Transection of the Pituitary Stalk in Man: Anatomical Changes in the Pituitary Glands of 21 Patients

Home , Pituitary stalk

J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.29.6.545 on 1 December 1966. Downloaded from J. Neurol. Neurosurg. Psychiat., 1966, 29, 545

Transection of the pituitary stalk in man: anatomical changes in the pituitary glands of 21 patients

J. HUME ADAMS, PETER M. DANIEL, AND MARJORIE M. L. PRICHARD From the Department ofPathology, The University and Western Infirmary, Glasgow, the Department of Neuropathology, Institute ofPsychiatry, Maudsley Hospital, London, and the Nuffield Institute for Medical Research, University of Oxford

In recent years hypophysectomy (Luft and Olive- MATERIAL AND METHODS crona, 1953) and adrenalectomy (which is nearly The pituitary glands from 21 patients were available for always combined with ovariectomy) have been used study. In 20 cases the operation was undertaken for the as methods of treating cases of advanced cancer of relief of carcinoma of the breast or prostate but in one the breast (Atkins, Falconer, Hayward, MacLean, (case 4 in Table I) the indication was a rapidly advancing and Schurr, 1966) and the results of comparative diabetic retinopathy. Fourteen of the patients were trials suggest that hypophysectomy is slightly the females and seven were males. The age range was from guest. Protected by copyright. two (Atkins, Fal- 36 to 67 years. The pituitary stalk was approached by more effective of these operations way of a frontal craniotomy, and in some cases the coner, Hayward, MacLean, Schurr, and Armitage, pituitary stalk was clipped before it was transected at as 1960). As an alternative to hypophysectomy neuro- low a level as was technically possible. In the majority surgeons have sometimes used the operation of pitui- of cases an inert and impermeable barrier or 'plate' tary stalk section for relieving such cases and also for (composed usually of dental acrylic resin or tantalum) treating patients with cancer of the prostate (see was placed between the cut ends of the stalk to prevent Schurr, 1966). In addition pituitary stalk section has the re-establishment of the hypothalamo-hypophysial been used for patients with advancing diabetic retino- portal circulation. pathy (Field, Schepens, Sweet, and Appels, 1962; The period of survival after stalk section ranged from 30 hours to 11 months (see Table I). The necropsies on Cullen, Harris, Munro, and Duncan, 1965; Fager, the majority of the cases were undertaken by P.M.D.or Rees, and Bradley, 1966). In assessing the merits J.H.A. but in a few cases the pituitary glands were sent of this particular operation as a means of depressing to us, and in these less detailed information is available. pituitary function, it is important to know the extent At necropsy care was taken to ensure that the plate did to which the structure of the gland is altered and its not become displaced during the removal of the brain, functions impaired. There are few reports in the and a check was made that there was no vascular con- literature on the effects of stalk section on the human nexion between the hypothalamus and the pituitary pituitary (Dandy and Goetsch, 1911; Russell, 1956; gland and that the position of the plate was satisfactory. Daniel, Prichard, and Schurr, 1958; Le Beau and The gland was then carefully dissected from the pituitary Foncin, 1960; Nicolaidis, 1962; Adams, Daniel, fossa.

The pituitaries were fixed whole, either in 10 % formol http://jnnp.bmj.com/ Prichard, and Schurr, 1963), although much saline or in mercuric formol. After fixation two of the information has now been obtained on the effects glands were cut sagittally into five blocks which were then of the operation on several species of animal (Harris, embedded in paraffin wax. Sections (7 ,t or 4 u) were 1950; Barrnett and Greep, 1951; Greep and taken from each block (Fig. la), mounted and stained Barrnett, 1951; Benoit and Assenmacher, 1953; with haematoxylin and eosin, or with iron haematoxylin Campbell and Harris, 1957; Daniel and Prichard, and Van Gieson's mixture, P.A.S., and a few other 1958; Holmes, 1961, 1962; Laszlo, David, and special stains. The remaining 19 glands were bisected Kovics, 1962; Adams, Daniel, and Prichard, in the horizontal plane and after embedding in paraffin 1963a, b, c; 1964a, b, c; 1966a, b; Jacobsohn, 1966). wax were cut serially. The serial sections (usually at on October 1, 2021 by 7 p or 4 jp) were all mounted, and selected sections at We have now had the opportunity of examining various levels were stained as above (Figs. lb and 2-7). the glands of 21 human patients surviving for For technical reasons the two pituitaries which were cut various periods of time after stalk section and in sagittally and two of those cut horizontally were un- this paper we report our findings in these pituitary suitable for measurement, but in the other 17 glands the glands. volumes of the anterior lobe, the infarct or the col- 545 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.29.6.545 on 1 December 1966. Downloaded from 546 J. Hume Adams, Peter M. Daniel, and Marjorie M. L. Prichard TABLE I MEASUREMENTS OF THE ANTERIOR AND POSTERIOR LOBES OF THE PITUITARY GLANDS OF PATIENTS AFTER TRANSECTION OF THE PITUITARY STALK Volumes in mm.3

Anterior Lobe Posterior Lobe Case No. Age Sex Survival after Plate' Whole Surviving Infarct