factor (EGF) may be implicated in the pathogenesis of BPH and there have been reports of regulation of growth factor expression by steroid hormones (Muccu and Stancel 1985, Traish and Wotiz

1987, Schuurmans et al. 1988, St. Arnaud et al. 1988,

Lippman and Dickson 1989). The first aim was to apply a more simple assay technique for EGFR, originally used for

100 breast tumour specimens (Nicholson 1988 et al.), to BPH

tissue. Established EGFR assays require larger sample

volumes and could not be used for small biopsy samples.

Many groups have failed to find a reliable radioligand

assay for androgen receptors, and a new hybrid ligand

assay was recently developed (Stebbings et al. 1987) .

Unfortunately when applied to prostatic specimens, especi

ally those collected by TURF, this assay was also found to

be unreliable and unsuitable for small volume samples. A

monoclonal antibody to the receptor protein has recently

been developed, but at present this technique is expensive

and largely untried.

The role of oestrogens and progesterone in the pathogen

esis of BPH (Walsh and Wilson 1976), and their relation

ship to EGF was selected for closer investigation. Their

receptors are more robust and easier to assay than andro

gen receptors and assays for small sample volumes are

established.

To study biochemical differences within areas of the

prostate, the aim was to obtain samples by ultrasound

guided biopsy and then apply the assays for EGF and ster

oid receptors to these small samples.

In order to compare morphometric parameters between

samples the aim was to develop and apply new techniques of

image analysis. Eventually it was hoped to derive a mor phometric index of BPH which could then be compared to the biochemical data for growth factor binding, steroid recep tors and serum hormone levels.

101 CHAPTER THREE

MATERIALS AND METHODS

102 Materials and Methods

I HUMAN TISSUE COLLECTION 105

II CHEMICALS 109

(I) Radiochemicals 109 (II) Radlolodlnatlon of EGF 109 (III) Radio-Inert steroids 112 (Iv) Other reagents 112 (v) Scintillant 112

III APPARATUS 113

IV PREPARATION OF TISSUE FOR RECEPTOR ASSAY 113

(I) Human tissue 113 (II) Tissue storage 114 (III) Homogenisation of tissue 114

V BINDING STUDIES 115

1. Epidermal Growth Factor Receptor Assay 115 (I) Incubation 116

(II) Removal of free ligand 116 (III) Single dose assay 117

103 2. Steroid Receptor Assays 117 (i) ER assay 118 (ii) PgR assay 119

3. Estimation of protein 119

(i) Stock solutions 119 ( ii) Reagents 120 (iii) Assay 120

VI SERUM HORMONE CONCENTRATIONS 121

1•Testosterone 121 2.DHT 122 3.SHB6 122 4.0estradiol 123

VII IMMUNOHISTOCHEMISTRY 124

VIII IMAGE PROCESSING AND ANALYSIS SYSTEM 126

1. Hardware 126 (i) Image processor and computer 126

(ii) Input 127

2. Steps in image analysis 129

104 I HUMAN TISSUE COLLECTION

Prostatic tissue samples were obtained from patients at

St. Bartholomew's Hospital, London and the Homerton Hospi tal, Hackney. Immediately prior to operation a group of patients underwent transrectal ultrasound examination and tissue biopsy of the prostate. This tissue was used in some of the assays, but tissue removed by transurethral resection and retropubic prostatectomy was also used.

Plasma was collected by venous puncture from all patients prior to operation and from a group of healthy young male volunteers in the Biochemistry Laboratory, St. Bartholo mew's Medical College.

Biopsv Technique

All biopsies from the prostate were taken using transrectal ultrasound control. The patient was placed in the left lateral position and the ultrasound probe (Kretz Combison

310A) was inserted into the rectum. The probe has the capability to image the prostate in two planes (trans verse and longitudinal). When the area to be biopsied has been selected an 18 gauge biopsy needle is placed down the biopsy port attached to the probe (Fig. 3(i)). The ultra sound machine shows a representation of the expected path of the biopsy needle on the screen image and when the needle is fired this can be seen on the screen due to its hyperechogenicity (Fig. 3(ii)). The biopsy needle is fired using a Bard "biopty" gun and a small core of tissue

105 is removed from the needle for subsequent analysis. An average of three biopsies is taken depending on the amount of tissue taken at each pass. All patients were treated with prophylactic broad spectrum antibiotics prior to this procedure.

Fig 3 (il

The Kretz Combison 310A transrectal ultrasound probe, showing biopsy port and biopsy needle.

106 Fig.3 (ia)

ÿ- mmïmmmmsu

Fig. 3 (la) Diagram showing the transrectal probe In position In the rectum In order to obtain longitudinal and

transverse scans of the prostate and biopsies.

107 Fig 3 (ii)

Transrectal ultrasound of the prostate in longitudinal plane showing the path of the biopsy needle as it fires into the outer zone of the prostate.

108 ^ CHEMICALS

(1) Radiochemicals

The following radiochemicals were obtained from Amersham International PLC, Amersham, Bucks:

[2,4,6,7 -^H]-oestradiol (spec, activity 85-110 Ci/mmol)

[1,2,6,7 -^H]-progesterone (spec, activity 80-110 Ci/mmol)

Radiochemicals were stored at -20°C in redistilled etha nol. Radiochemical purity was initially greater than 97% as assessed by thin layer chromatography on silica gel, paper chromatography in petroleum ether and by isotope dilution analysis.

(ii) Radioiodination of EGF

Radio-inert mouse EGF was obtained from the Sigma Chemical Co. Ltd. (Poole, Dorset). This was iodinated in the De partment of Biochemistry, St. Bartholomew's Medical Col lege by the author with assistance from Mr. S.Barker.

(a) Preparation of the column

30g of G50 sephadex (Pharmacia, Uppsala, Sweden) was added to 600ml of 40nM phosphate buffer (pH 7.2) containing

O.lSnM sodium chloride, 0.1% (w/v) thiomersal and 0.5%

bovine serum albumin (BSA) (w/v). This was allowed to

109 swell with gentle agitation overnight at room temperature.

The slurry was the poured into a 70cm x 26mm (i.d.) glass chromatography column (Pharmacia, Sweden) via a column reservoir. The packing was then equilibrated with 3 column volumes of phosphate buffer at a flow rate of lml/2.5mins.

This equilibration procedure was also carried out after each separation of labelled EGF had been completed so that the column could be repeatedly used.

b) The labelling reaction

A solution of iodogen reagent (Pierce, Oud-Beijerland. The

Netherlands) was made up to give a concentration of 1 mg/ml in dichloromethane. 30ul of this solution was used to coat the bottom 0.5-lcm of a 1.5ml microtube (Scotlab,

Bellshill, Scotland). This was allowed to evaporate to dryness or dried down under nitrogen. The coated tube was then placed on ice and lOug of receptor grade mouse EGF, dissolved in lOOul of distilled water was added followed by ImCi of Na^^^I. The reaction was mixed and left on ice for 12-15 minutes and stopped by removal of the reaction mixture from the tube, the iodogen remaining on the tube surface (Fraker and Speck 1978). The reaction mixture was then loaded onto a G50 sephadex column to separate la belled EGF from free Na^^^I.

110 (c) Separation of the reaction products

The reaction mixture was applied directly onto the column bed and allowed to be absorbed by the top layer of packing material. The column was then filled with phosphate buffer and sealed. The reaction mixture components were eluted using lOmM trizma base buffer containing 50mM NaCl, and 0.1% BSA (pH 7.4), at a flow rate of lml/2.5mins. 5 minute (2ml) fractions were collected over a 16 hour period by Frac-100 fraction collector (Pharmacia). A 5ul aliquot from each fraction was counted by NE 1600 gamma counter (Thorn Nuclear Enterprises). On a more routine basis the fractions were collected corresponding to the positions of the peaks of the elution profile. The specific activity of the ^^^I-EGF was calculated from the radioactivity in the peak fractions based on a recov ery of 90 per cent. Using this system the range of speci fic activities obtained was 370 to 510 Ci/mmol, with a mean + sd of 420 + 44 Ci/mmol.

Ill ( ill) Radio-Inert steroids

The Sigma Chemical Company Ltd. Poole, Dorset supplied the following steroids

Diethylstilboestrol Norethisterone Cortisol

(iv) Other reagents

All other general purpose reagents were of analytical grade or higher and were supplied by British Drug Houses (BDH) Ltd. (Poole, Dorset) or Sigma Chemical Company, (Poole, Dorset) except dextran T70 and sephadex G50 which was supplied by Pharmacia Fine Chemicals A.B. Uppsala, Sweden.

(v) Scintillant

Toluene containing 4g/l 2,5 Diphenyloxazole (PPG) and 0.05g/l of 1,4 -Di-2-(5 phenyl oxazolyl) benzene (POPOP) was used as scintillant.

112 III APPARATUS

Homogenisation was carried out using a Polytron(Kinematica

GmbH, Switzerland). Centrifugation procedures were per formed in a Microfuge (Heraeus) and a L2-65B Ultracentri fuge (Beckman RIIC Ltd, High Wycombe). Gamma radiation was counted using a standard gamma counter (NE 1600, Thorn Nuclear Enterprises). 1.5 ml plastic Eppendorf microcen trifuge tubes were supplied by Anderman Co. Ltd, Kingston- Upon-Thames, Surrey.

IV PREPARATION OF TISSUE FOR RECEPTOR ASSAY

(i)Prostate Specimens

Tissue taken by trucut biopsy was immediately snap frozen in liquid nitrogen. With tissue taken by transurethral resection the largest chips with the least diathermy damage were selected and dried on blotting paper. Each chip was then sectioned through its transverse axis and one half of the chip was snap frozen in liquid nitrogen in the operating theatre. The other half of the chip was fixed in formalin and sent for routine histopathology. Tissue taken by retropubic prostatectomy was also sec tioned prior to snap freezing, one half of each approxima tely cm^ specimen being sent for histopathology.

These were processed to paraffin in the Histopathology

113 Department at St.Bartholomew's Hospital in the usual way and Sum sections were stained with haematoxylin and eosin for subsequent microscopic examination and morphometric analysis. Some samples for immunocytochemistry were placed in frozen section block moulds with OCT mountant and the mould floated on liquid nitrogen.

(ii)Tissue storage

All specimens were stored in liquid nitrogen prior to receptor analysis, which was usually carried out within two weeks of sample receipt.

(iii)Homogenisation of Tissue

The frozen tissue was finely minced with a scalpel blade over a liquid nitrogen bath and then weighed. Buffer used was 50nM phosphate buffer (pH 7.4) containing 30% (v/v) glycerol, 1.5 nM EDTA, lOnM monothioglycerol, lOnM sodium molybdate and protease inhibitors (aprotinin lug/ml and soyabean trypsin inhibitor lug/ml). Buffer was added to the tissue in the ratio, tissue weight : buffer volume = 1:10-20). The mixture was then homogenised using a Poly- tron homogeniser with 3 x 15 second bursts with one minute cooling in between bursts. The homogenate was then allowed to stand for 30 minutes at 4°C and then whirli-mixed.

Where there was sufficient tissue, the homogenate was

114 initially centrifuged at 800 x g for 10 mins at 4°C. The supernatant was then carefully removed leaving behind the lipid layer and nuclear pellet, and was then re-centrifuged at 100,000 x g for 60 mins at 4°C. The resulting

supernatant was removed and retained for soluble oestrogen receptor assays. The pellet (membrane fraction) was resu spended in tris base buffer (pH 7.4) containing 50nM sodium chloride, 0.1% w/v bovine serum albumin and pro tease inhibitors and retained for EGFR assay. With speci mens where there was only a small amount of tissue (from trucut biopsy) the homogenate was centrifuged at 100,000 x g for 60 mins at 4°C. The resulting supernatant was re tained for oestrogen receptor assay and the pellet (parti culate fraction) was resuspended as above and retained for EGFR assay.

V BINDING STUDIES

1. Measurement of EGF binding sites

The concentration and dissociation constants of EGF bind ing sites were determined by a radioligand binding assay (Sainsbury et al. 1985) and calculated by Scatchard analy sis (Scatchard 1949). EGFR binding in small biopsy speci mens was determined by a single dose assay (SDA) (Nichol son et al. 1988).

115 (1) Incubation

All assays were performed in 1.5 ml plastic eppendorf tubes. 100 ul of the membrane fraction were incubated in duplicate with 100 ul of 1.2 nM eGF. The contents were made up to a final volume of 400 ul using assay buffer (Tris NaCl + 0.1% BSA). Separately and concurrently in a second set of tubes, 100 ul of the membrane fraction were incubated in duplicate not only with radiolabelled EGF but also with 100 ul of a variable amount of unla belled EGF acting as a competitor (final concentration 0- 100 nM in 10 increments). The contents were made up to 400 ul with assay buffer. The tubes were whirli-mixed and incubated for 2 hours at 26°C.

(ii) Removal of free ligand

After incubation the tubes were transfered to an ice bath and subsequently kept at 4 °C. 800 ul of ice cold assay buffer was then added, and the suspension centrifuged at 10000 xg for 5 minutes. The supernatant was discarded by aspiration and the pellet was washed with a further 800 ul of ice cold assay buffer and recentrifuged at 10000 xg for 5 minutes. The supernatant was again discarded and the radioactivity in the pellet was counted in a gamma coun ter.

116 (Ill) single dose assay

A single point assay using one dose of I EGF was used to measure EGFR binding In small tissue samples such as those obtained by biopsy. 100 ul aliquots of the particu late fraction, representing a crude membrane fraction were Incubated In duplicate with a single concentration of I EGF (4nM) and the final volume of the tube made up to 400 ul with assay buffer. Non specific binding was deter mined In a second set of tubes by Incubating 100 ul aliquots of the particulate fraction with I EGF (4nM) plus 100 ul of 100 fold excess unlabelled EGF (400nM). The tubes were Incubated for 2 hours at 26 °C and following this the unbound ligand was removed and the pellet counted In the same manner as above. The specific binding to the receptor was determined by subtracting the bound steroid (non specific binding) In the competed tubes from the total bound steroid In the non competed tubes. The number of binding sites was related to the protein content of the sample and expressed as fmol/mg protein. Rat liver was used as a positive control for all EGFR assays.

2. Steroid receptor assays

A single saturating dose assay (King et al. 1979, Pudde- foot et al. 1987) was used to measure soluble cytosolic oestrogen and progesterone receptors.

117 (1) ER assay

100 ul aliquots of cytosol were incubated in duplicate with 20 ul of 50 nM oestradiol and 30 ul of buffer (non competed tubes). lOOul aliquots of cytosol were added to a second set of tubes containing 20 ul of 50 nM oestra diol, 10 ul of buffer and 20 ul of 10“^M diethylstilboe strol as a competitor (competed tubes). The tubes were incubated for 18 hours at 4°C. After incubation free steroid was extracted by the addi tion of 200 ul of a suspension of 0.25% w/v dextran T70 coated charcoal in 10 mM tris, 1.5 mM EDTA (pH 7.4) buf fer. The tubes were vortexed and incubated for 10 minutes at 4°C. The charcoal was pelleted by centrifugation at 10000 xg for 5 minutes. A 200 ul aliquot of the superna tant was decanted into a scintillation vial, 4ml of scintillation fluid added, the vial mixed and the radioac tivity was measured. Specific binding was calculated by subtracting the non specific binding (competed tubes) from the total binding (non competed) tubes. The protein con tent of the cytosol was measured and the results expressed as fmol/mg protein. Rat uterine cytosol acted as a posi tive control tissue for ER assays.

118 (il) PgR assay

This assay procedure was identical to that for the single saturating dose analysis of the oestrogen receptor except in the following respects. 20 ul of lOOnM progesterone was used as the label, and 20 ul of 10“ nor ethisterone was used as the unlabelled competitor in the competed tubes. 10 ul of 10~^M was included in all tubes both competed, and non competed to block the non specific binding of progesterone to contaminants such as corti sol binding globulin. The dextran coated charcoal extra ction was performed for 5 minutes only. Rat uterine tissue was used for positive control tissue for PgR assays.

3.Estimation of protein

The protein concentration of all samples was measured by the method described by Lowry et al. (1951). The exact details of the method used in this study are outlined below.

(i ) Stock Solutions

A. 0.5% w/v Copper Sulphate 5H2 O B. 1% w/v Sodium Potassium Tartrate C. 2.5% w/v Sodium Carbonate in 0.1% w/v Sodium Hydroxide

119 (il) Reagents D. Equal volumes of A and B E. 25 parts of C and 1 part of D

F. Folin Ciocalteau's reagent diluted with an equal part of distilled water immediately prior to use.

(iii) Assay

A standard reference solution of bovine serum albumin (BSA) was prepared at a concentration of 1 mg/ml in dis tilled water, aliquoted into microcentrifuge tubes and kept at -20°C. On the day of assay, an aliquot was thawed out and diluted with distilled water to give a range of standard protein concentrations such that the tubes con tained 0, 25, 50, 75, 100, and 150 ug BSA in a volume of 500 ul. Twenty microlitres of the sample cytosol were diluted with 480 ul of distilled water and 20 microlitres of buffer were diluted in the same fashion to provide a buffer blank.

Two and a half millilitres of the mixture E were then added to each of the protein solutions (standards and cytosols), and left at ambient temperature for 10 minutes.

Then, 0.25 ml of reagent F were added to each of the tubes, which were mixed and allowed to stand for 30 minutes at room temperature. The absorbance of the samples and standards was then read at a wave length of 700 nm in a Pye-Unicam spectrophotometer. The absorbance of the

120 samples was corrected for the absorbance of the buffer blank. A standard curve was constructed for the known protein standards, and the protein concentrations of the cytosol were derived from this standard curve. Routinely this was performed using a linear regression analysis programme on a Casio Fx-3400 P scientific calculator.

VI SERUM HORMONE CONCENTRATIONS

Blood was obtained from post pubertal healthy males under the age of 40 and from patients with BPH, prior to opera tion or biopsy of the prostate, between 0800 hrs-1100 hrs.

The serum was separated and stored at -20°C until the time of analysis. The following serum concentrations were measured:- testosterone, DHT, oestradiol, SHBG. All these assays were performed by the staff of the Department of

Reproductive Physiology, St. Bartholomew's Hospital.

1. Testosterone

Serum testosterone was measured by radioimmunoassay fol lowing simple ether extraction. Standard or extracted sample in assay buffer was mixed with a tracer of testosterone-3 carboxymethyloxime (CMC)-iodohistamine and an antiserum to testosterone-3CM0-BSA raised in rabbits.

After incubation for one hour, separation was achieved by the addition of dextran coated charcoal (10% w/v). Results

121 were expressed as nmol/1. Cross reaction of the system was less than 0.25 % for a wide variety of steroids with the exception of DHT (28.5%) and androstenedione (2.1%). Assay sensitivity was 0.4 nmol/1, within assay precision 5-8%, and between assay precision 6-8%.

2. 5-aloha dihvdrotestosterone

Serum 5-alpha DHT was measured by radioimmunoassay follow ing ether extraction and oxidation steps (Howell et al.

198 6) . This was necessary to remove the effect of endoge nous binding proteins, and the interference of testoster one in the DHT assay. Standard or extracted sample was mixed with a tracer of [^H]-5-alpha DHT and antiserum raised to 5-alpha DHT-3CM0-BSA in sheep. After incubation overnight, separation was achieved by the addition of dextran coated charcoal (1% w/v). Results were expressed as nmol/1 after correction for procedural losses. The cross reaction of the system was less than 0.08% for all steroids tested with the exception of androstanedione

(10.6%), and androstenedione (3.4%). Assay sensitivity was

0.25 nmol/1, within assay precision 10% and between assay precision 15%.

3. Sex Hormone Binding Globulin

The principle of the measurement of serum SHBG is based upon saturation of its binding sites with 5-alpha DHT and

122 precipitation of the bound complex with saturated ammonium sulphate (Fattah and Chard, 1981). A fixed amount of trltlated labelled and unlabelled DHT was Incubated with standard or sample, and the fraction bound to SHBG was precipitated with ammonium sulphate. "Standards" were determined by dilution of pregnancy serum In preheated male serum. Results were expressed as nmol/1. There was no cross reaction within the system and assay sensitivity was 10 nmol/1, and within and between assay precision was 6- 8%.

4. 17-beta oestradiol

Serum oestradiol was measured by direct radioimmunoassay carried out at pH 4 to obviate serum binding effects (Perry 1986). Standard or sample was mixed with the tracer of oestradlol-6-CM0-[^^^I]-Iodohistamine and an antiserum to oestradiol-6-CMO-BSA raised In a rabbit. After a late addition Incubation procedure, separation was achieved by the addition of dextran coated charcoal (0.5% w/v). Re sults were expressed as pmol/1. The cross reaction of the system was less than 0.001% for a wide range of steroids with the exception of 6-keto-oestradlol-CMO (8.95%), 6- keto-oestradlol (7.73%), oestrone (1.7%) and oestrlol (1%). Assay sensitivity was 50 pmol/1 and within and between assay precision 8-10%.

123 VII IMMUNOHISTOCHEMISTRY

EGF Receptors

EGFR were Identified by an indirect immunoperoxidase tech nique, using a murine monoclonal antibody.

(i) Preparation of sections

Tissue was stored in liquid nitrogen and small specimens cut over a liquid nitrogen bath using a scalpel blade. The specimens were freeze mounted onto aluminium blocks prior to section cutting. A small drop (approx 500 ul) of dis tilled water was pipetted onto the bevelled surface of the block, which was then placed up to its waist in a liquid nitrogen bath. The specimen was then held with fine for ceps until the water droplet froze around the specimen, thus fixing it to the block. The blocks were then trans fered to a cryostat and 6um sections cut. These were picked up on slides coated with poly-L-lysine and marked with a Dako pen.

(ii) Fixing and Incubation

After fixing the sections in 3.7% formaldehyde for 15 minutes, they were washed gently in Tris buffered saline (TBS) (.05 M Tris base, .15 M NaCl pH 7.6), for two per iods of 5 minutes. The sections were then fixed in metha nol (-20°C) for 5 minutes and acetone (-20°C) for 2 min utes, and again washed in TBS. The sections were then

124 covered with normal rabbit serum (diluted 1:5 with TBS), to act as a blocking agent, for 10 minutes and then drained. Incubation was with EGFR 1 anti-human EGFR (Amersham pic.), diluted 1:2 with normal rabbit serum), or control antibody (murine IgG, diluted 1:200 with TBS), to evaluate non specific binding. Incubation took place in a humidified chamber at 4°C for 16 hours.

(iii) Staining After further washing in TBS, the sections were incubated at room temperature for 30 minutes with rabbit anti-mouse immunoglobulin (Dakopatts) added dropwise, to act as a bridging antibody. Following this the sections were washed in TBS buffer and incubated for 30 minutes with peroxidase anti-peroxidase (mouse) complex to bind to the bridging antibody. Following a final wash in TBS buffer peroxidase activity was developed by means of a chromogen substrate solution containing substrate reagent, diaminobenzidine incorporat

ing imidazole and hydrogen peroxide (Stuart and Warford 1983). This solution was added to the sections dropwise and incubated in a fume cupboard for 20 minutes and the sec tions were then washed in distilled water. The sections were counterstained with 1% haematoxylin for 5 minutes and mounted in glycergel. Human placenta which contains large numbers of EGFR, was used as a positive control (Downward et al. 1984).

125 VIII IMAGE PROCESSING AND ANALYSIS SYSTEM

1.Hardware

(i ) Image processor and computer

The IBAS II Image processing unit manufactured by Kontron Electronics of Munich is a dual processor system compris ing a host computer and an array processor. (Fig 3.(iii)) The 64 Kbyte host computer is interfaced with an Iomega fast image dump of 10 megabyte capacity, which utilises a floppy disc. It takes approximately a second to transfer an image of 512*512 pixels from the image dump to the image processing system. The host computer is linked to a microprocessor which controls an autostage and an autofo cus for the microscope, and other peripherals such as the digitising tablet and keyboard. There is a direct memory access interface to an array processor, which comprises 4 megabyte RAM, TV frame grabber and look up tables. The image analysis system is fed by the video camera (Fig.3 (iv)) with output to a colour/monochrome monitor. The data produced by the IBAS system can be statistically analysed via a serial interface with an IBM clone computer.

126 Fig. 3 (iii)

The IBAS II Image Array Processor and computer. The digi tising tablet is seen at the extreme right of the picture.

(ii) Input The microscope used was a Reichart Polyvar with flat field objectives (Fig.3 (iv)). The objective size used in this study was a x30 lens and a green filter was used to en hance contrast. The images were scanned by a measurement grade Bosch monochrome television camera fitted with a chalnicon tube, producing the input signal to the image

127 analyser. The light source was electronically controlled by a Siemens light control unit set at 90 volts. This unit

allowed the same amount of light to be used for each objective and a feedback to the T.V. camera prevented

overloading. The grey levels 0 to 255 (black to white) were checked using a plain slide and a density wedge.

Fig. 3 (iv)

The video camera linked to the microscope to produce the

digitised image.

128 2. Steps in Image analysis

The slides used for image analysis were haematoxylin and eosin (H&E) stained paraffin sections, prepared in the

Department of Histopathology using standard procedures.

There is always some inconsistency in the thickness of paraffin sections and sections cut at 4um may vary between

3-7um depending on the tissue and the operator. This variation in thickness on the amount of stain contained has a greater effect on the measured grey value than the variation of illumination. The variation of illumination over the whole field was found to be less than 10%.

The program consists of a set of program instructions and each instruction executes a menu function. Some of these instructions are standard to other image analysis programs which have been used in other histological studies using automated image analysis (Sowter et al. 1987). The others were written by the author with help from Mr.C.Sowter, and some of the functions are used several times during the program. Illustrations demonstrating this specially de signed program are given below (Figs. 3(v-xi).

1. The first set of instructions set up the program con trols: allowing input of patient details, set up micro scope scanning stage and execution of the program.

129 2. The T.V. camera is brought on line, clear overlay planes, select microscope field. The monochrome image is

digitised into 512 * 512 square pixels and stored in an

image memory in the array processor. Each pixel has a grey

value of between 0 (black) and 255 (white). The monochrome

image is displayed on the monitor and the first figure

(Fig.3 (v)) is taken from the microscope prior to digiti

sation.

Fig.3 (v)

130 3. At this stage it is necessary to interact with the image using a mouse on a digitised pad. This allows manual amendment of the image in order to comlete partial lumen or exclude areas of the field where tissue is not present.

The amended image is then saved. A median filter is ap plied to reduce noise and a digital filter produces high levels of contrast and so gives maximum separation between groups of grey values.

4. A binary image is then produced from the grey image,

and any holes in the binary image are filled in. The binary image is then eroded by three pixels using a circu

lar structuring element and stored (Fig.3 (vi)).

Fig.3 (vi)

131 5. The stored binary image is then inverted and dilated by three pixels using the same structuring element (Fig.3

(vii)) and the image is again stored.

Fig.3 (vii)

6. The contours of each acinus are then obtained by apply

ing Boolean algebra or image arithmetic to the two pre

vious images stored in the memory (Fig.3 (viii) and this

is combined with the original amended image which has been

held in the memory (Fig.3 (ix)).

132 Fig.3 (viii) B » i

Fig.3 (ix)

1 i

133 7. In order to calculate fractions from the image a Look Up table is loaded. This allows pixels with grey values within a certain predetermined range to be assigned a particular colour. Thus the image in Fig.3 (ix) is copied onto the original image (Fig.3 (v)) and all pixels with grey values 0-9 (outside the range of the original mono chrome image) are assigned red (Fig.3 (x)), leaving the original grey image unaffected. This can be seen to corre late well with the epithelial component of the image and this step allows this to be checked by the operator, a similar process is repeated using the binary image of the acinar lumen (Fig.3 (vi)), copied onto the amended image and grey values less than 10 assigned a different colour.

In Fig.3 (xi) the result of these two steps is shown; all pixels with grey values greater than 9 represent the stromal component of the field.

Fig.3 (x)

134 Fig.3 (xi)

Fields in each section are analysed with a random grid using the microscope stage. This ensures that there is no observer bias i.e. field to be studied cannot be selected.

The sequence shown in the illustrations shows the entire field occupied by tissue, however using a random grid this is not always the case. The blank area has to be removed by manual amendment using the mouse and digitised pad. The computer automatically counts pixels occupied by tissue and expresses this as a percentage. The computer then counts pixels in the total field occupied by individual

135 tissue elements first the glands (epithelium and acinar lumen) and then lumen alone, and expresses these results as percentages. These can be adjusted depending on the percentage of tissue in each field and simple subtrac tions: Tissue % - (Epithelium + Lumen %) = Stromal %, Epithelium + Lumen % - Lumen % = Epithelium %, calculates the individual fractions. Random fields are sampled until a stable running mean is obtained. The minimum number of fields sampled was 20 per

slide, (Fig.3 (xii)) and each field takes an average of 1 minute to analyse depending on the amount of manual ad justment required. Each tissue sample requires at least 30 minutes in order to obtain results with a low coefficient

of variation. This is still considerably faster than other

more tedious methods such as point counting.

Thus one of the aims of the study was achieved, namely to

develop a semi-automatic morphometric technique for com paring prostatic tissue samples using routine H&E paraffin

sections•

136 CHAPTER FOUR

VALIDATION OF THE METHODOLOGY

137 I INTRODUCTION 139

II EGFR ASSAY 139

1. Optimisation and specificity 140 (i) Duration of incubation 140 (ii) 125j gQp concentration 140

(iii) Effect of dilution and protein concentration 142

(iv) Binding specificity 143 (v) Effect of heat and proteaseenzymes 143

2. Validation of the single point assay 143 (i ) Reproducibility 143 (ii) EGFR SDA v SA 144

111 ASSAY OF STEROID RECEPTORS 145

1. Reproducibility of ER and PgR assays 145 2. Validation of SSDA v SA 145

138 I INTRODUCTION

The determination of EGFR and steroid receptor concentra tions in the prostate in this study, presents several problems. These stem from the small amounts of tissue which can be obtained by biopsy and the relatively low concentrations of receptors, particularly steroid recep tors, which may be found in this tissue. It has therefore been necessary to use preparation and assay techniques which provide reliable and reproducible results and which can be applied to small tissue samples. The validation of the methodology adopted includes data on the correlation between results using this method and the standard method used to measure receptor binding and specificity, when larger tissue specimens are available.

II EGFR ASSAY

The EGFR binding of a tissue fraction is usually measured by competition between a given concentration of labelled

EGF and serial increases in concentration of unlabelled

EGF (multipoint Scatchard analysis). However Scatchard

analysis (SA) requires a substantial quantity of tissue (500 - lOOOmg), which limits the evaluation of EGFR espe

cially in the prostate, where often only small samples may be obtained by biopsy. A single dose assay (SDA) requires much less tissue (100 - ISOmg) and is more practical for

analysis of large numbers of samples.

139 1.Optimisation and specificity

(i ) Duration of incubation

Incubation of the membrane fraction of BPH tissue at 26°C with InM labelled EGF (final concentration) showed that maximal EGF binding occured in 1.5 hours and remained stable for 3 hours.(Fig 4 (i)).

Fig.4 (i) Duration of incubation

Binding CPM 2500

2000

1500

1000

500

0 50 100 150 200 Mins

Total Binding I— Non-Specific Binding — Specific Binding

(ii) Optimal ^^^1 EGF concentration

Measurement of EGF binding (1.5 hours) with increasing concentrations of labelled ligand revealed that the use of either 1 or 2nM (final concentrations) of ^^5j ggp gave equivalent levels of total binding. However as the concen tration of labelled EGF increased above 2nM there was a

140 further increase in total binding which did not reach a constant level (Fig.4 (ii). This pattern of total binding results from the fact that EGFR occurs as two populations of binding site. A high affinity binding site which is saturable at 2nM and a lower affinity population which is not saturated even using 5nM I EGF.

As a result of these data an optimum of 2 hours incubation was used for all subsequent assays. A final concentration of labelled EGF of InM was used which allowed measurement of high affinity EGFR whilst minimising the contribution from low affinity EGF binding sites.

Fig. 4(11) 125 I EGF Concentration (3 samples)

Total Binding (dpm)

6000

4000

2000

1 2 3 4 1 125 I EGF concn. (nm)

• Sample 1 + Sample 2 * Sample 3

141 (iii) Dilution Dilution of a BPH membrane preparation revealed a linear relationship between the dilution factor and corresponding EGFR level to a dilution factor of less than 0.5 (protein concentration = 1.3 mg/ml) (Fig 4(iii)).

Maddy et al. (1987) also found a biphasic linear relation ship between EGF binding and serial dilutions correspond ing to protein concentrations between O.lmg/ml and Img/ml, with a second linearity occuring between 1.5mg/ml and 8 mg/ml. In this study, final protein concentrations of the crude membrane fractions were generally less than Img/ml, due to the small tissue samples available (mean = 0.54 mg/ml). However with larger samples the protein concentra tion was adjusted to a range of protein concentrations near Img/ml.

Fig.4 (ill) Influence of dilution

Specific binding CPM 1 6 r

0 0.2 0.4 0.6 0.8 1 1.2 Dilution Factor

(Original Protein Cone > 2.6mg/ml)

142 (iv) Specificity of the Binding The specificity of labelled EGF binding to BPH tissue membranes was demonstrated by incubating a membrane frac tion (lOOul) of BPH tissue in triplicate with ggp at

26°C for two hours in the absence and presence of excess unlabelled polypeptide as a competitor. Human insulin , prolactin and FSH concentrations of lOOOng/ml did not result in any significant inhibition of ggp binding.

(v) Effect of heat and protease enzymes If the particulate fraction was heated to cause protein dénaturation EGF binding was totally abolished. Similarly if protease enzymes such as trypsin or amylase were added to the preparation no EGF binding could be detected.

2. Validation of the single point assay

(i) Reproducibility

For the intra-assay reproducibility 3 BPH specimens were analysed in quadruplicate. Coefficients of variation ranged from 6.3-12.7% (Table 4 i.). For the inter-assay reproducibility 4 specimens were analysed in duplicate in 2 different series. The results of the two series were in good agreement, giving a correlation coefficient of 0.93.

(Table 4 ii.)

143 Table 4 Reproducibility

Intra-Assay Variation

* Mean Concentration fmol/mg protein x 4 samples

* Patient EGFR S.D. C.V.%

43.3 5.5 12.7

61.8 3.9 6.3

66.8 7.2 10.7

C.V.% > (Standard Deviation (SD)/ Mean) x 100%

Table 4 ii. Reproducibility

Inter-Assay Variation

Patient EGFR(1st Assay) E(;FR ( 2nd Assay)

1 41.3 35.9

2 49.8 55.5

3 35.5 27.1

4 71.2 64.3 r -0.93

(ii) EGFR using SDA v SA

EGF binding was analysed in 9 human BPH samples by Scat chard analysis (SA) and single dose assay (SDA). All the tissue samples were taken by transurethral resec tion and the diagnosis confirmed by histology. Comparison

144 between binding results using the two techniques showed a significant quantitative correlation (n=9, r=0.85, p<0.001). However the SSD assay underestimated EGFR bind ing by approximately 50%, compared to SA ( Table 5 ii..

Fig 5 (i)) The SDA does not have the specificity controls that are built into the standard methods for receptor analysis such as SA, which measures dissociation constant. However the quantitative correlation between EGFR levels measured by SA and SDA is satisfactory despite the underestimation by SDA found in these BPH specimens and previously in breast tumours (Nicholson et al. 1988).

Ill ASSAY OF STEROID RECEPTORS

(i ) Reproducibility

The assay of steroid receptors (ER and PgR) was monitored by sequential analysis of a reference preparation of human uterine cytosol. This cytosol was obtained by centrifuga tion (100,000g for 60 min) of an homogenate of human uterine tissue prepared in the glycerol assay buffer. Aliquots of this cytosol were frozen and stored in liquid nitrogen. Sequential analysis of ER in the uterine cytosol prepara tion gave a mean value of 52.2 fmol/mg protein from 34

145 assays (S.D. = 8.9) with an interassay coefficient of variation (CV%) = 17%. The mean value of PgR in this pre paration was 210.5 fmol/mg protein (S.D. = 26.4, CV =

12.5%).

(ii) Validation SSDA y ^ for Steroid Receptors

A validation of the ER single saturating dose assay (SSDA) was performed in our laboratory by comparing the values obtained by SSDA with those obtained by Scatchard Analysis (SA) in the cytosol of 45 breast tumours. Analysis of the data from all the tumours by linear regression gave a correlation coefficient = 0.969 which was highly signif icant (p <0.001, df=43) (Fig 4 (iv)). The results and figure are presented with kind permission of Mr.J.Pudde- foot (1990)• King et al. (1979) found an underestimation of receptor content by about 20% when comparing SSDA versus SA in breast tissue. This is expected to be the same in the prostate as some series report low levels of

ER.

146 Fig.4 (iv)

A comparison between the Scatchard and Single Dose methods of oestrogen receptor analysis.

300 1 SINGLE SAT. DOSE ANALYSIS (fmols mg protein)

200

100 r = 0.959 p<0.001

100 200 300 SCATCHARD ANALYSIS (fmols mg protein)

147 CHAPTER FIVE

BIOCHEMICAL RESULTS

148 BIOCHEMICAL RESULTS

I INTRODUCTION 150

II EGF BINDING STUDIES 151 1. EGFR in BPH and "Normal" prostate 151 2. Immunohistochemistry 161

3. Discussion 166

III STEROID RECEPTOR BINDING STUDIES 167 1. Introduction 167 2. Oestrogen receptor (ER) 168 3. Progesterone receptor (PgR) 170 4. Comparison of steroid receptor and EGFR binding 172 5. Discussion 174

IV REGIONAL DISTRIBUTION OF EGFR/STEROID RECEPTORS 175 1. Introduction 175

2. EGFR 176

3. ER 178 4. PgR 180 5. Discussion 181

V SERUM HORMONE LEVELS 184

1. BPH V Control 184 2. Correlation with EGFRand steroid receptors 185

3. Discussion 187

149 I INTRODUCTION

This chapter is concerned with the results from the bio chemical assays performed on prostatic tissue removed by transurethral resection of the prostate (TURP) in patients with BPH. The results of EGFR assays in a group of such patients are contrasted with those in a "control" group and these are compared with the immunohistochemical find ings. Steroid receptor assays were performed on the cyto solic fraction prepared from the same tissue and correla ted with the level of EGF binding. The regional distribu tion of EGFR and steroid receptors within prostate biopsy specimens from patients with BPH is then presented and discussed. The relationship of these biochemical para meters to serum steroid hormone levels and binding globu lins is reviewed. The next chapter is concerned with the morphometric results in BPH specimens, applying new computerised tech niques of image analysis. Quantitative estimates of tissue components are then compared with the biochemical findings for the same tissue. Not all samples from all patients were suitable for measurement of all biochemical or morphometric parameters. Not exactly the same group of patients or samples are presented throughout as indicated. This was due to lack of sufficient sample volume, handling errors or change in diagnosis with subsequent histological examination.

150 II EGF BINDING STUDIES IN BPH

1. EGFR in BPH and "Normal" prostate

A series of epidermal growth factor receptor (EGFR) assays was performed with the single dose assay, used by

Nicholson et al. (1988) for breast tumour tissue. The series consisted of 22 different samples of prostate removed by transurethral resection (TURP), for benign prostatic hyperplasia (BPH). BPH was confirmed histologi cally in 19 samples, 2 were found to have microfoci of well differentiated adenocarcinoma and 1 had granulomatous prostatitis. The results of these latter 3 samples were therefore disregarded.

The single dose assay using 1 nM ^^^I-EGF had a cut off value of 10 fmol/mg protein (Nicholson 1988). This cut off value was chosen by virtue of the fact that with a concen tration of 1 nM ^^^I-EGF in a volume of 400ul and a pro tein concentration of lOOug, specific binding of 0.25% represents a receptor concentration of 10 fmol/mg protein.

This is equivalent to approximately 200-3000 cpm (mean

950), where non specific binding is in the order of 600-

2,300 cpm (mean 1150), depending on the specific activity of the EGF. When an unpaired t-test between the 3 figures for total binding and 3 for non specific binding reaches the 95% confidence limit (p<0.05) and binding >0.25%

(equiv. to 10 fmol/mg) is reached then the specimen can be considered positive for EGFR (Nicholson et al. 1988).

151 In the previous chapter binding of EGF to the membrane fraction of EGF was found to be time and temperature dependent. Therefore all incubation in this series and in subsequent assays was undertaken at 26°C for 120 minutes.

Using this method and taking a cut off value of 10 fmol/mg protein, specific displaceable binding sites were found in all specimens (100%) of BPH analysed. The mean value for EGF binding in this series of BPH tissue samples 52.3 +/- 23.5 fmol/mg protein (range 15.9-86.4) appears low rela tive to other series (Maddy et al. 1987), with 2 specimens with values less than 20 fmol/mg protein (Table 5 i.). Other groups have demonstrated the presence of specific EGF binding in hyperplastic tissue using similar techni ques (Maddy et al. 1987, Eaton et al. 1988, Lubrano et al. 1989). The absolute values for EGF binding in BPH differ between these groups and from this series. This may be due to different techniques of tissue collection, storage, pre paration or analysis, which vary between laboratories. In the present preliminary studies (Chap.4), however, and in those of other groups, it has been shown that in BPH membrane fractions two binding sites exist (Hofman et al.

1984). Therefore, the main difference is probably due to measurement of the different populations of receptor. Eaton et al. (1988) quote a mean of 86.44 +/-18.12 fmol/mg protein, Maddy et al. (1987) 125 +/-7 fmol/mg protein and

Lubrano et al. (1989) 139.7 +/- 106.6 fmol/mg protein.

152 Eaton et al. (1989) used a single point assay with satura ting concentrations of ^^^I-EGF (15 nMol/1) with 100 fold excess of cold EGF. This assay therefore reflects the sum of low and high affinity receptor sites. Lubrano et al.

(1989) and Maddy et al. (1987) also used high concentra tions of labelled EGF (24 nMol/1 and 15 nMol/1) respect ively, but in the context of multipoint assays. Using latter methods figures for EGF binding will include a high proportion of low affinity binding sites. An assay based purely on saturation of all available binding sites would give an accurate estimate of this if only a single class of binding site existed.

Table 5 i^ T.U.R. P . Specimens EGFR Single Dose Assay

Ho^ Patients EGFR fmol/ma protein

1. CM 57.9 2. MD 27.2 3. AF 15.9 4. BK 42.9 5. FW 75.9 6. EL 58.5 7. RG 36.9 6. ET 17.0 9. JAs 52.9 10. HR 75.0 11. PB 86.4 12. JAt 29.7 13. JW 29.3 14. WW 47.1 15. FD 86.1 16. CG 81.4 17. EG 34.1 18. HT 60.5 19. MA 79.1

Mean 52.4 SO 23.5 SEN 5.4

153 Using the assay described previously only a proportion of the high affinity binding sites could be measured. This data was therefore compared with data obtained from the same tissue preparations using a multipoint radioligand binding assay. A multipoint analysis using 4nM demonstrated 2 specific binding sites with a typical curvilinear rather than a straight Scatchard plot, consis tent with two (high and low affinity) binding sites

(Hofmann et al. 1984) (Fig.S(ib)). Analysis of the Scatchard data was performed using an in- house program on an Amstrad microcomputer (Appendix II). However these results were checked using the LIGAND pro gram (Munson and Rodbard 1980) which gave similar results. In this small series there was a good correlation (r=0.85, p<0.008) between binding at InM (Single dose assay) and the capacity of the high affinity receptor site as calcu lated from the Scatchard analysis (Table 5 ii.,Fig 5(ii)).

154 Fig.5 (la)

Saturation curve specific ^^^I-labelled EGF binding to

human BPH. Membrane fractions incubated with radiolabelled EGF alone and in the presence of a range of concentrations of cold EGF.

EGFR saturation analysis Patient M.D.

B.Sp. (nM) 0.02

0.015

0.01

0.005

0 2 4 6 8 10 12 EGF dose (nM)

155 Fig.5 (ib) The biphasic plot with Scatchard analysis indicates two classes of binding site. Analysis of the linear components gives a value for specific binding (X intercept) and avidity of association (Kd=l/slope) between EGF and recep tors.

EGFR Scatchard plot Patient M.D.

0.012

0.01

0.008

/ 0.006

0.004

0.002

0 0.005 0.01 0.015 0.02 B.Sp (nM)

156 Table 5 (ii)

Comparison of results of EGFR Binding assayed by Scatchard

Analysis (SA) and Single Dose assay (SDA).

No. Patient Kd(nmol/l) EGFR(SA) EGFR(SDA)

1. WW 1.7 53.9 47.1 2. HG 0.9 80.5 34.1 3. RG 2.3 70.5 36.9 4. BK 2.1 90.5 42.9 5. MD 1.7 51.6 27.2 6. CM 1.3 112.1 57.9 7. ET 1.2 42.6 17.0 8. HT 2.3 142.1 60.5 9. JAs 2.5 109.2 52.9

Mean 1.9 83.7 41.8

SD 0.7 33.0 14.5

SEM 0.2 11.0 4.8

Fig.5 (ii) EGFR SA V SDA BPH TURP Specimens r-0 .8 5 SA fmol/mg protein 200 r p<0.008

100

0 10 20 30 40 50 60 70 SDA fmol/mg protein

157 These results show that the single dose assay at InM correlated well with high affinity binding, a relationship similar to that found by Nicholson et al. (1988) in breast tissue samples. This provides a valuable basis for mea surement of binding capacities in other prostate specimens and to compare these with other biochemical or morpho metric parameters which may affect EGFR binding.

EGF binding in the normal prostate

Specimens were obtained as previously described from 9 patients undergoing cystoprostatectomy for bladder tu mours. In these patients there was no clinical or patholo gical evidence of BPH. These samples were processed in exactly the same way as the BPH samples and EGF binding of the membrane fraction determined. Only 55.6% of the sam ples assayed were EGFR positive (>10 fmol/mg protein). A significantly lower level of EGFR binding was found in the normal specimens of prostate, 15.4 +/- 5.7 fmol/mg protein

(Table 5 iii.), p<0.0001 than in the BPH specimens (Fig.5

(ill)).

158 Table 5 iii. Normal prostate specimens EGFR Single dose assay

No. Patient EGFR (fmol/mg protein)

1 . WS 15.7 2 . MD 21.4 3. ES 0 4. JV 34.6 5. RW 9.8 6 . PD 14.2 7. NS 23.1 8 . VG 7.9 9. CG 8.4

Mean 15.4 SD 9.8 SEM 5.7

Fig.5 (iii) EGFR Binding (Single Dose Assay) Normal Prostate v BPH

EGFR fmol/mg protein p<0.0001 50

Normal BPH

159 It may have be expected that the EGFR binding levels in the TURP would be artefactually lower due to the tissue trauma during TURP, particularly due to the heating ef fects of the diathermy. This is contentious, although

Habib et al. (1986) did not find this to be the case for androgen receptors, if the prostatic tissue used in the assay was carefully selected. However, this helps to strengthen the hypothesis that there is increased EGF binding in BPH which may be due to up regulation of the EGF receptor in this disease.

160 2. Immunohistochemistry Results

Immunohistochemlstry was performed using an immunoperoxidase technique on cryo-sections obtained from the same TURP samples used for EGFR binding assays. The monoclonal antibody used was specific to the external domain of the EGF receptor. The pattern of immunoperoxidase activity in these sections is demonstrated in Fig.5 (v). The immuno peroxidase activity appears strongly at the cell mem branes. The monoclonal antibody is shown to be specific to the lining epithelial cells of the glandular elements of BPH, with more intense staining of the basal layers Fig.5 (vi). It has previously been demonstrated that there is no staining of the surrounding stroma either muscular or collagenous (Maddy et al. 1987). No staining was seen using mouse IgG as a control antibody (Fig.5 (iv)).

The intensity of staining and EGFR binding measured by radioligand binding assay in tissue from the same samples was compared. Three independent researchers graded the intensity of the staining using a scale of 1-4 where 4 represented intense staining.

There was good correlation between the intensity of staining and EGFR binding measured by radioligand binding assay (r = 0.89, p<0.01). Four specimens with high levels of EGFR binding >75 fmol/mg protein all showed intense immunohistochemical staining (Table 5 iv.).

161 Table 5 iv.

Immunohistochemistry v Single Dose Assay BPH

Patient Staining Intensity EGFR fmol/mg protein

1. CM 3.3 57.9 2. MD 2.0 27.2 3. AF 1.3 15.9 4. BK 2.0 42.9 5. FW 4.0 75.9 6. EL 3.0 58.5 7. RG 1.7 36.9 8. ET 1.0 17.0 9. JAs 2.3 52.9 10.HR 3.0 75.0 11. PB 4.0 86.4 12.JAt 2.3 54.7 13. JW 1.0 29.3 14. WW 2.0 47.1 15.FD 3.3 86.1 16. CG 4.0 98.2 17. HG 1.7 40.8 18.HT 2.0 60.5 19.MA 3.7 79.1

Immunohistochemistry was also performed using the same technique on the 9 samples of prostatic tissue taken from patients without evidence of BPH. The results were analy sed in the same manner Table 5 v.

Table 5 v.

Immunohistochemistry v Single Dose Assay Normal

Patient Staining IntensityEGFR fmol/mg protein

1. WS 1.0 15.7 2. MD 1.0 21.4 3. ES - 0 4. JV 2.3 34.6 5. RW 1.0 9.8 6. PD 1.3 14.2 7. NS 1.6 23.1 8. VG - 7.9 9. CD - 8.4

162 In both normal and BPH specimens there was uniform staining of the glandular epithelium which appears to be predomin antly on the surface of the cell membrane where the EGF receptors are located. There was no evidence of nuclear staining in any of the specimens. The intensity of stain ing in "normal" prostate was consistently lower than that of BPH, a relationship which is paralleled in the ligand binding assay results for the two groups (Table 5 v.).

Fig. 5 (iv)

Immunohistochemistry of BPH Tissue with control antibody

(murine IgG), showing no peroxidase activity (x230).

■V -

163 Fig.5 (v)

Immunohistochemistry of BPH tissue using murine EGF mono clonal antibody showing specific peroxidase activity of the glandular epithelium (x230).

■ m ."V

ir ' r ■

'■■ -i' , - • . - - ^ i . -- . -I,

164 Fig. 5 (vi)

Immunohistochemistry of BPH tissue showing intense peroxi dase activity of the glandular epithelium, particularly of the basal layer (x320).

165 3. Discussion

It is clear from this and other studies (Maddy et al.

1987, Eaton et al. 1988), that normal prostatic tissue and

BPH contains receptor proteins with a high affinity for EGF. This study also shows greater binding in the membrane fractions of prostatic tissue taken from patients with clinical and histological evidence of BPH, than in speci

mens of "normal" prostatic tissue. However there is diffi

culty in collecting good "normal" prostate tissue and dispute about what constitutes appropriate control tissue

in BPH studies. On the one hand it has been suggested that

the biochemical events which eventually lead to the deve

lopment of BPH begin at puberty (Griffiths et al. 1979).

Therefore it can be argued that control tissue should be obtained from pre-pubertal cadaver donors or post mortem. In this study the controls were age matched patients undergoing other procedures which require removal of the

prostate gland, such as cystoprostatectomy which has the

advantage of controlling for confounding factors associa

ted with ageing. However in this age group it is difficult

to find patients who do not have at least some histologi cal evidence of BPH. Subsequent studies address this problem further.

166 Ill STEROID RECEPTOR BINDING STUDIES

1. Introduction

In order to further investigate the role of oestrogens in

BPH, oestradiol receptors (ER) and progesterone receptors (PgR), were measured in the cytosolic extract of BPH tissue homogenate from the same 19 BPH specimens and 9 "normal” specimens whose membrane fraction was used for EGFR assay. A single dose assay (King et al. 1979, Pudde- foot et al. 1987), with overnight incubation with radioac tive ligand was used as described in detail previously (Chap.3). Threshold values were fixed at 5 fmol/mg protein for both ER and PgR cytosolic receptors. In previous studies of steroid receptors in BPH a threshold value of 3 fmol/mg protein was used (Lubrano et al. 1989). It is also known that the levels of ER in BPH cytosol are relatively low (Kirdani et al. 1984). At low levels the single dose assay in our laboratory may underestimate the cytosolic steroid receptor level by approximately 5 fmol/mg protein

(Chap.4). In order to increase specificity and eliminate false positive results but still maintain an acceptable level of sensitivity, a threshold value of 5 fmol/mg pro tein was thought to be a reasonable compromise. It was then possible to correlate the levels of steroid receptor binding with peptide growth factor (EGFR) in the same tissue.

167 2. Oestrogen Receptors

In this series of 19 BPH patients cytosolic ER were pre sent in 68% of the tissue samples assayed. The mean con centration was 20.0 +/- 4.3 fmol/mg protein (Table 5 vi.)

Table 5 vi. TURP specimens cytosolic ER (fmol/mg protein)

No. Patient ER

1. CM 10.1 2. MD 57.1 3. AF 41.6 4. BK 34.6 5. FW (3.1) 6. EL 16.6 7. RG 12.1 8. ET 52.8 9. JAs (0) 10. HR (0) 11. PB (4.6) 12. JAt 37.3 13. JW 33.6 14. WW 21.1 15. FD (3.7) 16. CG 10.2 17. HG 33.8 18. HT 15.2 19. MA (0)

Mean 17.7

S.D. 15.1

S.E.M. 3.5

% *ve 68%

168 In the 9 specimens taken from "normal" glands, cytosolic ER was found in all specimens (100%) analysed. However the levels of ER in these specimens were still low (mean = 23.5 +/- 3.6 fmol/mg protein), which is not significantly different (Mann-Whitney p<0.3) from the BPH specimens

(Table 5 vii.)

Table 5 vii. "Normal" prostate, cytosolic ER (fmol/mg protein)

No. Patient

1. WS 23.3 2. MD 15.6 3. ES 33.1 4. JV 38.6 5. RW 33.4 6. PD 8.2 7. NS 21.4 6. VG 9.2 9^ ÇS 28.8

Mean 23.5

S.D. 10.9

S.E.M. 3.6

% ♦ve 100%

Immunohistochemistry using the ERICA assay (Amersham UK) was also used on BPH tissue samples. However, in common with other reports, no specific staining could be demon

strated in any of the normal or BPH tissue samples.

Monoclonal antibodies raised against breast and endome trial ER do not react with prostate, suggesting an anti genic difference (Pertschuk et al. 1985).

169 3. Progesterone receptors (PgR)

Cytosolic progesterone receptors were present in 52.6% of the BPH tissue samples assayed. The mean concentration of these receptors was 10.3 +/- 2.6 fmol/mg protein (Table 5 viii. )

Table 5 viii.

Cytosolic PgR TURP specimens (fmol/mg protein)

No. Patient PgR

1. CM 20.6 2. MD 13.6 3. AF 11.5 4. BK 16.6 5. FW (0) 6. EL 11.0 7. RG (0) 8. ET 28.7 9. JAs (0) 10. HR (0) 11. PB (0) 12. JAt 15.5 13. JW (0) 14. WW 11.5 15. FD (0) 16. CG (0) 17. HG 23.7 18. HT 36.5 MA 101

Mean 10.3

S.D. 11.4

S.E.M. 2.6

% +ve 52.6%

170 There was also a poor yield of PgR positive samples in those specimens taken from "normal" prostates. Significant levels of specific cytosolic PgR binding could only be found in 44.4% (4/9) specimens, with a mean level of only 8.2 +/- 3.1 fmol/mg protein (Table 5 ix). There was no significant difference between PgR binding levels in BPH and "normal" prostate.

Table 5 ix.

Cytosolic PgR "Normal" Prostate (fmol/mg protein)

Wo. Patient PoR

1. WS 9.2 2. MD 20.5 3. ES (0) 4. JV 13.8 5. RW 24.3 6. PD (3.9) 7. NS (1.2) 8. VG (0) 9^ 101

Mean 8.2

S.D. 9.3

S.E.M. 3.1

% ♦ve 44.4%

171 4. Comparison of Steroid Receptor and EGFR binding proper ties of TURP specimens

The correlation of ER and PgR in terms of absolute values was poor (Spearman rho=0.68, p<0.05). However 77% of the samples that were ER positive were also PgR positive (Table 5 (x)), and all samples that were ER negative were also PgR negative indicating a high degree of association between the two. It seems that when ER is present in sufficient quantities in BPH to be detected by this assay then it is likely to be a functioning receptor. When comparing the steroid binding results with those for EGFR binding in the same tissue, there was a significant negative correlation between EGFR and cytosolic ER (rho = -0.72, p < 0.01). (FIG 5 (vii)). There was no correlation between EGFR and cytosolic PgR, although PgR is known to indicate a functioning oestrogen receptor.

172 Table 5 x.

TURP Results EGFR and Steroid Receptors (fmol/mg protein)

Patient EGFR PgR

1. CM 57.9 10.1 20.6 2. MO 27.2 13.4 13.9 3. AF 15.9 28.7 9.4 4. BK 42.9 34.6 16.6 5. FW 75.9 3.1 G 6. EL 58.5 16.6 11.0 7. RG 36.6 12.1 0 8. ET 17.0 52.8 28.7 9. JAs 29.7 0 0 10. HR 75.0 0 0 11. PB 86.4 4.6 0 12.JAt 54.7 37.3 15.5 13. JW 29.3 33.6 0 14. WW 47.1 21.1 11.5 15. FD 86.1 9.9 0 16.00 81.4 10.2 9.3 17. HG 40.8 33.7 23.7 18. HT 60.5 15.2 36.5 19.MA 79.1 0 0

Fig.5 (vil) EGFR V ER Binding BPH (TURP Specimens)

EGFR fmol/mg protein 100 Spearman P —0.72

p<0.01

4 0

10 20 30 40 50 60 ER fmol/mg protein 173 5. Discussion

The results of oestrogen receptor binding in BPH using a radioligand assay are similar to those of Emtage et al.

1989. They also reported higher levels in prostate cancer

(mean = 27.5 fmol/mg protein), but did not include any normal tissue in their study and could not demonstrate any relationship with androgen receptor binding. Lubrano et al. 1989 found a significant negative correlation between nuclear steroid receptors and binding capacity of EGFR. The negative correlation of EGFR and ER raises the poss ibility that these two receptor mechanisms may interact in some way in the pathogenesis of BPH. Oestrogens and andro gens, in vivo, may act at two different levels, enhancing the synthesis of EGF itself or its receptor, through paracrine (stromal-epithelial interaction) and endocrine mechanisms. The final result is a homologous down regula tion of EGFR. It has been demonstrated that adding EGF to cells possessing EGFR, the ligand receptor complex is internalised and degraded in lysosomes. Androgen and oestradiol stimuli may decrease EGF production and EGFR synthesis but allow basal EGFR to be exposed on the cellu lar surface. It is also possible that EGF through its own receptor may negatively regulate steroid receptor synth esis, activation and or translocation in the nuclei.

174 IV REGIONAL DISTRIBUTION OF EGFR / STEROID RECEPTORS WITHIN THE PROSTATE

1. Introduction

In order to investigate the regional distribution of EGFR and steroid receptors within the prostate, biopsies from

30 patients with BPH were studied. Samples of BPH tissue were taken by biopsy as previously described (Chap.3). These small samples were processed carefully but often yielded very low membrane protein concentrations. However EGF binding assays were possible in 27 cases. Three sam ples were disregarded due to the paucity of tissue ob tained. In 18 patients cytosolic steroid receptors (ER and PgR) were also measured in the cytosolic fraction of the same tissue homogenate used for EGFR assay. All assays were single dose assays as there was insufficient tissue for full Scatchard analysis. These assays were usually performed in duplicate, although in 8 samples only single assays could be performed due to shortage of tissue. The results are presented separately for each type of receptor assayed with the mean, standard deviation (SD) and stan dard error (SEM) for each set of measurements. The rela tionship between these receptors in each zone of the prostate is then examined.

175 2. EGFR Regional Distribution

The results show that there is a highly significant dif ference between EGFR binding in the inner zone of the prostate compared to the outer zone. In tissue taken from the inner zone specific binding was present in 89% (23/27) of the samples assayed. In the outer zone tissue only 54% (14/26) of the samples assayed demonstrated specific binding. The mean concentration of EGFR binding in the inner zone was 73.3 +/- 7.7 fmol/mg protein compared to 16.7 +/- 3.5 fmol/mg protein in the outer zone (Mann- Whitney W=889, p<0.0001. Table 5 xi.. Fig.5 (viii)). There was insufficient tissue to perform immunohistochemical confirmation of the differences in EGF binding between the two zones as all the samples were biopsy specimens. 5 specimens of normal prostate were also taken from pa tients undergoing cystoprostatectomy. Samples were taken from the portion of the gland closest to the urethra (inner zone) and some from near the capsule (outer zone). This was an attempt to obtain some control tissue as it was not possible to obtain transrectal ultrasound guided biopsy specimens from these patients. There were low levels of EGF binding in specimens from both inner and outer zones in these cases (mean 18.2 +/- 4.3 fmol/mg protein (inner zone) and 15.4 +/- 7.8 fmol/mg protein

(outer zone)), but no significant difference between results from the two areas.

176 Table 5 xi. Regional Distribution EGFR (fmol/mg protein)

Patient Inner Zone Outer

1. HT 64.9 0 2. HG 40.2 0 3. CG 127.8 53.8 4. FD 106.6 48.6 5. WW 42.9 - 6. JW 7.3 9.9 7. JAs 62.1 23.5 8. PB 106.7 22.8 9. SG 87.6 17.2 10. HS 67.2 6.8 11.HR 105.8 54.2 12.JAt 116.0 35.0 13.ET 67.5 16.4 14.RG 55.8 7.8 15.MG 65.8 0 16.EL 52.7 4.3 17.FW 114.6 0 18.BR 22.5 14.4 19.DH 126.2 29.4 20. JS 91.6 32.3 21.RW 119.3 3.8 22.PK 2.5 7.9 23.EL 116.8 19.2 24. AF 6.6 8.1 25.FS 118.2 31.7 26.HD 42.3 7.5 27.CM 58.5 27.2

Mean 73.3 16.7

40.2 18.0

S.E.N. 7.7 3.5

177 Fig. 5 (viii) Regionai Distribution EGFR Binding BPH (Biopsy Specimens)

EGFR fmol/mg protein p < 0 .0 0 0 1 100

Inner Zone Outer Zone

3. Oestrogen Receptor Distribution

Samples taken from the inner zone showed specific ER binding in 61% (11/18) compared to 50% (9/18) samples from the outer zone. The mean concentration of ER in the inner zone was 12.1 +/- 3.0 fmol/mg protein and 7.1 +/- 2.0 fmol/mg protein in the outer zone, but this difference is not statistically significant (Mann-Whitney W=361.5, p<0.4. Table 5 xii.. Fig.5 (ix)).

178 Table 5 xii. Regional Distribution ER (fmol/mg protein)

Patient Inner Zone Outer Z

1. HT 26.9 4.1 2. HG 31.6 29.3 3. CG 6.1 13.6 4. FD 0 0 5. WW 10.1 12.5 6. JW 35.6 10.2 7. JAs 4.8 7.9 8. PB 1.4 0 9. SG 9.7 0 10. HS 1.6 13.7 11. HR 20.6 3.9 12.JAt 4.0 0 13. RG 10.0 0 14. JS 15.3 12.0 15. RW 0 3.6 16. PK 26.4 12.0 17. FS 0 0 18.MD 25.1 16.5

Mean 12.1 7.1

S.D. 12.5 8.4

S.E.M. 3.0 3.0

Flg.5 (ix) Regional Distribution ER Binding BPH (Biopsy Specimens)

ER fmol/mg protein p -0.4 50

inner Zone Outer Zone

179 4. Progesterone Receptor Distribution

There was no statistically significant difference between

PgR status in the inner zone compared with the outer zones in BPH tissue sampled. Specific PgR binding could be detected in 50% (9/18) of tissue samples from the inner zone compared with 39% (7/18) samples from the outer zone. PgR concentrations were found to be very low in both inner (7.0 +/- 2.1 fmol/mg protein) and outer zones (3.4 +/- 1.2 fmol/mg protein) (Mann-Whitney W=366.5, p<0.3. Table 5 xiii., Fig.5 (x)). Table 5 xiii. Regional Distribution PgR (fmol/mg protein)

Patient Inner Zone Outer Zone

1. HT 31.5 0 2. HG 13.6 14.1 3. CG 11.7 3.9 4 FD 0 0 5. WW 4.3 0 6. JW 10.2 5.4 7. JAs 0 0 e. PE 0 0 9. SG 5.7 0 10.HS 16.0 0 11.HR 11.3 6.2 12.JAt 0 1.1 13.RG 2.6 6.5 14. JS 0 0 15.RW 2.9 6.4 16.PK 12.9 12.3 17. FS 4.7 0 18.ND 12.8 10.6

Mean 7.0 3.4

S.D. 8.7 4.9

S.E.M. 2.1 1.2

180 Fig.5 (x) Regional Distribution PgR Binding BPH (Biopsy Specimens)

PgR fmol/mg protein 40 r p -0.3

inner Zone Outer Zone

5. Discussion

It was possible to compare the results of the various assays in 18 patients who had all three performed on their prostatic tissue, both inner and outer zones (Table 5 xiv). The negative correlation that was shown in the previous section between EGFR and ER is maintained in this series but only in tissue taken from the inner zone

(Spearman rho=-0.72, p< 0.002) (Fig. 5 (xi)). In the outer zone this correlation did not exist (Spearman rho=-0.29, p<0.15) (Fig.5 (xii)). There was no significant correla tion between EGFR binding and PgR in either the inner zone

181 or outer zone specimens (Table 5 xiv.). This may be ac counted for by the many negative results for PgR in speci mens from both zones in this series (50% PgR -ve, inner zone and 61% PgR -ve, outer zone). However when the speci mens from the inner zone were ER positive the majority were also PgR positive (8/11), in the outer zone the minority (4/9) of ER positive specimens were also PgR positive. This parallels the distinction between diseased (tumour) tissue of the breast in which an inverse ER-EGFR relation ship can be found, when compared to normal breast where no such relationship is evident (Barker et al. 1989), thus highlighting the abnormality of the BPH tissue.

Fig.5 (xi) EGFR V ER Binding BPH Inner Zone

EGFR fmol/mg protein 140 r Spearman P —0.72 p < 0.002 120 r

100

0 10 20 30 40 ER fmol/mg protein

182 Table 5 xiv. Comparison of EGFR and Steroid Receptor binding in the inner and outer zones of the prostate

EGFR ER PgR Patient K OZ IZ OZ OZ

1. HT 64.9 0 26.9 4.1 31.5 0 2. HG 40.2 0 31.6 29.3 13.6 14.1 3. CG 127.8 53.8 6.1 13.6 11.7 3.9 4. FD 106.6 48.6 0 0 0 0 5. WW 42.9 - 10.1 12.5 4.3 0 6. JW 7.3 9.9 35.6 10.2 10.2 5.4 7. JAs 62.1 23.5 4.8 7.9 0 0 8. PB 106.7 22.8 1.4 0 0 0 9. SG 87.6 17.2 9.7 0 5.7 0 10. HS 67.2 6.8 1.6 13.7 16.0 0 11.HR 105.8 54.2 20.6 3.9 11.3 6.2 12.JAt 116.0 35.0 4.0 0 0 1.1 13. RG 55.8 7.8 10.0 0 2.6 6.5 14. JS 91.6 32.3 15.3 12.0 0 0 15. RW 119.3 3.8 0 3.6 2.9 6.4 16. PK 2.5 7.9 26.4 12.0 12.9 12.3 17. FS 118.2 31.7 0 0 4.7 0 18.MD 42.3 7.5 25.1 16.5 12.8 10.6

Fig.5 (xii) EGFR V ER binding BPH Outer Zone

EGFR fmol/mg protein Spearman yO—0.29 60

p-0.15 50

10 15 20 25 30 ER fmol/mg protein

183 V SERUM HORMONE LEVELS

1. BPH V Controls

Serum hormone levels were measured in 25 of the patients with BPH and in a control group of 15 post pubertal males under the age of 40. The hormones measured were testoster one (T), oestradiol (E2), dihydrotestosterone (DHT), and sex hormone binding globulin (SHBG), using the methods outlined in Chapter 3 (Table 5 xv.)

Table 5 xv.

Mean values of serum hormone levels (nmol/1 unless stated)

BPH (n=25) Controls (n=15)

T 14.0 -k/- 0.9 23.0 -K/- 1.6 DHT 0.40 -K/- 0.07 0.46 +/-0.07 E2 111.2 +/-20.4 80.4 +/-8.5 pmol/1

SHBG 43.8 +Z-2.5 22.3 +/-2.0 T/E2 ratio 0.13 0.29

There is a significantly lower testosterone/oestradiol ratio (T/E2) in the group of patients with BPH compared to the controls. This is accounted for by a highly signif icant fall in testosterone (Mann Whitney W=367.5, p<0.0001), rather than a rise in oestradiol (Mann Whitney

184 W=520, p<0.84) (Fig.5 (xiii)). There is no difference in the mean DHT levels, but there is a higher SHBG level in the BPH group (Mann Whitney W=674.5, p<0.0001).

Fig.5 (xiii) Steroid Hormone and Binding Globulin Levels in Control and BPH Groups

160 Control BPH 140

120

100

20 -3 -

T E2 SHBG E2 SHBG T and SHBG - nmol/l E2 • pmol/l

2. Serum hormone levels, steroid receptor and EGF receptor binding

The serum hormone levels were measured in the same group of BPH patients who subsequently underwent TURP. The resected tissue was then used for biochemical analysis (EGFR and ER assay), as described earlier in this chapter

This enabled comparison of the serum hormone levels and tissue receptor binding (EGFR n=25, ER n=15), in these patients.

185 However no direct correlation between the serum levels of oestradiol (Spearman rho=0.29), testosterone (rho=-0.02), DHT (rho=-0.02) and EGFR binding was found. The same was true for oestradiol (rho=-0.46) and prostatic ER binding

(Table 5 xvi.).

Table 5 xvi.

Serum Hormone levels, EGFR and ER binding in BPH (nmol/l), (E2 = pmol/1)

Patient E2 T DHT SHBG EGFR

1. GC 105 14.5 0.71 64 _ 118.7 2. JW 89 22.3 0.91 40 33.6 29.3 3. WW 31 13.1 0.27 58 21.1 47.1 4. HT 361 9.8 1.05 30 15.2 60.5 5. CG 93 11.8 0.80 39 10.2 81.4 6. JS 72 11.1 0.56 28 - 97.8 7. FS 427 10.5 0.20 33 - 121.6 8. MG 90 18.8 0.15 64 - 27.2 9. FD 47 10.9 0.18 28 3.7 86.1 10. BK 57 7.8 <0.01 58 34.6 42.9 11.JW 92 11.5 0.13 66 - 22.7 12.JAs 40 20.4 0.89 38 37.3 29.7 13.JD 53 16.5 0.23 45 - 96.8 14. PB 111 11.9 <0.05 47 4.6 86.4 15. SG 92 10.3 0.09 35 - 84.4 16. HS 59 12.0 0.04 50 - 109.5 17.AF 90 17.8 0.09 54 41.6 15.9 18.HR 125 16.3 0.14 36 - 75.0 19.JAt <20 37.3 <0.03 43 37.3 29.7 20.ET 75 18.0 0.76 60 52.8 17.0 21.RG 316 14.0 <0.01 46 12.1 36.9 22.RW 150 11.3 0.41 27 - 121.4 23.EL 64 14.9 0.94 28 16.6 58.5 24.PK 59 17.5 0.55 47 - 17.7

186 3. Discussion The measurement of serum hormone levels is usually only useful when comparing large populations, unless there are very clear differences. Previous studies comparing these hormone levels in patients with BPH, age matched controls and younger patients have not shown clear differences (Harman and Tsitouras 1980). Although this is a small group, the measurement of serum hormone levels was still valuable particularly as it offered an opportunity to com pare these with a new biochemical parameter in the same patient (EGFR). In common with other series (Vermeulen and de Sy 1976, Rannikko and Aldercreutz 1983), there was a lower mean testosterone level in the BPH group, leading to a shift in the testosterone/oestradiol ratio (Fig.5 (xiv)).

Fig.5 (xiv) Testosterone/Oestradiol Ratio Controls v BPH

0.4

0.35

0.3

0.25

0.2

0.15

0.1

0.05

0 Control BPH

187 However the testosterone/oestradiol ratio did not directly correlate with the level of EGFR binding in the prostate of these men, although a mean increase in EGFR binding in BPH has been shown previously. Tissue hormone levels may be more meaningful, particularly with the rise in SHBG in the BPH group, which may exert a controlling mechanism on the delivery of testosterone from the serum. However it was not possible to measure these due to the small quanti ties of tissue samples.

188 CHAPTER SIX

MORPHOMETRIC RESULTS

189 MORPHOMETRIC RESULTS

I INTRODUCTION 191

II MORPHOMETRY IN BPH ANDNORMAL PROSTATE 193

- using computerised image analysis

1. Glands 193 2. Epithelium and Acinar lumen 193

3. Stroma 196 4. Discussion 196

III COMPARISON WITH PATHOLOGIST 198

IV COMPARISON OF BIOCHEMICAL AND MORPHOMETRIC RESULTS 200

1. EGFR 200 2. Steroid receptors 203 3. Steroid hormone levels 206

190 I INTRODUCTION

The results of the morphometric analysis of each component of prostatic tissue using computerised image analysis are presented as a percentage of the total amount of tissue examined and have been adjusted for each field which did not contain 100% of tissue. Table 6 i. shows an example of the data from the analysis of tissue from one patient and the subsequent calculations required in order to reach an objective figure for each tissue component. This can then be used to compare results between patients, controls and with biochemical data for the same tissue samples. Results are available for 27 patients studied (Table 6. iv.). Although tissue samples were collected from 30 patients, 3 samples were disregarded as 2 had focal areas of carcinoma and 1 had histological evidence of granulomatous prostati tis.

The 5 control tissue specimens were taken from young post pubertal males (< age 40), who died of other causes and had no histological evidence of BPH (Normal A). These tissue specimens fixed and prepared in exactly the same manner as the BPH specimens. Unfortunately no tissue from these patients was available for biochemical analysis. The second control group of 9 normal prostates was taken from older patients (mean age 63) undergoing other procedures which included removal of the prostate (Normal B ). Tissue from this group was also used for biochemical analysis and the results are given in Chapter 5.

191 Table 6 i. Patient JA

Table shows data obtained and subsequent calculations for 20 histological fields examined using image analysis for patient JA. Prostatic tissue from all other patients included in the series was examined in the same way.

A B c DEF G %Tissue Glands %Glands Lumen %Lumen stroma Eplthelii (E+L) (B/AxlOO%) (D/AxlOO%) (100-C) (C-E)

67.5 15.4 22.8 8.3 12.3 77.2 10.5 99.2 6.6 6.6 1.5 1.5 93.4 5.1 99.2 8.2 8.2 3.1 3.1 91.8 5.1 46.4 7.4 15.9 2.1 4.5 84.1 11.4 46.7 12.9 27.6 4.7 10.0 72.4 17.6 82.5 13.7 16.6 7.2 8.7 83.4 7.9 49.2 8.5 17.2 3.4 6.9 82.8 10.3 84.3 24.7 29.3 14.3 16.9 70.7 12.4 92.3 26.2 28.4 15.1 16.3 71.6 12.1 43.2 5.6 13.0 2.9 6.7 87.0 6.3 94.5 22.6 23.9 13.1 13.9 76.1 10.0 99.2 8.6 8.6 2.3 2.3 91.4 6.3 99.2 22.7 22.7 12.9 12.9 77.3 9.8 56.2 10.1 18.0 4.2 7.5 82.0 10.5 99.2 7.8 7.8 2.4 2.4 92.2 5.4 82.7 14.5 17.5 5.8 7.0 82.5 10.5 89.7 15.5 17.2 7.3 8.1 82.8 9.1 99.2 26.4 26.4 14.3 14.3 73.6 12.1 68.5 17.9 26.1 8.7 12.7 73.9 13.4 99.2 23.7 23.7 10.4 10.4 76.3 13.3

Mean 18.8 8.9 81.1 9.9

S.D. 7.3 4.7 7.3 3.2

S.E.N. 1.6 1.1 1.6 0.7

192 II Results

1. Glands

In the 5 normal prostates the mean percentage of glands (epithelium + acinar lumen) was 47.9 +/- 3.1% (Table 6 ii.). This was lower (38.1 +/-2.1%) in the prostatic tissue removed from older males with no overt evidence of BPH, but this was not statistically significant (Table 6 iii.). However there was a significantly lower proportion of glandular tissue in the samples of BPH (26.3 +/-2.3), but with a range of 10.8-46.7% (Table 6 iv.).

2. Epithelium and Acinar Lumen

In order to determine whether the reduction in glandular tissue in these patients with BPH was predominantly due to glandular involution, with loss of lumen size, or loss of epithelium, the relative proportions of epithelium and acinar lumen were calculated. There seemed to be relatively less epithelium in the older controls, (12.8+/-1.2%) compared to the post puber tal group (16.7 +/-1.1%) (Tables 6 ii. and iii.). No sig nificant difference was found (Mann Whitney p>0.05) be tween either of these and the values obtained for BPH tissue (16.2+/-1.4%, Table 6 iv.).

193 However the acinar lumen component of the glandular tissue was greatly reduced in BPH tissue compared with both the normal groups (10.4+/-1.3% versus 31.2+/-2.4% and 25.3+/- 2.9%, p<0.0001).

Table 6 ii.

Morphometric Data for Normal Post Pubertal Prostate

No. %Glands %Lumen %Eplthellum %Stroma G/S Ratio

1. 48.9 29.2 19.7 51.1 0.96 2. 57.3 38.7 18.6 42.7 1.34 3. 42.6 28.6 14.0 57.4 0.74 4 39.7 25.2 14.5 60.3 0.66 5. 51.0 34.2 16.8 49.0 1.04

Mean 47.9 31.2 16.7 52.1 0.95

& D 6.9 5.3 2.5 6.9 0.27

SEM 3.1 2.4 1.1 3.1 0.12

Table 6 iii. Morphometric Data For :Normal Prostate (mean age 54)

Pt tGlands %Lumen %Epitheliuin %Stroma G/S Ratio

1. ws 32.7 16.9 15.8 67.3 0.48 2. MD 41.4 35.6 5.8 58.6 0.71 3. ES 39.7 26.4 13.3 60.3 0.66 4. JV 47.6 33.5 14.1 52.4 0.90 5. RW 42.3 31.9 10.4 57.7 0.73 6. PD 28.3 10.8 17.5 71.7 0.39 7. NS 35.2 19.1 16.1 64.8 0.54 8. VG 44.6 33.0 11.6 55.4 0.80 9. CS 31.1 20.7 10.4 68.9 0.45

Mean 38.1 25.3 12.8 61.9 0.61

S.D. 6.6 8.8 3.6 6.7 0.17

S.E.M. 2.1 2.9 1.2 2.2 0.06

194 Table 6 iv. Morphometric Data for BPH Patients

pt. %Glands %Lumen %Eplthelium %Stroma G/SRatlo

1. CM 29.7 7.2 22.5 78.3 0.34 2. MD 14.1 2.8 11.4 85.9 0.16 3. AF 16.9 4.5 12.4 83.1 0.20 4. BK 29.0 8.2 21.8 71.0 0.40 5. FW 36.2 12.4 23.8 63.8 0.56 6. EL 20.5 4.2 16.3 79.5 0.26 7. RG 24.5 20.3 5.9 75.5 0.32 8. ET 12.8 2.7 10.1 87.2 0.15 9. JAs 26.4 7.2 19.2 73.6 0.36 10. HR 26.3 8.4 17.9 73.7 0.36 11. PB 35.2 11.7 23.5 64.8 0.54 12.JAt 15.2 5.6 9.6 84.8 0.18 13.JW 13.7 7.0 6.7 86.3 0.16 14. WW 13.0 3.7 9.3 87.0 0.15 15.FD 41.2 23.2 18.0 58.8 0.70 16.CG 45.7 18.9 26.4 54.3 0.84 17.HG 11.5 5.4 6.1 88.5 0.13 18.HT 43.6 26.4 17.2 56.4 0.77 19.MA 46.7 19.4 27.3 53.3 0.88 20. HS 23.6 14.0 15.9 70.0 0.34 21. SE 15.8 6.9 8.9 84.2 0.19 22. FC 35.6 15.7 19.9 64.4 0.55 23.GR 19.7 8.2 11.5 80.3 0.24 24. WD 10.8 2.9 7.9 89.2 0.12 25. SG 44.8 15.6 29.2 55.2 0.81 26. WB 20.9 7.5 13.4 79.1 0.26 27.MS 36.5 10.4 26.1 63.5 0.57

Mean 26.3 10.4 16.2 73.8 0.39

S.D. 11.8 6.6 7.1 11.8 0.24

S.E.M. 2.3 1.3 1.4 2.3 0.05

195 3. Stroma

In the prostatic tissue examined from younger men the proportion of glandular tissue (epithelium + lumen) was

almost the same as of stromal tissue, with 47.9% of the

tissue occupied by glands and the remainder (52.1%) by stroma. This gives a mean glandular to stromal ratio (6/S ratio) of 0.95, which is lower in the older control group (0.61), due to a slightly lower proportions of epithelium and acinar lumen. However in BPH the significant reduction in acinar lumen volume is compensated by the rise in the proportion of stroma. In BPH, the stroma occupies a mean area of 73.8+/-2.3%, giving a mean G/S ratio of 0.39 but with a range from 0.12-0.88 (Table 6 iv).

4. Discussion

These results show that a new rapid and automated method for quantifying the tissue components of prostate speci mens has been successfully developed. Changing glandular/ stromal ratio with age and changes in the morphometric parameters with age have been demonstrated (Fig.6 (i). In making these measurements it is still necessary to examine at least 20 random fields, which is the main limiting factor in obtaining even more rapid results. Hopefully

this can be automated in the future.

196 Fig.6 (i) Morphometric Analysis of Normal Prostate and BPH - Using Image Analysis

100%

75% T 0 t a 1 50% G I a n 25% d

Ratio 0% Normal A Normal B BPH Glandular (L+E) stromal (8) 0.95 0.61 0.35

Computer assisted colour image analysis is presently being used to quantify the area densities of tissue components based on differential staining properties of the connective tissue and glands (Shapiro et al. 1991). This requires a double immunoenzymatic staining technique

to stain smooth muscle brown, epithelium dark red, connec tive tissue light brown, leaving the glandular lumen colourless. This is a complex procedure and may cause artefactual shrinkage of individual elements leading to

distortion of true ratios. The technique used in this study is unique as it can be applied to standard haematoxylin and eosin stained sections. The results confirm that

BPH is predominantly a stromal disease (Bartsch et al. 1979), but there is a wide range of glandular hyperplasia

197 Ill COMPARISON WITH PATHOLOGIST

An experienced pathologist was asked to examine the histological slides of prostate tissue taken from the BPH patients in order to confirm the diagnosis. He was also asked to grade the degree of glandularity (ie. whether or not there was predominantly glandular or stromal hyperpla sia ). After initial examination the pathologist felt that he would only be able to grade on a relatively crude scale within a narrow range (1-4), (where 1 = predominantly stromal hyperplasia and 4 = predominantly glandular hyper plasia). The results of this exercise are shown in Table 6

V . where the glandular/stromal ratio obtained from morpho metric analysis is compared to the pathologists "degree of glandularity". Despite the narrow range used the pathologist still found it difficult to distinguish grades solely on the basis of the histological assessment of the amount of glandular or stromal tissue present. The majority were judged to be in group 2, ie. mainly stromal hyperplasia, which is consis tent with the morphometric data (Table 6 iv.). He was more easily able to categorise the patients at the ex tremes of the series (ie. 1 or 4. Table 6 v.). It is important to note that the naked eye appearances of these sections did not conflict with the morphometric results obtained by image analysis. Although this exercise helps to validate the morphometric data, it also demonstrates

198 that the naked eye assessement of glandularity is general ly unhelpful. It cannot categorise the morphology in numerical terms with sufficient accuracy to be useful. Morphometric data obtained using the new technique de scribed above seems to achieve both of these criteria.

Table 6 v. Glandular/Stromal ratio y "Degree of G1andularity"

No. Patient G/S Ratio Glandularity

1. CM 0.34 2 2. ND 0.16 2 3. AF 0.20 2 4 BK 0.40 2 5. FW 0.56 3 6. EL 0.26 2 7. RG 0.32 2 8. ET 0.15 1 9. JAs 0.36 1 10. HR 0.36 1 11. PB 0.54 1 12. JAt 0.18 1 13. JW 0.16 2 14. WW 0.15 1 15. FD 0.70 3 16. CG 0.84 4 17. HG 0.13 2 18. HT 0.77 2 19. MA 0.88 4 20. HS 0.34 2 21. SE 0.19 2 22. FC 0.55 2 23. GR 0.24 1 24. WD 0.12 1 25. SG 0.81 3 26. WB 0.26 2 27. MS 0.57 3

199 IV CORRELATION OF BIOCHEMICAL AND MORPHOMETRIC RESULTS 1^ EGER

The morphometric data for normal and BPH tissue were com pared with the biochemical analysis of the same tissue. No relationship could be found between EGER and the amount of glandular tissue (epithelium alone or glandular acini), or the percentage of stroma in the normal tissue taken from older patients with no evidence of BPH (Normal B) (Table 6 vi).

Table 6 vi. Morphometric and biochemical analysis of normal prostate

Ho. Pt %G1 *L %E %S EGFR PoR

1. WS 32.7 16.9 15.8 67.3 15.7 23.3 9.2 2. MD 41.4 35.6 5.8 58.6 21.4 15.6 20.5 3. ES 39.7 26.4 13.3 60.3 0 33.1 0 4. JV 47.6 33.5 14.1 52.4 34.6 38.6 13.8 5. RW 42.3 31.9 10.4 57.7 9.8 33.4 24.3 6. PD 28.3 10.8 17.5 71.7 14.2 8.2 3.9 7. NS 35.2 19.1 16.1 64.8 23.1 21.4 1.2 8. VG 44.6 33.0 11.6 55.4 7.9 9.2 0 9. CS 31.1 20.7 10.4 68.9 8.4 28.8 0

Although these results pertain to a relatively small group of patients, they contrast with those obtained from exami nation of the prostatic tissue in patients with BPH. In the BPH group there is a significant correlation between the proportion of glandular tissue, epithelium alone, glandular/stromal ratio and EGER binding measured by

radioligand binding assay in that tissue (Table 6 vii. and

Eigs.6 (ii,iii,iv).

200 This is not surprising since EGFR is found in the epi thelium, and not in the stroma. EGFR binding studies were carried out on a homogenate of all cell types, therefore, the more epithelium, the more EGFR. BPH tissue has been shown to have a higher proportion of stroma, so it is sur prising to find higher EGFR levels in BPH than in normal tissue. The immunohistochemistry data suggests that this is due to overexpression or higher expression of EGFR within BPH epithelium compared to normal epithelium.

Table 6 vii. Morphometric and biochemical analysis of BPH

Wo. Pt. %G1 %Ep %st G/SRatio EGFR

1. CM 29.7 22.5 78.3 0.34 57.9 2. MD 14.1 11.4 85.9 0.16 27.2 3. AF 16.9 12.4 83.1 0.20 15.9 4 BK 29.0 21.8 71.0 0.40 42.9 5. FV 36.2 23.8 63.8 0.56 75.9 6. EL 20.5 16.3 79.5 0.26 58.5 7. RG 24.5 5.9 75.5 0.32 36.9 8. ET 12.8 10.1 87.2 0.15 17.0 9. JAs 26.4 19.2 73.6 0.36 75.1 10. HR 26.3 17.9 73.7 0.36 75.0 11. PB 35.2 23.5 64.8 0.54 86.4 12. JAt 15.2 9.6 84.8 0.18 54.7 13. JWi 13.7 6.7 86.3 0.16 29.3 14. WW 13.0 9.3 87.0 0.15 47.1 15. FD 41.2 18.0 58.8 0.70 86.1 16. CO 45.7 26.4 54.3 0.84 98.2 17. HG 11.5 6.1 88.5 0.13 34.1 18. HT 43.6 17.2 56.4 0.77 60.5 19. MA 46.7 27.3 53.3 0.88 79.1 20. HS 23.6 15.9 70.0 0.34 35.2 21. SE 15.8 8.9 84.2 0.19 22.9 22. FC 35.6 19.9 64.4 0.55 42.7 23. GR 19.7 11.5 80.3 0.24 36.8 24. WD 10.8 7.9 89.2 0.12 14.6 25. SG 44.8 29.2 55.2 0.81 87.6 26. WB 20.9 13.4 79.1 0.26 22.7 27. MS 36.5 26.1 63.5 0.57 79.7

201 Fig.6 (il) Epithelium % v EGFR Binding (BPH)

EGFR fmol/mg protein Spearman P ■ 0.7 8 120 p<0.001

100

0 0 5 10 15 20 25 30 % Epithelium

Fig.6 (iii) Glands (Epithelium + Acinar lumen) (BPH) V EGFR Binding Spearman P - 0 .8 2 EGFR fmol/mg protein 120 p<0.001

100

0 10 20 30 40 50 % Glands

202 Fig.6 (iv) G/S Ratio V EGFR Binding (BPH)

EGFR fmol/mg protein 120 Spearman P ■ 0 .7 9 p<0.001 100

0.2 0.4 0.6 0.8 G /S Ratio

2. steroid Receptors

The morphometric data and results from the assay of ster oid hormone receptors (ER and PgR) in the prostatic tissue were analysed in the same way. As with EGFR binding in the normal prostatic tissue there was no relationship between the assay of the steroid receptors and any of the morpho metric parameters (Table 6 vi.). In the BPH tissue there is a good correlation between the amount of stroma in the specimen and the number of ER

(Table 6 viii., Fig.6 (v)) and a corresponding negative correlation with the glandular/stromal ratio (Fig.6 (vi)).

This relationship is found irrespective of whether PgR is detectable in the same specimen.

203 These results are to be expected since a significant nega tive correlation between EGFR and ER has already been shown, and EGFR seems to be directly related to the amount of epithelial or glandular tissue.

Table 6 viii.

Morphometric and steroid receptor analysis of BPH

No. Pt. %G1 %Ep *st G/SRatio Fo r

1. CM 29.7 22.5 78.3 0.34 10.1 20.6 2. MD 14.1 11.4 85.9 0.16 57.1 13.9 3. AF 16.9 12.4 83.1 0.20 41.6 11.5 4. BK 29.0 21.8 71.0 0.40 34.6 16.6 5. FW 36.2 23.8 63.8 0.56 3.1 0 6. EL 20.5 16.3 79.5 0.26 16.6 11.0 7. RG 24.5 5.9 75.5 0.32 12.1 0 8. ET 12.8 10.1 87.2 0.15 52.8 28.7 9. JAs 26.4 19.2 73.6 0.36 0 0 10. HR 26.3 17.9 73.7 0.36 0 0 11. PB 35.2 23.5 64.8 0.54 4.6 0 12. JAt 15.2 9.6 84.8 0.18 37.3 15.5 13. JWi 13.7 6.7 86.3 0.16 33.6 0 14. WW 13.0 9.3 87.0 0.15 21.1 11.5 15. FD 41.2 18.0 58.8 0.70 3.7 0 16. CO 45.7 26.4 54.3 0.84 10.2 0 17. HG 11.5 6.1 88.5 0.13 33.8 23.7 18. HT 43.6 17.2 56.4 0.77 15.2 36.5 19. MA 46.7 27.3 53.3 0.88 0 0

204 Fig.6 (v) % Stroma v ER Binding (BPH)

ER fmol/mg protein 60 Spearman P ■ 0 .7 5

p<0.01 50

40 50 60 70 80 90 100 %Stroma

Fig.6 (vi) G /S Ratio V ER Binding (BPH)

EGFR fmol/mg protein 60 Spearman P ■ -0 .7 4

p<0.01 50

0.2 0.4 0.6 0.8 G/S Ratio

205 3. Steroid hormone levels

Serum was obtained for steroid hormone assay in 17 of the patients whose BPH tissue was suitable for morphometric analysis. Although this was a small group it was an oppor tunity to try to determine if the testosterone/oestradiol ratio had any direct influence on the glandular/stromal ratio in BPH (Table 6 ix.). If a high or low testoster one/oestradiol ratio could predict whether or not there was predominantly glandular or stromal hyperplasia in the patients prostate gland, it would have potentially useful clinical implications. The higher the glandular/stromal ratio, the more sensitive the prostatic tissue to androgen deprivation (Coffey 1988) and this might mean that a blood test could potentially predict the response to medical treatment with aromatase inhibitors or 5-alpha reductase inhibitors. Unfortunately no such simple relationship could be shown in this group (Fig.6 (vii)) and a study in a larger group of patients may be indicated. Morphometric data from prostatic specimens using image analysis would allow this to be done in the future. However the testosterone/oestra diol ratio does not seem to alter (6th/7th decade of

life), until well after BPH has begun to develop (4th/5th decade).

206 Table 6 ix.

Morphometry and serum steroid hormones in BPH

Wo. Pt T/E2 ratio G/S ratio

1. JW 0.25 0.16 2. WW 0.42 0.15 3. HT 0.02 0.77 4. CG 0.13 0.84 5. FD 0.23 0.70 6. BK 0.14 0.40 7. JWi 0.12 0.16 8. JAs 0.51 0.36 9. PB 0.11 0.54 10. SG 0.11 0.81 11. HS 0.20 0.34 12. AF 0.20 0.20 13. HR 0.13 0.36 14. JAt 1.86 0.18 15. ET 0.24 0.15 16. RG 0.04 0.32 17. EL 0.23 0.26

Fig.6 (vii) G /S Ratio V T/E2 Ratio (BPH)

T /E 2 Ratio 1 r

0.8

0.6

0.4

0.2

0.5 1.5 G/S Ratio

207 CHAPTER SEVEN

GENERAL DISCUSSION

208 GENERAL DISCUSSION

1.Introduction 210

2. Paracrine growth control 210

3. Regional variation in the prostate 212

4. Growth factors and steroid hormones 214

5. Biomorphometry 217

6. Conclusion/future studies 220

209 GENERAL DISCUSSION

1. Introduction

This final section reviews the findings of this study, comparing and contrasting the results with previous work. In doing so it aims to identify the ways in which these studies help to further understanding of the pathogenesis of BPH, and how the findings relate to existing theories. The possible application of this work to clinical practice and particularly medical management of BPH is reviewed. Finally, the extension of this work and the possibility of future studies are discussed.

2. Paracrine Growth Control

The demonstration of epidermal growth factor binding in homogenates of BPH or by immunocytochemistry is not new (Maddy et al. 1987, Eaton et al. 1988, Fowler et al. 1988). However there are differences between researchers in their assays and which receptor (high or low affinity) is measured. This has led to differences in absolute values obtained for BPH samples and carcinoma samples (Maddy et al. 1987, Eaton et al. 1988). The assay developed in this study predominantly measured high affinity receptors but is not without interference from low affinity receptors. The assay was shown to be

210 reproducible, and assay conditions remained stable between samples. We can be confident, therefore, that the differ ences between specimens are genuine. This Is Important since It allows confidence In the comparison between patients and the relationship to steroid receptors and morphometry. Although slightly lower values were obtained In TURP specimens EGFR could still be measured. The characteristics of EOF binding In BPH and normal tissue In this study are similar to those reported for other mammalian tissues (Menard et al. 1987, Gill et al. 1987, Carpenter 1987) and those by Maddy et al. (1987) for human BPH. Two receptor sites may be present In the same cell population or may represent the epithelial (high affinity site) and mesenchymal (low affinity site) In human BPH (Lubrano et al. 1989). The evidence from studies on breast cancer Is that the high affinity site Is prefer entially expressed In epithelial cells (Salnsbury et al.

1985, King and Cuatrecasas 1982). The Immunohlstochemlcal data In this study clearly show that the EGF receptors detected by a monoclonal antibody to the external domain of the receptor, are confined to the prostatlc acinar epi thelium. This finding confirms the results of previous studies (Maddy et al. 1987). The ability to demonstrate EGF binding In BPH, by any method, does not necessarily Indicate that EGF functions as a major mitogen In BPH. Indeed, some believe that from the evidence of other radiolabelled binding studies, using

211 a mitogen found in BPH and young adult prostates (Jacobs et al. 1979), that EGF is not the major growth promoting agent in extracts of prostatic tissue (Story et al.

1983). However, both the immunohistochemical and binding studies presented here seem to suggest that EGFR is more highly expressed in BPH epithelium compared with ’normal' epithelium. It is also likely that EGF acts in synergy with other growth factors such as basic fibroblast growth factor (bFGF), insulin like growth factor (IGF-1) or the transforming growth factors (TGF alpha and beta).

3. Regional Variation in the prostate

A particular feature of BPH and a potential area for fur ther research has been the observation, originally made many years ago, that BPH principally affects the central portion of the gland (Morgagni 1761) whereas the surround ing gland is spared. Since then detailed studies have shown that the area of origin of BPH is confined to a

group of glands surrounding the urethra, otherwise known as the transitional zone (McNeal 1972).

There are also some minor morphological differences be tween the two regions of the gland but these do not ac count for the pathological differences. These studies are a further attempt to investigate the biochemical differ

ences which may give a clue to the pathological behaviour of the regions of the prostate. Previously, Habib et al.

212 (1983) have demonstrated some differences in the distribu tion of the enzyme 5-alpha reductase which converts testosterone to DHT. More recently Begun et al. (1989) were the first group to show a variation of a growth factor (bFGF) within the prostatic zones. This is the first study to show regional variation in EGF binding within the prostate. The increased EGF binding in the inner zone where BPH arises supports a role for EGF in the pathogenesis of this condition. There certainly appears to be up-regulation of EGF receptor binding but it is diffi cult to know the significance of this in the development of BPH and which stimuli provoke up-regulation. There are differing opinions about which prostatic tissue consti tutes a 'normal* control and admittedly the lack of a good control group raises a question mark over the validity of this finding. However it is not possible, due to ethical considerations, to take biopsy specimens from the normal prostate in either young or elderly males. The normal prostatic tissue that was obtained was often of poor quality and sampling of each area was not controlled by ultrasound, although tissue was removed as accurately as possible from the two zones. There was no difference in EGF binding between the two zones in the normal glands but this may be an artefact due to ischaemia time during the open operation. However BPH tissue taken by retropubic prostatectomy also had increased EGF binding despite similar ischaemia times. Aged matched controls with no

213 histological evidence of BPH are difficult to find. In the absence of ideal control material, post pubertal prostatic tissue from tissue donors (less than age 40) is often used. Using the patients own prostate as the control by sampling tissue from both normal and hyperplastic areas under ultrasound control is an excellent compromise as both areas are in the same hormonal milieu. By using this technique the results of this section of the study are more credible and allow a progression to further invest igation of the interaction of EGF with other steroid receptors and hormones.

4. Growth factors and Steroid Hormones

Although the aetiology of BPH may be multifactorial, steroid hormones are undoubtedly involved in both the normal and abnormal development of the prostate. These hormonal influences may be straightforward, such as a reduced testosterone/oestradiol ratio with ageing (Vermeu- len and De Sy 1976), or due to more complicated local biochemical modifications (Sciarra et al. 1988). It is possible that steroid hormones act through their own receptors to promote epithelial and stromal hyperplasia by influencing expression of growth factors and their recep tors (McKeehan et al. 1984). This study was designed to examine the relationship between steroid hormones and growth factors.

214 Unfortunately at this time there is no reliable means to assay androgen receptors in prostatic tissue, principally due to the heat labile nature of the receptor and the binding of ligands used in the assays to other proteins and receptors. The recent advent of a monoclonal antibody to the receptor may help to resolve this problem. The "hybrid ligand" technique developed in the Department of Biochemistry at St. Bartholomew's Hospital by Stebbings et al. (1987) to measure cytosolic androgen receptors was used on some prostatic homogenates. However, poor inter and intra assay reproducibility led to this approach being abandoned, until the development of a more reliable tech nique . In contrast, the assay of cytosolic oestrogen receptors in BPH using standard techniques is well established (Emtage et al. 1989). Oestrogens are known to be essential for the development of BPH in dogs (Walsh and Wilson 1976). The increased levels of oestradiol in BPH and sig nificantly lowered testosterone/oestradiol ratio in our patients is the same as in other larger series (Rannikko and Aldercreutz 1983, Zumoff et al. 1982). Furthermore, the abundance of oestradiol in the stromal tissue of BPH suggests an important role for oestrogen in the aetiology of human BPH (Kozak et al. 1982). However the mechanisms by which this may be mediated are not known. Regulation of EGFR expression in mammary cells by oestro gens In vivo and In vitro has been reported (Imai et al.

215 1982, Mukku and Stancel 1985). Following castration there are increased levels of EGFR in rat prostates (Traish and

Wotiz 1987, St.Arnaud et al. 1988). The results from the patients in this study show that ER seems to be important in regulating EGFR expression as they demonstrate an inverse correlation between ER and EGFR. This relationship is all the more significant as it is only present in the inner zone of the prostate in patients with BPH.

The negative correlation between ER and EGFR in this series contrasts with the positive correlation demonstra ted i n v i t r o between steroid hormones and EGFR. Up- regulation of EGFR has been reported after androgen treat ment of LNCap (prostate cancer) cells (Schuurmans et al

1988) and after progestin exposure of mammary carcinoma cell lines (T47D, MCF7) (Murphy et al. 1986) .

From these results it is possible to hypothesise that there may be a paracrine mechanism (stromal-epithelial interaction) between oestrogens and their receptors and

EGFR. This could be due to oestrogenic enhancement of EGF synthesis. It is known that EGF forms a ligand-receptor complex with EGFR which is internalised and degraded by lysosomes (Fernandez-Pol 1981,1982). In this way, in creased levels of EGF may cause down regulation of EGFR, whereas reduced EGF production may result in increased

EGFR synthesis.

It is unlikely that EGF negatively regulates ER synthesis or its activation/translocation in the nucleus, because

216 EGF has been shown to stimulate phosphatidyllnositol turnover (Sawyer and Cohen 1987), generating diacylglycer- ol which activates protein kinase C. The phosphorylation of EGFR by protein kinase C results in attenuation of EGF binding to its receptor and internalisation (Bell 1986, Sibley et al. 1988). However in human breast cancer cells (MCF-7) protein kinase C may phosphorylate PR and ER which increases the total number of steroid binding sites (Boyle and Van der Valt 1988, Knabbe et al. 1988). The results of this study at this stage favour the hypothesis of steroid hormone regulation of EGFR expression in human BPH. EGF may then be involved in the autocrine or paracrine regula tion of prostatic growth.

5. Biomorphometry

An attempt has been made to correlate the quantitative morphometric data with the biochemical values of the benign hyperplastic tissue, to determine whether or not steroid binding or growth factor binding is reflected in the morphologic architecture. The development of an effi cient method to evaluate the tissue components of each specimen has helped to make this possible. This is the first study to use automated image analysis to provide quantitative morphometric data from tissue which has also undergone biochemical analysis. It is also unique in that a measurement of a peptide growth factor receptor was included in the analysis.

217 Light microscopic stereological measurements of the normal prostate and BPH show that tissue overgrowth in BPH is mostly due to an increase in stromal tissue (Bartsch et al. 1979). The results of the morphometric analysis of BPH in this study are consistent with these findings. There is undoubtedly a shift in the relative proportions of glandu lar and stromal tissue in BPH. In fact a gradual change with age has been demonstrated by comparing the results for BPH patients compared with those for two normal con trol groups (Fig.6 (i)>). There is also glandular hyper plasia, with papillary epithelial hyperplasia (Fig.l (v)), which may be florid. This represents a more significant component of the hyperplastic prostate gland in some patients, and there is a range from 10.8% up to 46.7% of glandular tissue in our series. Coffey (1988) suggested that those patients with higher glandular to stromal ratios respond better to androgen deprivation. This is an important observation when considering medical treatment of BPH using any form of hormonal manipulation.

When the morphometric parameters are compared to the data for EGFR binding, it confirms what would be expected from the immunohistochemical studies of EGFR. Specific staining for the receptor (Fig.5 (v)) was confined to the glandular epithelium and there was a strong correlation between the amount of epithelium in the specimen and EGFR binding measured by biochemical assay. This same was not true in the normal prostate specimens but this may simply be due

218 to low levels of EGFR binding in this group. It is unlike ly that there are any significant differences in receptor distribution and this was confirmed by immunohistochemi stry of normal prostatic tissue.

Fractionation studies suggest that ER are predominantly found in the stromal portion of the prostate (Kreig et al. 1981). Comparison of the ER binding and morphometric data would support this finding. There is a strong positive correlation between ER binding and the amount of stroma in the tissue specimen. The latter results question the significance of the negative correlation which was found to exist between ER and EGFR. This relationship could simply have been due to the relative amounts of each tissue (glands/stroma) pre sent in the specimens examined rather than up or down regulation of either receptor. However this correlation does not exist in the tissue taken from the outer zone nor is it present in the normal prostatic tissue. This favours the development of the relationship with the advent of

BPH. Paracrine control may be involved in ultimately determining the relative amounts of glandular and stromal

hyperplasia with in the prostate.

The biomorphometric approach to the study of BPH patho genesis in this study has demonstrated the importance of relating the biochemical data to the morphometric para meters in order to understand these complex relationships.

219 6 . Cone1usion/Future Studies

The simplified methods of tissue analysis which have been developed should allow the clinical interest in growth factor and morphometric determination to be investigated in larger populations. This will be particularly important as new medical therapies for BPH are introduced. Biomor phometric studies of the kind described in this thesis should ultimately lead to a better understanding of the pathogenesis of BPH. It may be possible to select groups of patients who will respond to these new treatments and to stratify each patient according to the biochemical or morphometric characteristics of the BPH. As an example of this approach Shapiro et al. (1991) have studied the biomorphometry and the clinical response of patients with BPH who are treated with alpha adrenoceptor blockade. A direct relationship was observed between the percentage area density of prostate smooth muscle and the percentage increase in peak urinary flow rate after treat ment with the drug (Fig.7 (i)). An indirect relationship was observed between the percentage area density of prostate smooth muscle and percentage decrease in total symptom score in response to alpha blockade, although alpha-1 receptor density was similar in prostatic tissue obtained from asymptomatic and symptomatic men. They con cluded that the clinical response to alpha blockade was related quantitatively to the smooth muscle of the gland.

220 which supports the hypothesis that the therapeutic re sponse to alpha blockade is associated with relaxation of smooth muscle.

Fig.7 (i)

CORRELATION BETWEEN % A 0^.% AND % SMOOTH MUSCLE % Smooth Mutelo 50 I—

• 50 • 10 30 70 110 ISO 190 230 %&0.

Graph to show the correlation between percent increase in peak flow rate (% Qmax) and % of smooth muscle in the prostatic tissue taken at transurethral resection (Shapiro et al. 1991)

221 This observation may be more clinically relevant when non invasive imaging correlates with the morphometry. Early studies on this relationship, comparing the echo images and histology were promising (Hasegawa and Kumazama 1987).

As imaging quality improves (Jones et al. 1990), there is now a genuine possibility that treatment of the patient may be tailored to the type of prostatic hyperplasia.

Recent clinical studies in the Department of Urology at

St.Bartholomew's Hospital (Kirby et al. 1991) suggest that patients with BPH may respond by prostatic shrinkage and improved uroflow to treatment with the 5-alpha reductase inhibitor, finasteride. This drug blocks the conversion of the testicular androgen testosterone to dihydrotestoster one (DHT). DHT, not testosterone, is the key androgen modulating prostatic growth and maintenance, with testosterone acting as a pro-hormone (Bruchovsky and

Wilson 1968). Withdrawal of 5-alpha reduced androgen should eventually lead to prostatic shrinkage, which would relieve bladder outflow obstruction and improve symptoms.

However this may take several months and it will be impor tant to select patients who are most likely to respond to this new form of treatment. Patients with BPH that is predominantly glandular hyperplasia rather than stromal would theoretically respond better (Coffey 1988). It is hoped that a biomorphometric study of these patients will confirm this expectation and help selection of responders.

Furthermore the biochemical changes in the prostatic

222 tissue of patients being treated with this drug, particu larly the steroid receptor and growth factor binding, may help the understanding of the complex biochemical events leading to the development of BPH. This study has further demonstrated that steroid hormones can no longer be con sidered the sole regulators of prostatic growth. Other cellular moderators such as the growth factors probably act in synergy with steroid hormones or they may act independently. Epidermal growth factor is merely a part of the network of endocrine and paracrine growth control in the prostate. Despite this complex network of cellular regulation, disturbance of any one of these regulatory factors can produce abnormality of growth control. Other growth factors need to be studied and more specific assays of steroid receptors are now available. Such research should lead to a better understanding of the mechanisms responsible for abnormal prostatic growth and ultimately to the development of more effective treatments.

223 Appendix I

Validation Results

TB = Total Binding (dpm)

NSB = Non Specific Binding (dpm)

SB = Specific Binding (dpm)

CSA = Corrected Specific Activity (dpm/fmol) Protein = Protein (mg/100 ul)

Fig, 4 (i ) Duration of Incubation

Time (mine) TB NSB SB CSA Protein

30 1012 788 224 1384 0.122

60 1593 1303 290 1384 0.122

90 1843 1509 334 1384 0.122

120 1982 1609 373 1384 0.122

180 1898 1593 315 1384 0.122

Fig. 4 (ii) EGF concentration (nmol)

NM TB Mean TB

1 2053 2019 2321 2131

2225 2487 2193 2302

3898 3762 3600 3753

6098 6221 6395 6238

6833 6995 6873 6900

224 Fig 4 (iii) Influence of Dilution

DP = Dilution Factor

Assay Date (AD) = 19/12

TB MSB SB CSA Protein DF EGER

25721 11091 14630 1400 0.261 1 40.0

21095 11951 9144 1400 0.144 0.5 45.3

16121 10957 5164 1400 0.082 0.25 44.9

11981 9998 1983 1400 0.029 0.125 48.8

Table 4 Reproducibi1ity Intra-assay variation

TB NSB SB CSA Protein EGFR

Patient A AD - 12/12

16432 8923 7509 1518 0.121 40.8 15974 8071 7903 1518 0.121 43.0 17701 7993 9708 1518 0.121 52.8 16985 9831 7154 1518 0.121 38.9

Patient B AD - 12/12

19734 9976 9758 1518 0.097 66.2 18856 9571 9585 1518 0.097 63.0 18971 9935 9036 1518 0.097 61.3 19023 10645 8378 1518 0.097 56.8

Patient C AD - 13/12

17288 7729 9559 1501 0.089 71.5 14821 7324 7497 1501 0.089 56.1 17348 8067 9281 1501 0.089 69.5 18484 9135 9349 1501 0.089 70.0

225 Table 4 il. Reproducibi1ity Inter Assay Variation

TB NSB SB CSA Protein EGFR

Patient A AD - 14/12 and 18/12

14921 7880 7041 1484 0.115 41.3 13074 6909 6165 1416 0.121 35.9

Patient B AD - 7/12 and 11/12

18111 10572 7539 1608 0.094 49.8 16534 8253 8281 1536 0.097 55.5

Patient C AD - 14/12 and 18/12

10956 5843 5113 1484 0.097 35.5 9274 4971 4303 1416 0.112 27.1

Patient D AD - 7/12 and 13/12

17431 8043 9388 1608 0.082 71.2 18927 7929 10998 1500 0.114 67.3

226 Appendix II EGFR Results

TURP Specimens EGFR Assay

Date = Assay Date (date/month) TB = Total Binding (dpm)

NSB = Non Specific Binding (dpm) SB = Specific Binding (dpm)

CSA = Corrected Specific Activity (dpm/fmol) Protein = Protein (mg/100 ul) EGFR = EGFR Binding (fmol/mg protein) Table 5(i)

No. Pt. Date TB NSB SB CSA Protein EGFR

1. CM 5/12 18049 9271 8778 1646 .092 57.9 2. MD 5/12 17921 13085 4836 1646 .108 27.2 3. AF 5/12 12271 8991 3280 1646 .125 15.9 4. BK 5/12 12127 5842 6285 1646 .089 42.9 5. FW 5/12 19870 9994 9876 1646 .079 75.9 6. EL 6/12 16294 7530 8764 1627 .092 58.5 7. RG 6/12 14287 9714 4573 1627 .076 36.9 8. ET 6/12 12651 9991 2660 1627 .096 17.0 9. JAs 6/12 23185 13103 10082 1627 .117 52.9 10. HR 6/12 20897 9543 11354 1627 .093 75.0 11. PB 9/12 19005 8011 10994 1571 .081 86.4 12. JAt 9/12 14062 9951 4111 1571 .088 29.7 13. JW 9/12 10592 6070 4522 1571 .098 29.3 14. WW 9/12 15729 8257 7472 1571 .101 47.1 15. FD 9/12 19923 8015 11908 1571 .088 86.1 16. GO 10/12 18813 6218 11895 1553 .094 81.4 17. HG 10/12 15201 10167 5034 1553 .095 34.1 18. HT 10/12 17593 8101 9492 1553 .101 60.5 19. MA 10/12 19215 9015 10200 1553 .083 79.1

227 TURP Specimens EGFR binding using Scatchard analysis (SA)

Table 5 (ii) Protein = ug/tube

No. Pt TB SB CSA Kd(nmol) Protein EGFR

1. WW 2345 1158 1573 386 1570 383 1479 292 1335 148 1274 87 1241 54 1004.3 1.7 79.7 53.9

2. HG 1017 580 873 436 843 406 744 307 712 275 610 173 551 114 512 75 1127.2 0.9 56.8 80.5

3. RG 1612 746 1450 584 1173 307 1138 272 1030 164 1026 160 985 119 969 103 1127.2 2.3 92.5 70.5

4. BK 1661 533 1553 425 1497 369 1379 251 1288 160 1242 114 1195 67 1148 20 1127.2 2.1 76.2 90.5

5. MD 2008 1015 1923 930 1883 890 1542 549 1387 394 1229 236 1221 228 1153 160 1127.2 1.7 138.7 51.6

228 No. Pt TB SB CSA Kd Protein EGFR

6. CM 1459 1039 1171 751 926 506 862 442 634 414 705 285 625 205 581 161 1127.2 1.3 87.5 112.1

7. ET 2345 1158 1573 386 1570 383 1479 292 1335 148 1274 87 1241 54 1227.2 1.2 101.0 42.5

8. HT 1017 597 873 453 863 443 744 324 733 313 712 292 610 190 632.8 2.3 57.0 142.1

9. JAs 1488 862 1329 703 1261 635 1193 567 1098 472 985 359 919 293 792 166 662.7 2.5 97.5 109.2

229 TURP Specimens Scatchard Analysis (SA)

Calculation using in-house program on Amstrad p.c. (Patient MD Table 5 (ii)).

138.75ug/tube EGF Receptor assay Label date 5/15 Scatchard Assay date 6/16 Enter Sp.Act.: 84 uCi/ug Sample:Patient M .D. Sp.Act.; 1127.28 dpm/fmol Post-lab 32 days No.Points 8Corr.Sp.A 779.0144 dpm/fmol Corr.: — .160486

sample Total Bound B.Sp B.Sp Corr.BSp Free B/F no. (nM) (dpm) (dpm) (nM) (nM) (nM)

1 .3 2008 1015 .0032573 .0032573 .2967427 .0032573 2 .453 1923 930 .0029845 .0045067 .4484933 .0100484 3 .612 1883 890 .0028562 .0058266 .6061734 .0096121 4 .925 1542 549 .0017618 .0054323 .9195677 .0059075 5 1.55 1387 394 .0012644 .0065328 1.543467 .0042326 6 2.8 1229 236 .0007574 .0070688 2.792931 .0025309 7 5.3 1221 228 .0007317 .0129266 5.287073 .0024449 8 10.3 1153 160 .0005135 .0176291 10.28237 .0017145

12 100.3 993 .0179067 0 r= -.762337

Kd= -1.68689 nM Sites= 51.62283 fmol/mg protein

230 Appendix III Steroid receptor results

TURP Specimens ER Assay (Table 5 vi)

AF = Aliquot Factor

Aliquot Factor = Assay vol x Charcoal vol / Count fraction SAF = Specific Activity Factor (dpm/mol) Protein = mg/ml

Steroid Receptor Binding = SB x AF x SAF / Protein (fmol/mg protein)

Wo. Pt TB NSB SB SAF Prot.

1. CM 538 403 135 1.75 4.74 1.10 10.1 2. MD 972 318 654 1.75 4.74 0.95 57.1 3. AF 992 349 643 1.75 4.74 1.28 41.6 4. BK 877 314 563 1.75 4.74 1.35 34.6 5. FW 523 489 34 1.75 4.74 0.92 3.1 6. EL 569 390 179 1.75 4.74 0.89 16.6 7. RG 611 427 184 1.75 4.74 1.26 12.1 8. ET 895 322 573 1.75 4.74 0.90 52.8 9. JAs 439 519 - 1.75 4.74 1.31 0 10.HR 474 692 - 1.75 4.74 1.22 0 11. PB 589 534 55 1.75 4.74 0.98 4.6 12.JAt 878 532 346 1.75 4.74 0.77 37.3 13.JW 754 366 418 1.75 4.74 1.03 33.6 14. WW 650 401 249 1.75 4.74 0.98 21.1 15.FD 761 700 61 1.75 4.74 1.35 3.7 16. CG 853 719 134 1.75 4.74 1.09 10.2 17. HG 794 513 281 1.75 4.74 0.69 33.8 18.HT 636 482 154 1.75 4.74 0.84 15.2 19.MA 434 422 - 1.75 4.74 1.05 0

231 "Normal" prostate ER Assay (Table 5 vii )

No. Pt TB NSB SB SAF Prot. ro

1. WS 946 541 354 1.75 4.7 1.60 23.3 2. MD 621 421 200 1.75 4.7 1.06 15.6 3. ES 905 397 253 1.75 4.7 1.27 33.1 4. JV 884 399 485 1.75 4.7 1.04 38.6 5. RW 737 363 374 1.75 4.7 0.93 33.4 6. PD 533 369 164 1.75 4.7 1.65 8.2 7. NS 976 402 574 1.75 4.7 2.22 21.4 8. VG 790 583 207 1.75 4.7 1.86 9.2 9. CS 677 325 352 1.75 4.7 1.14 28.8

TURP Specimens PgR Assay (Table viii.)

No. Pt TB NSB SB SAF Prot. PgR

1. CM 3044 2798 245 1.75 5.29 1.10 20.6 2. MD 2702 2559 143 1.75 5.29 0.95 13.9 3. AF 1789 1630 159 1.75 5.29 1.28 11.5 4. BK 1819 1577 242 1.75 5.29 1.35 16.6 5. FW 3041 3401 - 1.75 5.29 0.92 0 6. EL 1791 1685 106 1.75 5.29 0.89 11.0 7. RG 2141 2984 - 1.75 5.29 1.26 0 8. ET 2559 2278 279 1.75 5.29 0.90 28.7 9. JAs 1940 1992 - 1.75 5.29 1.31 0 10.HR 1788 1754 - 1.75 5.29 1.22 0 11. PB 1975 2003 - 1.75 5.29 0.98 0 12.JAt 1851 1721 130 1.75 5.29 0.77 15.5 13.JW 2005 1994 - 1.75 5.29 1.03 0 14. WW 3044 2921 122 1.75 5.29 0.98 11.5 15.FD 2721 3556 - 1.75 5.29 1.35 0 16.CG 2402 2492 - 1.75 5.29 1.09 0 17. HG 1780 1603 177 1.75 5.29 0.69 23.7 18.HT 1959 1628 332 1.75 5.29 0.84 36.5 19.MA 1055 2712 - 1.75 5.29 1.05 0

232 "Normal" Prostate PgR Assay (Table 5 ix. )

No. Pt TB NSB SB SAF Prot. PgR

1. WS 2651 2493 158 1.75 5.29 1.60 9.2 2. MD 2852 2617 235 1.75 5.29 1.06 20.5 3. ES 2239 2392 - 1.75 5.29 1.27 0 4. JV 1746 1590 155 1.75 5.29 1.04 13.8 5. RW 1895 1650 245 1.75 5.29 0.93 24.3 6. PD 1865 1795 70 1.75 5.29 1.65 3.9 7. NS 1128 1098 30 1.75 5.29 2.22 1.2 8. VG 1422 1900 - 1.75 5.29 1.86 0 9. CS 1903 1975 - 1.75 5.29 1.14 0

Appendix IV Regional distribution

Regional distribution of EGFR and Steroid Receptors EGFR Inner Zone (Table 5 xi.)

Patient ra NSB SB CSA Protein EGFR

1. HT 3276 1892 1384 410 .052 64.9 2. HG 4243 1823 2420 1627 .037 40.2 3. CG 5967 2463 3504 403 .068 127.8 4. FD 5224 3160 2064 403 .048 106.6 5. WW 2035 893 1142 604 .044 42.9 6. JW 11117 10568 549 1627 .046 7.3 7. JAs 12587 8084 4503 1646 .044 62.1 8. PB 17008 10407 6601 1627 .038 106.7 9. SG 12853 6792 6061 1608 .043 87.6 10. HS 15177 10957 4220 1608 .039 67.2 11.HR 25932 19487 6445 1785 .034 105.8 12.JAt 11986 6506 5480 1627 .029 116.0 13.ET 12927 9237 3690 1609 .034 67.5 14.RG 12269 8589 3680 1609 .041 55.8 15.MG 2354 1670 684 400 .026 65.8 16.EL 9122 5651 3471 1646 .040 52.7 17.FW 17320 7555 9765 1609 .053 114.6 18.BK 2130 1691 439 374 .052 22.5 19.DM 6386 3863 2523 400 .050 126.2 20. JS 18432 8859 9573 1536 .068 91.6 21.RW 15124 7061. 8063 1536 .044 119.3 22. PK 6446 6216 230 1536 .061 2.5 23.ELa 15144 5858 9286 1500 .053 116.8 24. AF 19208 18571 637 1785 .054 6.6 25. FS 25955 16463 9492 1785 .045 118.2 26.MD 1589 882 702 429 .039 42.3 27.CM 4639 2484 2155 658 .056 58.5

233 EGFR Outer Zone (Table 5 xi.)

Patient TB NSB SB CSA Protein EGFR

1. HT 1998 2016 410 .057 0 2. HG 2078 2078 - 1627 .037 0 3. CG 3320 1932 1388 403 .064 53.8 4. FD 3818 3053 765 403 .039 48.6

5. WW -- . - --- 6. JW 9842 8894 948 1027 .059 9.9 7. JAs 12610 10060 2550 1646 .066 23.5 8. PB 11925 9879 2046 1627 .055 22.8 9. SG 10714 9388 1326 1608 .048 17.2 10. HS 10584 10103 481 1608 .044 6.8 11.HR 22404 18631 3773 1785 .039 54.2 12.JAt 8452 6855 1597 1627 .028 35.0 13.ET 10029 9076 953 1608 .059 16.4 14.RG 9019 8543 476 1608 .038 7.8 15.MG 1387 1386 - 400 .020 0 16.EL 5805 5553 252 1646 .036 4.3 17.FW 7599 7586 13 1608 .043 0 18. BK 2072 1758 314 374 .058 14.4 19.DM 2596 1666 930 400 .079 29.4 20. JS 12107 9171 2936 1536 .059 32.3 21.RW 8191 7868 353 1536 .056 3.8 22.PK 7074 6317 757 1536 .062 7.8 23.EL 10765 8808 1957 1500 .068 19.2 24. AF 18038 17516 522 1785 .036 8.1 25. FS 21418 18645 2773 1785 .049 31.7 26.MD 730 659 71 429 .022 7.5 27.CM 2765 2192 573 658 .032 27.2

234 ER, Inner zone (Table 5 xil. )

Patient TB NSB SB SAF Protein ER

1. HT 675.5 418.4 257.1 4.7 79 26.9 2. HG 851.5 466.4 385.1 4.7 101 31.6 3. CG 721.1 651.0 70.1 4.7 95 6.1 4. FD 429.2 582.6 - 4.7 102 0 5. WW 538.0 403.5 134.5 4.7 110 10.1 6. JW 829.7 490.6 339.1 4.7 79 35.6 7. JAs 521.6 485.4 36.2 4.7 62 4.8 8. PB 812.2 788.6 23.6 4.7 142 1.4 9. SG 556.0 451.9 104.1 4.7 89 9.7 10. HS 725.6 712.4 13.2 4.7 69 1.6 11.HR 629.7 403.1 226.6 4.7 91 20.6 12.JAt 785.6 719.7 65.9 4.7 135 4.0 13.MD 689.0 392.1 296.9 4.7 98 25.1 14. SE 818.5 701.3 117.2 4.7 105 9.3 15. FC 431.2 299.7 131.5 4.7 146 7.5 16.WD 625.0 291.3 333.7 4.7 95 29.1 17. FS 755.9 918.2 - 4.7 162 0 18. PK 771.4 500.9 270.5 4.7 85 26.4 19.RW 259.7 230.0 - 4.7 116 0 20.HA 523.3 407.2 116.1 4.7 129 7.5 21. JS 598.6 392.7 205.9 4.7 111 15.3 22. RG 711.5 597.2 114.3 4.7 95 10.0

ER Outer Zone (Table 5 xil.)

Patient TB NSB SB SAF Protein m

1. HT 526.7 478.1 148.6 4.7 98 4.1 2. HG 632.0 313.9 318.1 4.7 90 29.3 3. CG 569.2 451.0 145.2 4.7 88 13.6 4. FD 240.0 296.4 - 4.7 109 0 5. WW 528.3 356.9 171.4 4.7 113 12.5 6. JW 590.4 468.9 121.5 4.7 98 10.2 7. JAs 570.8 478.7 92.1 4.7 96 7.9 8. PB 860.9 947.7 - 4.7 161 0 9. SG 513.9 543.8 - 4.7 97 0 10. HS 537.0 356.5 181.0 4.7 109 13.7 11.HR 512.4 455.3 57.1 4.7 121 3.9 12.JAt 232.9 298.2 - 4.7 149 0 13.MD 517.8 319.4 198.4 4.7 100 16.5 14. SE 954.6 839.2 115.4 4.7 117 8.2 15. FC 456.2 381.7 74.5 4.7 123 5.0 16.WD 789.1 432.6 356.5 4.7 95 31.1 17. FS 677.9 659.2 - 4.7 114 0 18.PK 532.2 366.9 165.3 4.7 114 12.0 19.RW 533.3 491.2 42.1 4.7 97 3.6 20.MA 649.2 601.0 48.2 4.7 116 3.4 21.JS 717.2 584.3 132.9 4.7 92 12.0 22.RG 233.5 314.2 - 4.7 99 0

235 PgR Inner Zone (Table 5 xiii.)

Patient TB NSB SB SAF Protein PgR

1. HT 1862.1 1592.7 269.4 5.3 79 31.5 2. HG 2951.2 2802.3 148.9 5.3 101 13.6 3. CG 1650.1 1529.2 120.9 5.3 95 11.7 4. FD 1544.2 1699.8 - 5.3 102 0 5. WW 1929.7 1878.3 51.4 5.3 110 4.3 6. JW 3071.2 2983.4 87.8 5.3 79 10.2 7. JAa 1827.1 1950.8 - 5.3 62 0 8. PB 959.4 1728.0 - 5.3 142 0 9. SG 1955.8 1901.2 54.6 5.3 89 5.7 10. HS 2928.3 2809.0 119.3 5.3 69 16.0 11.HR 1604.2 1492.2 112.0 5.3 91 11.3 12.JAt 1173.7 1955.3 - 5.3 135 0 13.MD 2604.5 2469.2 135.3 5.3 98 12.8 14. SE 1009.9 1057.3 - 5.3 105 0 15. FC 1789.4 1564.2 225.2 5.3 146 14.2 16. WD 1899.5 1620.3 279.2 5.3 95 27.2 17. FS 2223.0 2142.4 81.6 5.3 162 4.7 18. PK 1464.7 1345.9 118.8 5.3 85 12.9 19.RW 1135.2 1098.6 36.6 5.3 116 2.9 20.HA 1721.6 2916.8 - 5.3 129 0 21. JS 1984.2 1999.0 - 5.3 111 0 22.RG 1209.8 1182.6 - 5.3 95 2.6

PgR Outer Zone

Patient TB NSB SB SAF Protein PgR

1. HT 1753.1 1944.8 5.3 98 0 2. HG 1859.2 1722.2 137.0 5.3 90 14.1 3. CG 1421.5 1384.7 36.8 5.3 88 3.9 4 FD 1971.2 3596.7 - 5.3 109 0 5. WW 1092.4 1083.9 - 5.3 113 0 6. JW 1625.3 1567.6 57.7 5.3 98 5.4 7. JAs 1172.6 1223.4 - 5.3 96 0 8. PB 1094.6 1952.1 - 5.3 161 0 9. SG 1455.0 1449.2 - 5.3 97 0 10. HS 1628.7 1759.6 - 5.3 109 0 11.HR 1492.8 1411.8 81.0 5.3 121 6.2 12.JAt 1001.7 992.7 17.0 5.3 ' 149 1.1 13.MD 1406.7 1291.9 114.8 5.3 100 10.6 14. SE 1595.1 1427.0 168.1 5.3 117 13.3 15. FC 1679.5 1699.2 - 5.3 123 0 16. WD 1955.0 1832.6 122.4 5.3 95 11.9 17. FS 1211.7 1229.0 - 5.3 114 0 18.PK 1692.3 1541.4 150.9 5.3 114 12.3 19.RW 1559.8 1492.8 67.0 5.3 97 6.4 20.MA 1095.3 1052.6 42.7 5.3 116 3.4 21. JS 1289.7 1488.6 - 5.3 92 0 22.RG 1168.8 1099.4 69.4 5.3 99 6.5

236 Appendix V Steroid Hormones

Controls (Table 5 X V . )

Wo. Testosterone DHT Oestradiol SHBG

1. 33.0 0.41 137 25 2. 30.8 0.27 107 33 3. 24.1 0.74 62 32 4. 10.9 0.62 53 16 5. 18.3 0.21 100 16 6. 22.4 0.87 102 14 7. 22.9 0.18 31 33 8. 21.1 0.90 64 33 9. 32.0 0.23 116 17 10. 25.1 0.14 111 10 11. 31.3 0.09 112 23 12. 18.9 0.43 52 18 13. 15.1 0.55 68 15 14. 22.4 0.83 40 25 15. 22.0 0.44 51 25

237 Bibliography

Aherne W, Dunnill M S 1966a Quantitative aspects of pla cental structure. Journal of Pathology and Bacteriology

91: 123-139

Aherne W, Dunnill M S 1966b Morphometry of the human pla centa. British Medical Bulletin 22: 5-9

Aherne W 1967 Methods of counting discrete tissue compo nents in microscopical sections. Journal of the Royal Microscopical Society 87: 493-508

Aherne W 1968 A method of determining the cross sectional areas of muscle fibres. Journal of Neurological Science 7 519-528

Aherne W 1970 Quantitative methods in histology. Journal of Medical Laboratory Technology 27: 168-170

Aherne W 1975 Some morphometric methods for the central nervous system. Journal of Neurological Science 24: 221-

241

Aherne W, Dunnill MS 1982 Morphometry. London, Edward

Arnold.

238 Albarran J, Motz B 1902 Contribution a 1'etude de 1'anato mie macroscopique de la prostate hypertrophiée. Annals de

Maladie d'Organes Genitourin. (Paris) 20: 769-817

Anderson K M, Liao S 1968 Selective retention of dihydro

testosterone by prostatic nuclei. Nature 219: 277-284

Aragona C, Freisen H 1975 Specific prolactin binding sites

in the prostate of rats. Endocrinology 97: 677-684

Armenian H K, Lilienfeld A M, Diamond E L, Bross I D J

1974 Relation between benign prostatic hyperplasia and

cancer of the prostate. A prospective and retrospective

study. Lancet 2: 115-117.

Ashley J B 1966 Observations on the epidemiology of pro

static hyperplasia in Wales. Br.J Urol 38: 567-569

Asselin J, Labrie F, Gourdeau Y 1976 Binding of [^] me-

thyltrienolone (R1881) in rat prostate and human BPH.

Steroids 28: 449-459

Atkison P R, Weidman E R, Bhaumick B, Bala R M 1980 Re

lease of somatomedin-like activity by cultured WI-38 human

fibroblasts. Endocrinology 106: 2006-2012

Auf G, Ghanadian R 1982 Characterisation and measurement of

cytoplasmic and nuclear oestradiol 17-beta receptor pro

teins in human BPH. Journal of Endocrinology 93: 305-317

239 Bahr G F, Bloom G, Friberg U 1957 Volume changes of tissues in physiological fluids during fixation in osmium tetrox- ide or formaldehyde and during subsequent treatment.

Experimental Cellular Research 12: 342-349

Barker S, Puddefoot J R, Panahy C, Goode A W, Vinson G P 1989 Epidermal growth factor receptors and oestrogen receptors in the cancerous breast. Journal of Endocrino logy (suppl) 121: 19

Barrack E R, Coffey D S 1980 The specific binding of estrogens and androgens to the nuclear matrix of sex hor mone responsive tissues. Journal of Biological Chemistry 255: 7265-7275

Barrack E R, Bruchovsky P, Walsh P C 1983 Subcellular dis tribution of androgen receptors in human normal, benign hyperplastic and malignant prostatic tissue: characterisa tion of nuclear-salt resistant receptors. Cancer Research

43: 1107-1116

Bartsch G, Fischer E, Rohr H P 1975 Ultrastructural mor phometric analysis of the rat prostate (ventral lobe). Urological Research 3: 1-11

240 Bartsch G, Frick 3, Rohr H P 1976 Stereology, a new mor phological method of prostatic function and disease. In Marberger H, Haschek H, Schirmer H K A, Colston J A C and

Witkin E (eds) Progress in Clinical and Biological Re search. New York, Alan R. Liss 123-141.

Bartsch G, Rohr H P 1977 Ultrastuctural stereology. A new approach to the study of prostatic function. Investigative Urology 14: 301-306

Bartsch G, Muller H R, Oberholzer M, Rohr H P 1979 Light microscopic stereological analysis of the normal human prostate in benign prostatic hyperplasia. Journal of Urology 122: 487-491

Bartsch W, Becker H, Pinkenburg F A 1979 Hormone blood levels and there interrelationships in normal men and men with benign prostatic hyperplasia (BPH). Acta Endocrinolo- gica 90: 727-736

Bartsch W, Klein H, Sturenburg H J 1987 Metabolism of androgens in human benign prostatic hyperplasia: aromatase and its inhibition. Journal of Steroid Biochemistry 27: 557-564

241 Bashirelahi N, Beckerman T, Young J D 1983 Distribution of promegestone in benign prostatic hypertrophy in regard to age, race and histological differences. Journal of Steroid Biochemistry 19: 1757-1761

Baulieu E E 1978 Cell membrane, a target for steroid hormones. Molecular and Cellular Endocrinology 12: 247-254

Beatson G T 1896 On the treatment of inoperable cases of carcinoma of the mamma: suggestions for a new method of treatment with illustrative cases. Lancet 2: 104-107

Begun F P, Story M T, Jacobs S C, Lawson R K 1989 Distri bution of basic fibroblast growth factor within the prostate. Journal of Urology (suppl) 141: 549

Belis J A, Adelstein L B, Terry W F 1983 Influence of estradiol on accessory reproductive organs in the castra ted male rat. Journal of Andrology 4: 144-149

Belis J A, Lizza E F, Tarry W F 1984 Progesterone recep tors in the prostate. In: Kimball F A, Buhl A E, Carter D B (eds) New approaches to the study of benign prostatic hyperplasia. Progress in Clinical and Biological Research 145: 345-361.

Bell R M 1986 Protein kinase C activation by diacylglycer- ol second messengers. Cell 45: 631-634

242 Berenblum I 1978 Established principles and unresolved problems in carcinogenesis. Journal of the National Cancer Institute 60: 723-726

Berry R J and Jones D R 1982 Prostatic hypertrophy and carcinoma. In: Miller D L and Farmer R D T (eds) Epidemio logy of Diseases. Blackwell Scientific Publications, London. 261-265.

Berry S J, Coffey D S, Walsh P C, Ewing L R 1984 The development of human benign prostatic hyperplasia with age. Journal of Urology 132: 474-479

Blacklock N J, Bouskill K 1977 The zonal anatomy of the prostate in man and in the rhesus monkey. Urological Research 5: 163-167

Blacklock N J 1982 Surgical anatomy of the prostate. In: Chisholm G D, Williams D I (eds) Scientific Foundations of Urology, Heinemann, London, 113-125.

Boni-Schnetzler M, Pilch P F 1987 Mechanism of epidermal growth factor receptor autophosphorylation and high-affin ity binding. Proceedings of the National Academy of Sci ence U.S.A. 84: 7832-7836

243 Bonne C, Raynaud J P 1975 Methyltrienolone, a specific ligand for cellular androgen receptors. Steroids 26: 227-

232

Bonne C, Raynaud J P 1976 Assay of androgen binding sites by exchange with methyltrienolone R1881. Steroids 27: 497-

507

Borthwick N M, Smellie R M S 1975 The effects of oestradiol 17-beta on the ribonucleic acid polymerases of immature rabbit uterus. Biochemical Journal 147: 91-101

Bourke J B and Griffin J P 1966 Hypertension, diabetes mellitus and blood groups in benign prostatic hypertrophy. British Journal of Urology 38: 18-23

Bourke J B, Griffin J P 1968 Diabetes mellitus in patients with benign prostatic hyperplasia. British Medical Journal 4: 492-493

Boyle D M, Van der Valt L A 1988 Enhanced phosphorylation of progesterone receptor by protein kinase C in human breast cancer cells. Journal of Steroid Biochemistry 30:

239-244

244 Boyns A R, Phillips M E A, Hiller S G, Cameron E H D,

Griffiths K, Shahmanesh M, Feneley R C L, Hartog M 1974

Plasma prolactin, GH, LH, FSH, TSH and testosterone during treatment of prostatic carcinoma with oestrogens. European Journal of Cancer 10: 445-449

Bradbury S 1983 Commercial image analysers and the charac terisation of microscopical images. Journal of Microscopy 131: 203-210

Bremner W J, Vitiello M V, Prinz P N 1983 Loss of circadian rhythm in blood testosterone levels with aging in normal men. Journal of Clinical Endocrinology and Metabolism 53: 1278-1281

Brendler C B, Isaacs J T, Follansbee A L, Walsh P C 1984 The use of multiple variables to predict response to endocrine therapy in carcinoma of the prostate. Journal of Urology 131: 694-700

British Prostate Group Study 1979 Evaluation of plasma hormone concentrations in relation to clinical staging in patients with prostatic cancer. British Journal of Urology 51: 382-389

245 Brochu M, Belanger A 1987 Comparative study of plasma

steroid and steroid glucoronide levels in normal men and men with benign prostatic hyperplasia. The Prostate 11:

33-40.

Brody J A 1984 Facts, projections, and gaps concerning data

on aging. Public Health Reports 99: 468-475

Brody J A 1985 Prospects for an ageing population. Nature

315: 463-466

Brown T R, Rothwell S W , Migeon C J 1981 Comparison of

methyltrienolone and dihydrotestosterone binding and

metabolism in human genital skin fibroblasts. Journal of

Steroid Biochemistry 14: 1013-1022

Bruchovsky N, Wilson J D 1968a The conversion of

testosterone to 5 alpha-androstan-17 beta-ol-3-one by rat

prostate in vivo and vitro. Journal of Biological Chemi

stry. 243: 2012-2021

Bruchovsky N, Wilson J D 1968b The intranuclear binding

of testosterone and 5 alpha-androstan-17 beta-Ol-3-one by

rat prostate. Journal of Biological Chemistry 243: 5953-

5960

246 Bruchovsky N and Lieskovsky G 1979 Increased ratio of 5 alpha-reductase: 3 alpha (beta)-hydroxysteroid dehydroge nase activities in the hyperplastic human prostate. Jour nal of Endocrinology 80: 289-301

Buchi K, Villee C A 1976 Influence of heating on the acti vation and transformation of the estradiol receptor of the rat uterus. Journal of Steroid Biochemistry 7: 539-544

Cabot A T 1896 The question of castration for enlarged prostate. Annals of Surgery 24: 265-309

Carpenter G, Cohen S 1979 Epidermal Growth Factor. Annual Review of Biochemistry 48: 198-216

Carpenter G 1987 Receptors for epidermal growth factor and other polypeptide mitogens. Annual Review of Biochemistry 56: 881-914

Catalona W J and Scott W W 1978 Carcinoma of the prostate: a review. Journal of Urology 119: 1-8

Chamberlain 3, Jagarinec N, Ofner P 1966 Catabolism of [4- 14C] testosterone by subcellular fractions of human prostate. Biochemical Journal 99: 610-616

247 Chaproniere D M, McKeehan D M, McKeehan W L 1986 Serial culture of single adult human prostatic epithelial cells in serum-free medium containing low calcium and a new growth factor from bovine brain. Cancer Research 46: 819-

824

Chung L W K, Cunha G R 1983 Stromal-epithelial interac tions: II. Regulation of prostatic growth by embryonic urogenital sinus mesenchyme. Prostate 4: 503-507

Chung L W K, Matsuura J, Rocco A K, Thompson T C, Miller G J, Runner M N 1984 A new mouse model for prostatic hyper plasia: Induction of adult prostatic overgrowth of fetal urogenital sinus implants. In: Kimball F A, Buhl A E, Carter D B (eds) New Approaches to the Study of Benign Prostatic Hyperplasia. New York, Alan R. Liss.

Chung L W K, Chang S M, Bell C, Zhau H, Ro J Y, von

Eschenbach A C 1988 Prostatic carcinogenesis evoked by cellular interaction. Environmental Health Perspectives

77: 23-28

Clark J H, Peck E J 1976 Nuclear retention of receptor- oestrogen complex and nuclear acceptor sites. Nature 260:

635-637

248 Clemmons D R, Underwood L E, Van Wyk J J 1981 Hormonal control of immunoreactive somatomedin production by cul tured human fibroblasts. Journal of Clinical Investigation

67: 10-19

Cline M J, Buttitora H, Yokota J 1987 Protooncogene abnor malities in human breast cancer: correlations with anato mic features and clinical course of disease. Journal of Clinical Oncology 5: 999-1006

Cochet C, Gill G N, Meisenhelder J, Cooper J A, Hunter T 1984 C-kinase phosphorylates the epidermal growth factor receptor and reduces its epidermal growth factor- stimulated tyrosine protein kinase activity. Journal of Biological Chemistry 259: 2553-2556

Coffey D S 1988 Endocrine control of normal and abnormal growth of the prostate. In: Raifer J (ed) Urologie Endo crinology, Saunders 170-195.

Cohen S 1962 Isolation of a mouse submaxillary gland protein accelerating incisor eruption and eyelid opening

in the newborn animal. Journal of Biological Chemistry 237: 1555-1562

Cohen S, Taylor J M 1974 Recent Progress in Hormone

Research 30: 551-574

249 Cole M 1966 The metals research quantimet (QTM). Micro

scope 15; 148-169

Cooper A P 1836 Principles and practice of surgery. E.Cox

London 333-335.

Cowan R A, Cook B, Cowan S K 1979 Testosterone 5 alpha- reductase and the accumulation of dihydrotestosterone in benign prostatic hyperplasia. Journal of Steroid Biochemi stry 11: 609-613

Cowan R A, Cowan S K, Grant J K 1975 The specificity of 5 dihydrotestosterone binding in human prostatic cytosol preparations. Biochemical Society Transcripts 3: 537-540

Cowan R A, Cowan S K, Grant J K 1977 Binding of methyl-

trienolone (R 1881) to a progesterone receptor-like compo nent of human prostatic cytosol. Journal of Endocrinology 74: 281-289

Cowan W M, Wann D F 1973 A computer system for the mea

surement of cell and nuclear sizes. Journal of Microscopy

99: 331-348

Crabb J W, Armes L G, Carr S A, Johnson C M, Roberts G D,

Bardoli R S, McKeehan W L 1986 Complete primary structure

of prostatropin: a prostatic epithelial cell growth fac

tor. Biochemistry 25: 4988-4993

250 Cunha G R 1972 Tissue interactions between epithelium and mesenchyme of urogenital and integumental origin. Anatomi cal Record 172: 529-541

Cunha G R 1976 Epithelial-stromal interactions in develop ment of the urogenital tract. International Review of Cytology 47: 137-194

Cunha G R, Lung B 1979 The importance of stroma in morpho genesis and functional activity of urogenital epithelium. In Vitro 15: 50-71

Cunha G R, Chung L W K, Shannon J M, Reese B A 1980 Stro mal-epithelial interactions in sex differentiation. Biolo gical Reproduction 22: 19-42

Cunha G R, Chung L W K, Shannon J M, Toguchi O, Fujii H 1983 Hormone induced morphogenesis and growth: role of mesenchymal-epithelial interactions. Recent Progress in Hormone Research 39: 559-598

Dahnert W F, Hamper U M, Eggleston J C, Walsh P C, Sanders

R C 1968 Prostatic evaluation by transrectal ultrasonogra phy with histopathological correlation: the echopenic appearances of early carcinoma. Radiology 158: 97-102

251 Davies P, Eaton C L 1989 Binding of epidermal growth factor by human normal, hypertrophic and carcinomatous prostate. Prostate 14: 123-132

DeKlerk D P, Heston W D W, Coffey D S 1975 Studies on the role of macromolecular synthesis in the growth of the prostate. In: Benign Prostatic Hyperplasia, NIAMDD Work shop Proceedings, DHEW Publication, U.S. Govt. Printing Office, Washington D.C. 43-53.

DeKlerk D P, Coffey D S 1978 Quantitative determination of prostatic epithelial and stromal hyperplasia by a new technique. Biomorphometrics. Investigative Urology 16: 240-245

DeKlerk DP, Coffey D S, Ewing L L, McDermott I R, Reiner W G, Robinson C H, Scott W W, Talalay P, Walsh P C, Whea ton L G, Zirkin B R 1979 Comparison of spontaneously and experimentally induced canine prostatic hyperplasia. Journal of Clinical Investigation 64: 842-849

Delesse M A 1847 Procédé mechanique pour determiner la composition des roches. C.R.Académie Science (Paris) 25: 544-545

Deming C L, Neumann C 1939 Early phases of prostatic

hyperplasia. Surgery Gynecology Obstetrics 68: 155-160

252 Demurs T, Kuzumaki N, Oda A, Fujita H, Ishibashi T, Koyan- gi T 1989 Establishment and characterization of monoclonal antibody against androgen receptor. Journal of Steroid Biochemistry 33: 845-851

Derbes V De P, Leche S M, Hooker C W 1937 The incidence of benign prostatic hypertrophy among the whites and negroes in New Orleans. Journal of Urology 38: 383-388

De Voogt D J, Dingjan P 1978 Steroid receptors in human prostatic cancer: a preliminary evaluation. Urological

Research 6: 151-158

Deyl Z, Rosmus J, Adam M 1976 Pituitary and collagen. In: Everitt A V, Burgess J A (eds) Hypothalamus, Pituitary and Aging. Illinois, Charles C. Thomas, 171-192.

Diamond D A, Berry S J, Umbricht C, Jewett H J, Coffey D S 1982 Computerised image analysis of nuclear shape as a prognostic factor for prostate cancer. The Prostate 3:

321-322

Dickson R B, Rosen N, Gelman E P, Lippman M E 1987 Receptor and signal transduction-related proto-oncogenes in breast cancer. Trends in Pharmacological Science 8: 372-375

253 Dix C T, Jordan V C 1980 Modulation of rat uterine steroid hormone receptors by oestrogen and antioestrogen. Endocri nology 107: 2011-2020

Doisy E A, Veler C D, Thayer S 1929 Folliculin from urine of pregnant women. American Journal Physiology 90: 329-330

Downward J, Yarden Y, Mayes E, Scrage G, Totty N, Stock- well P, Ullrich A, Schlessinger J, Waterfield M D, 1984

Close similarity of epidermal growth factor receptor and v-erb-B oncogene protein sequences. Nature 3 07: 521-527

Drafta D, Proca E, Zamifir V 1982 Plasma steroids in benign prostatic hypertrophy and carcinoma of the prostate. Journal Steroid Biochemistry 17: 689-693

Dube J Y, Chapdelaine P, Tremblay P R 197 6 Comparative binding specificity of methyltrienolone in human and rat prostate. Hormone Research 7: 341-347

Eaton C L, Davies P, Phillips M E A 1988 Growth factor involvement and oncogene expression in prostatic tumours.

Journal of Steroid Biochemistry 30: 341-345

Edwards D P, Murthy S R, McGuire W L 1980 Effects of estro gen and antiestrogen on DNA polymerase in human breast cancer. Cancer Research 40: 1722-1726

254 Egender G, Rapf C, Feichtinger I, Mikuz G, Bartsch G,

Frommhold H 1984 Vergleichende histopathologische und sonomorphologische Prostatauntersuchungen. ROFO: For- tschritte auf dem Giebiete der Rontgenstrahlen und der Nuclearmedizin (Stuttgart) 140: 60-66

Ekman P, Snochowski M, Dahlberg E 1979a Steroid receptors in metastatic carcinoma of the prostate. European Journal of Cancer 15: 257-262

Ekman P, Snochowski M, Dahlberg E, Bression D, Hogberg B, Gustoffson J A 1979b Steroid receptor content in cytosol from normal and hyperplastic human prostates. Journal of Clinical Endocrinology and Metabolism 49: 205-215

Ekman P, Barrack E R, Walsh P C 1982 Simultaneous measure ment of progesterone and androgen receptors in human prostate: A microassay. Journal of Clinical Endocrinology and Metabolism 55: 1089-1099

El Etreby M P, Mahrous A T 1979 Immunocytochemical techni que for the detection of prolactin (PRL) and growth hor mone (GH) in hyperplastic and neoplastic lesions of dog prostate and mammary gland. Histochemistry 64: 279-286

255 Elson S D, Browne C A, Thorburn G D 1984 Identification of epidermal growth factor-like activity in human male repro ductive tissues and fluids. Journal of Clinical Endocrino logy and Metabolism 58: 589-594

Emtage L A, Dunn P J S, Rowse A D 1989 Androgen and oe strogen receptor status in benign and neoplastic prostate disease. British Journal of Urology 63: 627-633

Ewing L L, Berry S J, Higginbottom E G 1983 Dihydro testosterone content of beagle prostatic tissue: Effect of age and hyperplasia. Endocrinology 113 2004-2009

Exner H E 1981 Progress in stereology - Methods, instru mentation and international co-operation. In: Contemporary Stereology. Proceedings 3rd European Symposium for Stereo

logy.

Farnsworth W E, Brown J R 1963 Metabolism of testosterone by the human prostate. Journal of the American Medical Association 183: 436-439

Farnsworth W E 1988 Prolactin effect on the permeability

of human benign hyperplastic prostate to testosterone. The Prostate 12: 221-225

256 Fattah D I, Chard T 1981 Simplified method for measuring sex hormone binding globulin. Clinical Chemistry 27: 1277-

1279

Ferguson J D, Gibson E C 1956 Prostatic smear diagnosis.

British Medical Journal 1: 822-825

Fernandez-Pol J A 1981 Epidermal growth factor: relation ship between receptor down-regulation in cultured NRK cells and epidermal growth factor enhancement of phosphor ylation of 170,000 molecular weight membrane protein in vitro. Biochemistry 20: 3907-3912

Fernandez-Pol J A 1982 Modulation of a growth factor dependent protein phosphorylation in cell membrane pre paration by receptor down regulation. Journal of Cellular Biochemistry 19: 205-222

Fisher C 1967 An image analysing computer. Bio-medical Engineering 351-357

Fisher C 1971 Analysing images by computer. New Scientist

23: 676-678

Fowler J E, Lau J L, Ghosh L, Mills S E, Mounzer A 1988

Epidermal growth factor and prostatic carcinoma: an immunohistochemical study. Journal of Urology 139: 857-861

257 Fraker P J, Speck J C 1978 Protein and cell membrane iodinations with a sparingly soluble chloroamide 1,3,4,6- tetrachlor-3a,6a-diphrenylglycoruril. Biochemical and

Biophysical Research Communications 80: 849-857

Franks L M, Barton A A 1960 The effects of testosterone on the ultrastructure of the mouse prostate in vivo and in organ cultures. Experimental Cellular Research 19: 35-50

Franks L M 1954 Benign nodular hyperplasia of the prostate: a review. Annals of The Royal College of Sur geons 14: 92-106

Franks L M, Riddle P N, Carbonell A W, Gey G 0 1970 A comparative study of the ultrastructure and lack of growth capacity of the adult human prostate epithelium mechanic ally separated from its stroma. Journal of Pathology 100: 113-119

Franks L M 1975 Benign prostatic hyperplasia; gross and microscopical anatomy. In: Grayhack J, Wilson J, Scher- benske M S (eds) Benign Prostatic Hyperplasia, Bethesda, DHEW publication (NIH), 76-1113.

Furth J 1975 Hormones as etiological agents in neoplasia.

In: Becker F F (ed) Cancer: A comprehensive treatise. Vol

1 Plenum Press, New York and London, 89-134.

258 Garraway W M, Collins G N, Lee R J 1991 High prevalence of benign prostatic hypertrophy in the community. Lancet 338: 469-471

Geller J, Albert J, Lopez D, Geller A, Niwayama G 1976 Comparison of androgen metabolites in benign prostatic hypertrophy (BPH) and normal prostate. Journal of Clinical Endocrinology 43: 686-688

Ghanadian R 1982 Mechanism of action of androgens. In: Chisholm G D, Williams D I (eds) Scientific Foundations of Urology, 2nd Edition, London, Heinemann Medical 138- 146.

Ghosh-Dastidar P, Coty W A, Griest R E, Woo D D, Fox C F 1984 Progesterone receptor sub-units are high affinity substrates for phosphorylation by epidermal growth factor receptor. Proceedings of the National Academy of Science U.S.A. 81: 1645-1658

Giannopoulous G, Gorski J 1971 Oestrogen receptors. Quanti tative studies on the transfer of estradiol from cytoplas mic to nuclear binding sites. Journal of Biological Chemi stry 246: 2524-2529

Gill G N, Bertics P J, Santon J B 1987 Epidermal growth factor and its receptor. Molecular and Cellular Endocrino- loçry 51: 169-186

259 Gil-Vernet S 1953 Biologie y Pathologie de le Prostate, Madrid, Pez-Montelvo.

Glegoleff A A 1933 On the geometrical methods of quantita tive minéralogie analysis of rocks. Transcripts of the

Institute of Minerology (USSR) 59: 5-18

Glynn R J, Campion E W, Bouchard G R, Silbert J E 1985 The development of benign prostatic hyperplasia among volun teers in the normative aging study. American Journal of Epidemiology 121: 78-90

Gonor S E, Lakey W H, McBlain W A 1984 Relationship be tween concentrations of extractable and matrix bound nuclear androgen receptor and clinical response to endo crine therapy for prostatic adenocarcinoma. Journal of Urology 131: 1196-1201

Gorski J, Toft D, Shymala G, Smith D, Notides A 1968 Hor mone receptors: studies on the interaction of oestrogens with the uterus. Recent Progress in Hormone Research 24: 45-80

Gotoh K, Nishi M 1965 Ultrasonic diagnosis of prostatic cancer. Acta Urologica Japan 11: 87-90

260 Grayhack J T, Bunce P L, Kearns J W, Scott W W 1955 Influ ence of the pituitary on prostatic response to androgen in the rat. Bulletin of The Johns Hopkins Hospital 96: 154-

163

Grayhack J T, Lebowitz J M 1967 Effect of prolactin on citric acid of lateral lobe of prostate of Sprague-Dawley rat. Investigative Urology 5: 87-94

Greenwald P, Kirmss V, Polan A K, Dick V S 1974 Cancer of the prostate among men with benign prostatic hyperplasia. Journal of the National Cancer Institute 53: 335-340

Gregory H, Willshire I R, Kavanagh J P, Blacklock N J, Chowdury S, Richards R C 1986 Urogastrone-epidermal growth factor concentrations in prostatic fluid of normal indi viduals and patients with benign prostatic hypertrophy. Clinical Science 70: 359-363

Griffiths K, Davies P, Harper M E, Peeling W B, Pierre- point C G 1979 The etiology and endocrinology of prostate cancer. In: Rose D (ed) Endocrinology of Cancer, CRC Press 2: 1-55.

261 Griffiths K, Davies P, Eaton C L, Harper M E, Peeling W B, Turkes A O, Turkes A, Wilson D W, Pierrepoint C G 1987 Cancer of the prostate: Endocrine factors. In: Oxford Review of Reproductive Biology, Oxford University Press,

192-259

Habib F K, Beynon L, Chisholm G D, Busuttil A 1983 The distribution of 5 alpha-reductase and 3 alpha (beta)- hydroxysteroid dehydrogenase activities in the hyperplas tic human prostate gland. Steroids 41: 41-53

Habib F K, Odoma S, Busuttil A, Chisholm G D 1986 Androgen receptors in cancer of the prostate. Correlation with stage and grade of tumour. Cancer 57: 2351-2356

Habib F K 1989 Prostate: Mechanisms of normal and abnormal metabolism. In: Chisholm G D, Innes Williams D (eds) Scientific Foundations of Urology, London, Heinmann 358 —365 •

Habib F K, Chisholm G D 1989 Management of advanced prostate cancer: the biological basis of endocrine treat ment. In: Bloom H J G (ed) Urological Oncology: Dilemmas and Developments. New York, John Wiley.

262 Hafiez A A, Lloyd C W, Bartke A 1972 The role of prolactin in the regulation of the testis function: the effects of prolactin and luteinising hormone on the plasma levels of testosterone and androstenedione in hypophysectomised rats. Journal of Endocrinology 52: 327-332

Hall T L, Fu Y S 1985 Biology of disease: Applications of quantitative microscopy in tumour pathology. Laboratory Investigation 53: 5-21

Hammond G L, Kontturi M, Vikho P 1978 Serum steroids in normal males and patients with prostatic diseases. Clini cal Endocrinology 9: 113-121

Hammond G L 1978 Endogeneous steroid levels in the human prostate from birth to old age: A comparison of normal and diseased states. Journal of Endocrinology 78: 7-19

Hanid A, Slavin G, Mair W, Sowter C, Ward P, Webb J, Levi J 1981 Fibre type changes in striated muscle of alcoho lics. Journal of Clinical Pathology 34: 991-995

Harman S M, Tsitouras P D 1980 Reproductive hormones in aging men. I.Measurement of sex steroids,basal luteinising hormone, and Leydig cell responses to human chorionic gonadotrophin. Journal of Clinical Endocrinology and

Metabolism 51: 35-40

263 Harper M E, Sibley PEC, Peeling W B, Griffiths K 1981 The immunocytochemical detection of growth hormone and prolactin in human prostatic tissues. In: Murphy G P, Sandberg A A, Karr J P (eds) The Prostatic Cell: Stucture and Function, Part B, New York, Alan R.Liss, 115-128.

Hasegawa Y, Kumazawa J 1987 Relationship between echogram and histology of benign prostatic hyperplasia. European Urology 13: 378-381

Haug H 1955 Die treffermethode ein verfahren zur quantita- tiven analyse im histologischen schnitt. Z. Anatomie En- twickl-Gesch 118: 302-321

Heathfield H A, Winstanley G, Kirkham N 1988 Computer assisted breast cancer grading. Journal of Biomedical Engineering 10: 379-386

Henning A 1957 Fehler der oberflachenbestimmung von Kernen bei endlicher schnittdicke. Mikroskopie 12: 7-18

Higgins S J, Gehring U 1978 Molecular mechanisms of ster oid hormone action. Advances in Cancer Research 28: 313- 397

Hill W C 0 1953-1966 Primates. Comparative anatomy and taxonomy. 6 Vols, Edinburgh, Edinburgh University Press.

264 Hofmann G E, Rao C V, Barrows G H, Schultz G S, Sanfilippo J S 1984 Binding sites for epidermal growth factor in human uterine tissues and leiomyomas. Journal of Clinical Endocrinology and Metabolism 58: 880-884

Horton R J 1976 Androgen hormones and prehormones in young and elderly men. In: Grayhack J T, Wilson J D and Scher- benske M J (eds) Benign Prostatic Hyperplasia. Proceedings of a workshop sponsored by the Kidney Disease and Urology Program of the NIAMDD. Washington, D.C. U.S. Govt. Print ing Office 183-188.

Houston B, Chisholm G D, Habib F K 1985 Evidence that human prostatic 5 alpha-reductase is located exclusively in the nucleus. FEBS Letters 185: 231-235

Howell R, Besser G, Wass J, Perry L, Chard T 1986 Testosterone: dihydrotestosterone ratios in men during replacement therapy with testosterone oenanthate and testosterone undecanoate. Journal of Endocrinology (suppl) 108: 85

Hsuan J J, Panayotou G, Waterfield M D 1989 Structural basis for epidermal growth factor receptor function. Progress in Growth Factor Research 1: 23-32

265 Huggins C, Clark P T 1940 Quantitative studies of prostatic secretion II: The effect of castration and of estrogen injection on the normal and on the hyperplastic prostate glands of dogs. Experimental Medicine 72: 747-767

Huggins C, Stevens R A 1940 The effect of castration on benign hypertrophy of the prostate in man. Journal of

Urology 43: 705-714

Huggins C, Hodges C V 1941 Studies of prostatic cancer: effect of castration, of estrogen and of androgen injec tion on serum phosphatases in metastatic carcinoma of the prostate. Cancer Research 1: 293-297

Huggins C, Russell P S 1946 Quantitative effects of hypophysectomy on testis and prostate of dogs. Endocrinology

39: 1-7

Huggins C 1967 Endocrine induced regression of cancer. Science 156: 1050-1054

Hulka B S, Hammond JE, Di Ferdinando G 1987 Serum hormone levels among patients with prostatic carcinoma or benign prostatic hyperplasia and clinic controls. The Prostate 11: 171-182

Hunter J 1786 Observations on certain parts of the animal oeconomy. 1st edition Bibliotheca Osteriana , London.

266 Hunter T, Cooper J A 1985 Protein-tyrosine kinases. Annual Review of Biochemistry 54: 897-930

Imai Y, Leung C K, Friesen H G, Shiu R P C 1982 Epidermal growth factor receptors and effect of epidermal growth factor on growth of human breast cancer cells in long term tissue culture. Cancer Research 42: 4394-4398

Isaacs J T, Brendler C B, Walsh P C 1983 Changes in the metabolism of dihydrotestosterone in the hyperplastic human prostate. Journal of Clinical Endocrinology and Metabolism. 56: 139-146

Jacobs S C, Pinka D, Lawson R K 1979 Prostatic osteoblas tic factor. Investigative Urology 17: 195-199

Jensen E V, Jacobsen H I 1962 Basic guides to the mechan ism of estrogen action. Recent Progress in Hormone Re search 18: 387-414

Jensen E V, Suzuki T, Kawashima T, Stumpf WE, Jungblut P W, Desombre E R 1968 A two-step mechanism for the inter- ction of estradiol with rat uterus. Proceedings of the national Academy of Science USA 59: 632-638

267 Jensen E V 1991 Overview of the nuclear receptor family. In: Parker M G (ed) Nuclear Hormone Receptors: Molecular Mechanisms, Cellular Functions, Clinical Abnormalities.

Academic Press, London 1-14.

Jones D R, Parkinson M C, Griffiths G J, Davies R LI, Peeling W B 1990 Origin and structure of benign prostatic hyperplasia. British Journal of Urology 66: 506-508

Jores L 1894 Ueber die hypertrophie des sogenannten mit- tleren lappens der prostata. Archives Pathologies: Anato mie Physioliologie Klinische Medicin 135: 224-247

Kabalin J N, Peehl D M, Stamey T A 1989 Clonal growth of human prostatic epithelial cells is stimulated by fibro blasts. Prostate 14: 251-263

Kaburagi Y, Marino M B, Kirdani R Y 1987 The possibility of aromatization of androgen in the prostate. Journal of

Steroid Biochemistry 26: 739-742

Kambal A 1977 Prostatic obstruction in Sudan. British

Journal of Urology 49: 139-141

Karr J P, Wajsman Z, Madajewicz S, Kirdani R Y, Murphy G

P, Sandberg A A 1979 Steroid hormone receptors in the prostate. Journal of Urology 122: 170-175

268 Kavanaugh J P, Chowdhury S D, Richards C, Gregory H 1983 Urogastrone (epidermal growth factor) in human prostatic fluid. Proceedings of the Annual Conference of Society for the Study of Fertility Abstract 100.

King A L, Cuatrecasas P 1982 Resolution of high and low affinity epidermal growth factor receptors. Journal of Biological Chemistry 257: 3053-3060

King C R, Borrello I, Bellot F, Comoglio P, Schlessinger J 1988 EOF binding to its receptor triggers a rapid tyrosine phosphorylation of the erbB-2 protein in the mammary tumour cell line SK-BR-3 Embo Journal 7: 1647-1651

King R J B, Redgrave S, Hayward J L, Millis R R, Rubens R D 1979 The measurement of receptors for oestradiol and progesterone in human breast tumours. In: King R J B (ed) Steroid receptor assays in human breast tumours: methodological and clinical aspects. Alpha Omega Publishing, Cardiff, 55-72.

King R J B, Greene G L 1984 Monoclonal antibodies localize oestrogen receptor in the nuclei of target cells. Nature

307: 745-747

269 Kirby R S, Eardley I, Bryan J, Miller P D, Christmas T 1991 A urodynamic study of the value of finasteride in the management of bladder outflow obstruction due to BPH.

Journal of Urology (suppl) 145: 213

Kirdani R Y, Pontes E J, Murphy G P 1984 Correlation of estrogen and androgen receptor status in prostatic disease measured by high pressure liquid chromatography. Journal of Steroid Biochemistry 20: 401-411

Knabbe C, Lippman M E, Greene G L, Dickson R B 1988 Phos phorylation of the estrogen receptor in MCF7 human breast cancer cells? In: Bresciani F, King R J B, Lippman ME, Raynaud J P (eds) Progress in Cancer Research and Therapy, Raven Press, New York 39-42.

Kosak I, Bartsch W, Kreig M, Voigt K D 1982 Nuclei of stroma: site of highest estrogen concentration in human benign prostatic hyperplasia. The Prostate 3: 433-438

Kreig M, Bartsch W, Janssen W, Voigt K D 1979 A compara tive study of binding metabolism and endogenous levels of androgens in normal, hyperplastic and carcinomatous human prostate. Journal of Steroid Biochemistry 11: 615-624

270 Kreig M, Klotzl G, Kaufmann J 1981 Stroma of human benign prostatic hyperplasia: preferential tissue for androgen metabolism and oestrogen binding. Acta Endocrinologica 96:

422-432

Kreig M, Bartsch W, Thomsen M, Voigt K D 1983 Androgens and estrogens: Their interaction with stroma and epitheli um of human benign prostatic hyperplasia and normal prostate. Journal of Steroid Biochemistry 19: 155-161

Lacassagne A 1932 Apparition de cancers de la mamelle chez les souris male soumise a des injections de colluculine. Compt. Rendu Acadamie D.Sc.(iii) 195: 630-632

Lahotnen R, Bolton N J, Kontturi M, Vihko R 1983 Nuclear androgen receptors in the epithelium and stroma of human benign prostatic hypertrophic glands. The Prostate 4: 129- 139

Laurence D M, Landau R L 1965 Impaired ventral prostate affinity for testosterone in hypophysectomised rats.

Endocrinology 73: 1119-1125

Leake A, Chisholm G D, Habib F K 1983 Characterisation of the prolactin receptors in human prostate. Journal of

Endocrinology 99: 321-328

271 Le Duc I E 1939 The anatomy of the prostate and pathology of early benign hypertrophy Journal of Urology 42: 1217- 1241

Lee F 1987 Transrectal ultrasound in the diagnosis, sta ging, guided needle biopsy, and screening for prostate

cancer. Progress in Clinical and Biological Research 237:

73-109

LeFebvre Y A, Goldsteyn E J, Michiel T L 1985 Androgen interactions with intact nuclear envelopes from the rat ventral prostate. Journal of Steroid Biochemistry 23: 107-

111

Lipkin L E 1966 The analysis, synthesis and description of biological images. Annals of the New York Academy of Science 51: 984-1011

Lippman M E, Dickson R B 1989 Mechanisms of growth control in normal and malignant breast epithelium. Recent Progress

in Hormone Research 45: 383-440

Lostroh A J, Li C H 1957 Stimulation of the sex accessor ies of hypophysectomised male rat by non-gonadotrophic hormones of the pituitary gland. Acta Endocrinologica Copenhagen 25: 1-16

272 Lowry 0 H, Rosebrough N J, Farr A L, Randall R J 1951

Protein measurement with the folin phenol reagent. Journal of Biological Chemistry 193: 265-275

Lowsley O S 1912 The development of the human prostate gland with reference to the development of other struc tures at the neck of the urinary bladder. American Journal of Anatomy 13: 229-349

Lubrano C, Petrangeli E, Catizone A, Santonati A, Concoli- no G, Rombola M, Frati L, Di Silverio F, Sciarra F 1989

Epidermal growth factor binding and steroid receptor content in human benign prostatic hyperplasia. Journal of

Steroid Biochemistry 34: 499-504

Lytton B, Emery J M, Harvard B M 1968 The incidence of benign prostatic hypertrophy. Journal of Urology 99: 639-

645

MacCorquodale D W, Thayer S, Doisy E A 1936 Isolation of the principal oestrogenic substance of liquor folliculi.

Journal of Biological Chemistry 115: 435-448

Maddy S Q, Chisholm G D, Hawkins R A, Habib F K 1987

Localization of epidermal growth factor receptors in the human prostate by biochemical and immuncytochemical me thods. Journal of Endocrinology 113: 147-153

273 Maehama S, Li D, Nanri H, Leykam J F, Deuel T F 1986

Purification and partial characterisation of a prostate- derived growth factor. Proceedings of the National Academy of Science U.S.A. 83: 8162-8166

Mainwaring W I P 1969 The binding of [1,2-^H] testosterone within the nuclei of the rat prostate. Journal of Endocri nology 44: 323-333

Mainwaring W I P 1977 The Mechanism of Action of Andro gens. Springer-Verlag, Berlin.

Margolis B, Rhee S G, Felder S, Mervic M, Lyall R, Levit zki A, Ullrich A, Zilberstein A, Schlessinger J 1985 EGF induces tyrosine phosphorylation of phophoslipase C-II: A potential mechanism for EGF receptor signalling. Cell 57:

1101-1107

Marriotti A, Mawhinney M 1981 Preliminary studies of the hormonal control of male accessory sex organ epithelial collagen. In: Murphy G P, Sandberg A A, Karr J P (eds) The

Prostatic Cell. Structure and Function. New York, Liss

Inc. 133-136

Mawdesley-Thomas L E, Healey P 1969 The computer joins the microscope. New Scientist 6: 286-287

274 McKeehan W L, Adams P S, Rosser M P 1984 Direct mitogenic effects of insulin, EGF, glucocorticoid, cholera toxin, unknown pituitary factors and possibly prolactin but not androgen on normal rat prostate epithelial cells in serum- free, primary cell culture. Cancer Research 44: 1998-2010

McNeal J E 1968 Regional morphology and pathology of the prostate American Journal of Clinical Pathology 49: 347- 357

McNeal J E 1969 Origin and development of carcinoma in the prostate. Cancer 23: 24-34

McNeal J E 1972 The prostate and prostatic urethra : A morphological synthesis. Journal of Urology 107: 1008-1016

McNeal J E 1978 Origin and evolution of benign prostatic enlargement. Investigative Urology 15: 340-345

McNeal J E 1981 The zonal anatomy of the prostate.

Prostate 2: 35-49

Meikle A W, Stringham J D, Olsen D C 1978 Subnormal tissue

3 alpha-androstanediol and androsterone in prostatic hyperplasia. Journal of Clinical Endocrinology. 47: 909-

913

275 Menard D, Pothier P , Gallo-Payet N 1987 Epidermal growth factor receptors during post natal development of the mouse colon. Endocrinology 121: 1548-1554

Meyhoff H H, Hald T 1978 Are Doctors able to assess prostatic size? Scandinavian Journal of Urology and Ne phrology 12: 219-221

Miller W R, Telford J, Hawkins R A 1983 Binding of [^H]- methyltrienolone (R1881) by human breast cancers. European Journal of Cancer and Clinical Oncology 19: 1473-1478

Mobbs B G, Johnson I E, Connolly J G 1975 In vitro assay of androgen binding in human prostate. Journal of Steroid Biochemistry 57: 371-384

Moore R A 1935 The morphology of small prostatic carcino ma. Journal of Urology 33: 224-234

Moore R A 1936 The evolution and involution of the prostate gland. American Journal of Pathology 12: 599-623

Moore R A 1944 Benign hypertrophy and carcinoma of the prostate : occurence and experimental production in ani mals. Surgery 16: 152-160

276 Moore R J, Gazak J M, Quebbeman J F, Wilson J D 1979

Concentration of dihydrotestosterone and 3-androstanediol in naturally occurring and androgen induced prostatic hyperplasia in the dog. Journal of Clinical Investigation

64: 1003-1019

Morgagni G 17 69 The seats and causes of disease investiga ted by anatomy. Millar A, Cadell T (eds) Book 3 Johnson and Payne, London 460-462.

Morphin R F, DiStefano S, Bercovici J P 1978 Comparison of testosterone, 5 alpha-dihydrotestosterone and

5 alpha-androstane-3 beta,17 beta-diol metabolisms in human normal and hyperplastic prostates. Journal of Ster oid Biochemistry 9: 245-252

Morgan C, Spruell S 1983 Increased trends of prostate operations in Alabama. Journal of the Medical Association of Alabama 5: 17-20

Morrison A S 1978 Prostatic hypertrophy in Greater Boston.

Journal of Chronic Diseases 31: 357-362

Muccu V R, Stancel G M 1985 Regulation of epidermal growth factor receptors by oestrogens. Journal of Biological

Chemistry 260: 9820-9824

277 Muhlbock O 1972 Role of hormones in the aetiology of breast cancer. Journal of the National Cancer Institute 48: 1213-1216

Munson P J, Robard D 1980 LIGAND: a versatile computerised approach for characterization of ligand binding systems. Analytical Biochemistry 107: 220-239

Muntzig J, Liljekvist J, Murphy G P 1979 Chalones and stroma as possible growth limiting factors in the rat ventral prostate. Investigative Urology 16: 399-402

Murphy J B, Emmott R C, Hicks L L 1980 Estrogen receptors in the human prostate, seminal vesicles, epididymis, testis and genital skin: a marker for estrogen-resonsive tissue? Journal of Clinical Endocrinology and Metabolism 50: 838-842

Murphy L J, Vrhovsek E, Lazarus L 1983 Characterization of specific growth hormone binding sites in mouse fibro blasts. Endocrinology 113: 750-757

Murphy L J, Sutherland R L, Stead B, Murphy L C, Lazarous L 1986 Progestin regulation of epidermal growth factor receptor in human mammary carcinoma cells. Cancer Research

46: 728-734

278 Myatt L, Elder M G, Lim L 1979 The binding of rat uterine cytosol oestrogen receptor to oligothymidylate-cellulose: further evidence for the presence of a cytosol activating factor. Biochemical Society Transcripts 7: 909-910

Mydlo J H, Bulbul M A, Ridon V M, Heston W D, Fair W R

1988 Expression of basic fibroblast growth factor mRNA in benign prostatic hyperplasia and prostatic carcinoma.

Prostate 13: 241-247

Nicholson S, Sainsbury J R C, Needham G K, Chambers P, Farndon J R, Harris A L 1988 Quantitative assays of epi dermal growth factor receptor in human breast cancer: Cut off points of clinical relevance. International Journal of Cancer 42: 36-42

Odoma S, Chisholm G D, Nicol K, Habib F K 1985 Evidence for the association between blood prolactin and androgen receptors in BPH. Journal of Urology 133: 717-720

Peck E J, Burgner J, Clark J H 1973 Oestrophilic binding sites of the uterus. Relation to uptake and retention of oestradiol in vitro. Biochemistry 12: 4596-4603

Pertschuk L P, Eisenberg K B, Macchia R J 1985 Heterogene ity of steroid binding sites in prostatic carcinoma: morphological demonstration and clinical implications.

Prostate 6: 35-47

279 Peters C A, Barrack E R 1987 Androgen receptor localisa tion in the human prostate: demonstration of heterogeneity using a new method of steroid receptor autoradiography. Journal of Steroid Biochemistry 27: 533-541

Peutilla A, McDowell E M, Trump B F 1975 Effects of fixa tion and post fixation treatment on volume of injured cells. Journal of Histochemistry and Cytochemistry 23:

251-255

Pirke K M, Doerr P 1975 Age related changes in free plasma testosterone, dihydrotestosterone and oestradiol. Acta Endocrinologica 80: 171-178

Ploem J S, Verwoerd N, Bennet J, Koper G 1979 An automated microscope for quantitative cytology combining television image analysis and stage scanning microphotometry. Journal of Histochemistry and Cytochemistry 27: 136-143

Pontes J E, Karr J S, Kirdani R Y, Murphy G P, Sandberg A

A 1982 Estrogen receptors and the clinical correlations with human prostatic disease. Urology 19: 399-403

Pradham B K, Chandra K 1975 Morphogenesis of nodular hyperplasia-prostate. Journal of Urology 113: 210-215

280 Pratt W B, Sanchez E R, Bresnick E H, Meschinchi S, Scher- rer L C, Dalman F C, Welsh M J 1989 Interaction of the glucocorticoid receptor with the M 90,000 heat shock protein: an evolving model of ligand mediated receptor transformation and translocation. Cancer Research (suppl)

49: 2222-2229

Price D, Williams Ashman H G 1961 The accessory reproduc tive glands of mammals. In: Young W C (ed) Sex and Inter nal Secretions. Third Edition, Baltimore, Williams and Wilkins Co. Chapter 6. 366-488.

Price D 1963 Comparative aspects of development and struc ture in the prostate. In: Biology of the Prostate and Related Tissues. United States National Cancer Institute Monographs 12: 1-14

Puca G A, Sica V, Nola E 1974 Identification of a high affinity nuclear acceptor site for oestrogen receptor of

calf uterus. Proceedings of the National Academy of Sci

ence USA 71: 979-983

Puddefoot J R, Anderson E, Vinson G P, Gilmore O J A 1987 Heterogeneity of oestrogen receptors in human breast tumours. In: Peeters H (ed) Protides of the Biological Fluids. Proceeds of the 35^^ colloquium.

Pergamon Press, Oxford, 307-310.

281 Rannikko S, Aldercreutz H 1983 Plasma oestradiol, free testosterone, sex hormone binding capacity, and prolactin in benign prostatic hyperplasia and prostatic cancer. The Prostate 4; 223-229

Richardson I M 1965 Prostatic hyperplasia and social class. British Journal of Preventive and Social Medicine 18: 157-162

Reischauer F 1925 Die Entstehung der sogenannten Prostata- hypertrophie. Virchows Archives [Pathol.Anat.Physiol] 256: 357-389

Robson M C 1964 The incidence of benign prostatic hyper trophy and prostatic carcinoma in cirrhosis of the liver Journal of Urology 92: 307-310

Rohr H, Oberholzer M, Bartsch G, Keller M 1976 Morphometry in experimental pathology: methods baseline data and applications. In: Richter G.W. and Epstein M.A. (eds.)

International Review of Experimental Pathology, New York:

Academic Press 233-325.

Rotkin I D 1976 Epidemiology of benign prostatic hypertro

phy: review and speculations. In: Grayhack J T, Wilson J

D, Sherbenske M J (eds) Benign Prostatic Hyperplasia. DHEW

Publ.No. (NTH) 76-1113, US Govt.Printing Office, Washing

ton DC, 105-117.

282 Rotkin I D 1983 Origins, Distribution and Risk of Benign Prostatic Hypertrophy. In: Hinman F. (ed) Benign Prostatic Hypertrophy. Springer-Verlag, New York. 10-21.

Rubens R, Dhont M, Vermeulen A 1974 Further studies on

Leydig cell function in old age. Journal of Clinical Endo crinology and Metabolism 39: 40-45

Saitoh M, Watanabe H, Ohe H 1981 Ultrasonically guided puncture for the prostate and seminal vesicles with transrectal real-time linear scanner. Journal of Kyoto Univers ity of Medicine 90: 47-53

Sainsbury J R C, Malcolm A J, Appleton D R, Farndon J R, Harris A L 1985 Presence of epidermal growth factor recep tor as an indicator of poor prognosis in patients with breast cancer. Journal of Clinical Pathology 38: 1225-1228

Sandberg A A 1981 Some experimental results with prolac tin: an overview of effects in the prostate. In: Murphy G P, Karr J P, Sandberg A A (eds) The Prostatic Cell: Struc ture and Function: Part B. New York, Alan R Liss 15-28.

Sawyer S T, Cohen S 1987 Enhancement of calcium uptake and phosphatidylinositol turnover by epidermal growth factor in A-431 cells. Journal of Biological Chemistry 262: 6280-

6287

283 Scatchard G 1949 The attraction of protein for small mole cules amd ions. Annals of New York Academy of Science 51: 660-672

Schlessinger J 1988 The epidermal growth factor receptor as a multifunctional allosteric protein. Biochemistry 27: 3119-3123

Schuurmans A L G, Bolt J, Mulder E 1988 Androgens stimu late both growth rate and epidermal growth factor receptor activity of the human prostate tumour cell LNCaP. The Prostate 12; 55-63

Schweikert H U 1979 Conversion of androstenedione to estrone in human fibroblasts from prostate, genital and non genital skin. Hormones and Metabolism Research 11; 635-640

Schweikert H U, Totzauer P, Rohr H P, Bartsch G 1985

Correlated biochemical and stereological studies on tes tosterone metabolism in the stromal and epithelial com partment of human benign prostatic hyperplasia. Journal of

Urology 134: 403-407

Schweikert H U, Tunn U W 1987 Effects of the aromatase inhibitor testolactone on human benign prostatic hyperpla sia. Steroids 50: 62-65

284 Sciarra F, Concolino G, Tenaglia R, Di Silverio F 1988 Biochemical modifications in benign prostatic hyperplasia. In: Di Silverio F, Steg A (eds) International Workshop of Urology. Acta Medica, Rome 277-283.

Sekine H, Oka K, Takehara Y 1982 Transrectal longitudinal ultrasonotomography of the prostate by electronic linear scanning. Journal of Urology 127: 62-65

Shain S A, Boesel R W, Lamm D L, Radwin H M 1980 Cytoplas mic and nuclear androgen receptor content of normal and neoplastic human prostates and lymph node métastasés of human prostatic adenocarcinoma. Journal of Clinical Endo crinology and Metabolism 50: 704-711

Shapiro E, Lepor H, Becich M 1991 The relationship between histology and clinical response to alpha blockade in men with symptomatic BPH. Journal of Urology 145 (Suppl): 211

Shikata H, Utsumi N, Hiramatsu M 1984 Immunohistochemical localization of nerve growth factor and epidermal growth factor in guinea pig prostate gland. Histochemistry 80: 411-416

Shymala G, Gorski J 1969 Oestrogen receptors in the rat uterus. Studies on the interaction of cytosol and nuclear binding sites. Journal of Biological Chemistry 244: 1097-

1103

285 Sibley PEC, Harper M E, Peeling W B, Griffiths K 1984 Growth hormone and prostatic tumours: localization using a monoclonal human growth hormone antibody. Journal of Endocrinology 103: 311-315

Sibley D R, Benovic J L, Caron M G, Lefkowitz R J 1988 Phosphorylation of cell surface receptors: a mechanism for regulating signal transduction pathways. Endocrine Review 9: 38-66

Siiteri P K, Wilson J D 1970 Dihydrotestosterone in prostatic hypertrophy. I. The formation and content of dihydrotestosterone in the hypertrophic prostate of man. Journal of Clinical Investigation 49: 1737-1745

Siiteri P K, Wilson J D 1974 Testosterone formation and metabolism during male sexual differentiation in the human embryo. Journal of Clinical Endocrinology and Metabolism

38: 113-125

Slavin G, Sowter C, Robertson K, McDermott S, Paton K 1980 Measurement in jejunal biopsies by computer-aided micro scopy. Journal of Clinical Pathology 33: 254-261

Smith T, Chisholm G D, Habib F K 1982 Failure of human benign prostatic hyperplasia to aromatise testosterone. Journal of Steroid Biochemistry 17: 119-120

286 Sowter C, Slavin G, Rosen D 1987 Morphometry of bladder carcinoma: I. The automatic delineation of urothelial nuclei in tissue sections using an IBAS II image array processor. Journal of Pathology 153: 289-297

Sporn M B, Roberts A B 1985 Autocrine growth factors and cancer. Nature 313: 745-747

St Arnaud R, Poyet P, Walker P, Labrie F 1988 Androgens modulate epidermal growth factor receptor levels in the rat ventral prostate. Molecular and Cellular Endocrinology 56: 21-27

Stebbings W S L, Anderson E, Puddefoot J R, Farthing M J G, Vinson G P 1987 A hybrid ligand method for androgen receptor measurement. Journal of Steroid Biochemistry 31: 181-186

Steins P, Kreig M, Hollman H J, Voigt K D 1974 In vitro studies of testosterone and 5 alpha dihydrotestosterone binding in benign prostatic hypertrophy. Acta Endocrinolo gica 75: 773-778

Stone N N, Fair W R, Fishman J 1986 Estrogen formation in human prostatic tissue from patients with and without benign prostatic hyperplasia. The Prostate 9: 311-318

287 story M T, Jacobs S C, Lawson R K 1983 Epidermal growth factor is not the major growth-promoting agent in extracts of prostatic tissue. Journal of Urology 130: 175-179

Story M T, Esch F, Shimasaki S, Sasse J, Jacobs S C,

Lawson R K 1987 Amino-terminal sequence of a large form of basic fibroblast growth factor isolated from human benign prostatic hyperplastic tissue. Biochemical and Biophysical Research Communications 142: 702-709

Stuart A E, Warford A 1983 Staining of human splenic sinusoids and demonstration of unusual banded structures by monoclonal antisera. Journal of Clinical Pathology 30: 1176-1180

Swenerton K D 1982 Endometrial adenocarcinoma In: Furr B J A (ed) Clinics in Oncology. W.R.Saunders Co Ltd, London 215-232.

Swyer G I M 1944 Post-natal growth changes in the human prostate. Journal of Anatomy 78: 130-145

Syms A J, Harper M E, Griffiths K 1985 The effect of prolactin on human BPH epithelial cell proliferation. The

Prostate 6: 145-153

288 Takahashi H, Ouchi T 1963 The ultrasonic diagnosis in the field of urology. Proceedings of the Japanese Society of Ultrasonics in Medicine 3: 7-10

Takahashi H, Ouchi T 1964 The ultrasonic diagnosis in the

field of urology. Proceedings of the Japanese Society of Ultrasonics in Medicine 4: 35-37

Tarin D 1972 Tissue interactions in morphogenesis, mor- phostasis and carcinogenesis. Journal of Theoretical Biology 34: 61-67

Tenniswood M 1986 Role of epithelial-stromal interactions in the control of gene expression in the prostate: An hypothesis. Prostate 9: 375-385

Thompson S A, Rowley D R, Heidger P M 1979 Effects of estrogen upon the fine structures epithelium and stroma in

the rat ventral prostate gland. Investigative Urology 17: 83-89

Thornton M O, Fredrickson R, Matal J, Mawhinney M 1985 Collagen and the proliferation and differentiation of rat

ventral prostate epithelial cells. Biology of Reproduction

33: 213-227

289 Tilley W D, Keightley D D, Marshall V R 1980 Oestrogen and progesterone receptors in benign prodtatic hyperplasia in humans. Journal of Steroid Biochemistry 13: 395-399

Tisell L E, Salander H 1975 The lobes of the human prostate. Scandinavian Journal of Urology and Nephrology 9: 185-191

Todaro G L, Fryling C, De Larco J E 1980 Transforming growth factors produced by certain human tumour cells: polypeptides that interact with epidermal growth factor receptors. Proceeding of the National Academy of Science U.S.A. 77: 5258-5262

Toft D, Gorski J 1966 A receptor molecule for oestrogens: isolation from the rat uterus and preliminary characteri zation. Proceedings of the National Academy of Science USA 55: 1574-1581

Trachtenberg J, Hicks L L, Walsh P C 1980 Androgen and

oestrogen receptor content in spontaneous and experiment ally induced canine prostatic hyperplasia. Journal of Clinical Investigation 65: 1051

Trachtenberg J, Bruchovsky P, Walsh P C 1981 Androgen

receptor content of normal and hyperplastic human prostate. Journal of Clinical Endocrinology and Metabolism

54: 17-21

290 Trachtenberg J, Walsh P C 1982 Correlation of prostatic nuclear receptor content with duration of response and survival following hormonal therapy in advanced prostatic cancer. Journal of Urology 127: 466-471

Traish A M, Wotiz H H 1987 Prostatic epidermal growth factor receptors and their regulation by androgens. Endo crinology 121: 1461-1467

Tucker J H 1979 An image analysis system for cervical cytology automation using nuclaer DNA content. Journal of

Histochemistry and Cytochemistry 27: 613-620

Tullner W W 1963 Hormonal factors in the adrenal-dependent growth of the rat ventral prostate. National Cancer Insti tute Monographs 12: 211-223

Tunn U VJ, Kaivers P, Schweikert H U 1985 Conservative treatment of human benign prostatic hyperplasia. In: Bruchovsky N, Chapeldaine A, Neumann F (eds) Regulation of Androgen Action. Congressdruck R. Bruckner, Berlin 87-

90.

Turner-Warwick R T, Whiteside C G, Arnold E P, Bates C P, Worth P H L, Milroy E G J, Webster J R, Weir J 1973 A urodynamic view of prostatic obstruction and the results of prostatectomy. British Journal of Urology 45: 631-645

291 Ullrich A, Coussens L, Hayflick J S, Dull T J, Gray A, Tam

A W, Lee J, Yarden Y, Liberman T A, Schlessinger J, Down

ward J, Mayes E L V, Whittle N, Waterfield M D, Seeburg P

H 1984 Human epidermal growth factor cDNA sequence and

aberrant expression of the amplified gene in A431 epider

moid carcinoma cells. Nature 309: 418-425

Ushiro H, Cohen S 1980 Identification of phosphotyrosine as a product of epidermal growth factor-activated protein kinase in A-431 cell membranes. Journal of Biological Chemistry 255: 8363-8365

Van Aubel O G J M, Bolt-de Vries J, Blankenstein M A, Kate F J W, Schroder F H 1985 Nuclear androgen receptor content in biopsy specimens from histologically normal, hyperplas tic and cancerous human prostate tissue. Prostate 6: 185- 194

Van Wagenen G 1936 The coagulating function of the cranial lobe of the prostate gland in the monkey. Anatomical

Record 66: 411-421

Vermeulen A, Rubens R, Verdonck L 1972 Testosterone secre tion and metabolism in male senescence. Journal of Clini

cal Endocrinology 34: 730-735

292 Vermeulen A, De Sy W 1976 Androgens in patients with benign prostatic hyperplasia before and after prostatect

omy. Journal of Clinical Endocrinology and Metabolism 43:

1250-1254

Walsh P C, Wilson J D 1976 The induction of prostatic hypertrophy in the dog with androstanediol.

Journal of Clinical Investigation 57: 1093-1097

Walsh P C, Hutchins G M, Ewing L L 1983 The tissue content of dihydrotestosterone in human prostatic hyperplasia is not supranormal. Journal of Clinical Investigation 72:

1772-1777

Walsh P C 1984 Human benign prostatic hyperplasia: Etio

logical considerations. In: Kimball F A, Buhl A E (eds)

New approaches to the study of benign prostatic hyperpla

sia. Alan R.Liss Inc., New York. 1-25

Walton W H 1952 Automatic counting of microscopic parti cles. Nature 169: 518-520

Watanabe H, Katoh H, Katoh T 1968 Diagnostic application of ultrasonotomography for the prostate. Japanese Journal of Urology 59: 273-279

293 Watanabe H, Katoh H, Tanaka M, Katoh H 1971 Diagnostic application of ultrasonotomography to the prostate. In vestigative Urology 8: 548-559

Weibel E R 1963 Principle and methods for the morphometric

study of the lung and other organs. Laboratory Investiga tion 12: 131

Weibel E R, Kistler G S, Scherle W F 1969 Practical

stereological methods for morphometric cytology. Journal

of Cell Biology 30: 23-28

Weibel E R 1969 Stereological principles for morphometry in

electron microscopic cytology. International Review of

Cytology (New York) 26: 235-302

Weid G L, Bartels P H, Dyten H I, Pishotta L, Bibbo M 1982

Rapid high-resolution cytometry. Analytical and Quantita tive Cytology 4: 257-262

Welshons W V, Lieberman M E, Gorski J 1984 Nuclear locali

zation of unoccupied oestrogen receptors. Nature 307: 747-

749

Weiss J 1989 Atlas of transrectal ultrasonography. GEM

Medical Systems. New York

294 White J W 1895 The results of double castration in hyper trophy of the prostate. Annals of Surgery 22: 1-80

Wild J J, Reid J M 1955 Echographic tissue diagnosis.

Proceedings 4th Annual Conference On Ultrasonic Therapy:

1-26

Wilkin R P, Bruchovsky N, Shnitka T K 1980 Stromal 5-alpha reductase activity is elevated in benign prostatic hyper plasia. Acta Endocrinologica 94: 284-288

Wilkin R P, Bruchovsky N, Shnitka T A, Rennie P S, Comeau

T L 198 0 Stromal 5 alpha-reductase activity is elevated

in benign prostatic hyperplasia Acta Endocrinologica 94:

284-288

Wilson J D, Gloyna R E 1970 The intranuclear metabolism of testosterone in the accessory organs of reproduction.

Recent Progress in Hormone Research 26: 309-3 3 6

Wilson J D 1980 The pathogenesis of benign prostatic hyperplasia. American Journal of Medicine 68: 745-756

Witorsch R J 1981 Visualisation of prolactin binding sites

in prostate tissue. In: Murphy G P, Sandberg A A, Karr J P

(eds) The Prostatic Cell: Stucture and Function, Part B,

New York, Alan R Liss, 89-113.

295 Yamanka H, Kirdani R Y, Saroff J, Murphy G P, Sandberg A A

1977 Hormonal control of zinc uptake and binding in the rat dorsolateral prostate. Investigative Urology 14: 492-

495

Yarden Y, Harari I, Schlessinger J 1985 Purification of an active EGF receptor kinase with monoclonal antireceptor antibodies. Journal of Biological Chemistry 260: 315-319

Yarden Y, Schlessinger J 1987a Self-lf-phosphorylation of epidermal growth factor receptor: Evidence for a model of intermolecular allosteric activation. Biochemistry 26:

1434-1442

Yarden Y, Schlessinger J 1987b Epidermal growth factor induces rapid, reversible aggregation of the purified epidermal growth factor receptor. Biochemistry 26: 1443-

1451

Yarden Y, Ullrich A 1988 Growth factor receptor tyrosine kinases. Annual Review of Biochemistry 57: 443-478

Zava D T, Landrum B, Horwitz K B, Mcguire W L 1979 Androgen receptor assay with [3H] methyltrienolone (R1881) in the presence of progesterone receptors. Endocrinology 104:

1007-1012

296 Zuckerman S 1936 The endocrine control of the prostate.

Proceedings of the Royal Society of Medicine 29: 1557-1568

Zumoff B, Strain G W, Kream J, O'Connor J, Rosenfeld R S,

Levin J, Fukushima D K 1982 Age variation of the 24-hour mean plasma concentration of androgens, estrogens, and gonadotrophins in normal adult men. Journal of Clinical

Endocrinology and Metabolism 54: 534-539

297 Presentations/Publications

The following presentations to learned societies and publications have so far resulted from work presented in or directly related to this thesis.

Presentations

British Endocrine Societies AGM (1990) "Epidermal growth factor binding in the prostate" Miller P D, Barker S, Puddefoot J R, Vinson G P, Kirby R S

"Control of EGF receptors in breast cancer cell lines" Barker S, Puddefoot J R, Miller P D, Vinson G P

Guy's Urological Research Meeting (1990) "Studies on growth factor binding in the prostate" Miller P D, Barker S, Kirby R S

British Prostate Group (1990) "Biomorphometry in benign prostatic hyperplasia" Miller P D, Kirby R S

EORTC Hormonal Manipulation of Cancer: Symposium (1990) "Growth factor receptors and prostate cancer"

Miller P D, Barker S, Puddefoot J R, Kirby R S, Vinson G P

298 American Urological Association AGM (1990) "Regional variation of growth factor binding within the prostate" Miller P D, Eardley 1, Barker S, Kirby R S

Urological Society of Australasia (1990) "Growth factors in benign and malignant prostatic hyper plasia" Miller P D, Eardley I, Barker S, Kirby R S

European Association of Urology (1990) "Clinical implications of growth factor binding in prostatic neoplasia" Miller P D, Vale J, Barker S, Kirby R S

British Association of Urological Surgeons AGM (1990) "Application of a new technique to determine the relative amounts of glandular and stromal tissue in BPH" Miller P D, Sowter C, Slavin G, Kirby R S

International Cancer Congress (1990) "Prostatic neoplasia and epidermal growth factor" Miller P D, Barker S, Puddefoot J R, Vinson G P

International Cancer Congress (1990) "EGF receptor regulation in breast cancer"

Barker S, Puddefoot J R, Miller P D, Goode A W, Vinson G P

299 International Congress of Morphometry in Morphological Diagnosis (1990)

"Biomorphometry using image analysis in benign prostatic hyperplasia"

Miller P D, Sowter C, Kirby R S, Slavin G

London Hormone Group (1990) "Biochemical studies on hormone manipulation in benign prostatic hyperplasia" Miller P D, Barker S, Puddefoot J R, Vinson G P

British Prostate Group (1991) "Relationship of growth factor binding and oestrogen receptors in benign prostatic hyperplasia" Miller P D, Liu S, Vinson G P, Barker S, Kirby R S

Radiology and Oncology (1991) "Growth control in prostatic neoplasia" Miller P D, Liu S, Kirby RS

Radiology and Oncology (1991) "Computerised image analysis in prostatic neoplasia" Miller P D, Sowter C, Slavin G, Kirby R S

Urological Society Of Australasia (1992)

"Growth factors in benign prostatic hyperplasia - update"

Miller P D, Barker S, Kirby R S

300 Publications

Barker S, Puddefoot J R, Miller P D, Vinson G P 1990 Control of EOF receptors in breast cancer cell lines. Journal of Endocrinology (suppl) 124: 153

Miller P D, Barker S, Puddefoot J R, Kirby R S, Vinson G P 1990 Epidermal growth factor binding in the prostate. Journal of Endocrinology (suppl) 124: 111

Miller P D, Eardley I, Barker S, Kirby R S 1990 Regional variation of epidermal growth factor binding within the prostate. Journal of Urology 143 (4): 598

Miller P D, Puddefoot J R, Barker S, Kirby R S, Vinson G P 1990 Growth factor receptors and prostate cancer. European Journal of Cancer 26 (2): 185

Miller P D, Puddefoot J R, Barker S, Kirby R S, Vinson G P

1990 EGF in BPH and prostate cancer. EORTC Monographs

Barker S, Panahy C, Puddefoot J R, Miller P D, Goode A W, Vinson G P 1990 EGF receptor regulation in breast cancer. Journal of Cancer Research and Clinical Oncology 116: 456

Miller P D, Barker S, Puddefoot J R, Vinson G P 1990 Prostatic neoplasia and epidermal growth factor. Journal

of Cancer Research and Clinical Oncology 116: 458

301 Miller P D, Vale J, Barker S, Kirby R S 1990 Clinical implications of growth factor binding in prostatic neopla sia. European Urology 18(20): 660

Eaton C L, Miller P D 1991 Epidermal growth factor in the pathogenesis of benign prostatic hyperplasia. Prospectives

1(8): 6-9

Miller P D, Sowter C, Kirby R S 1992 Automated image analysis in the morphometry of benign prostatic hyperpla sia. British Journal of Urology. Submitted for publication

302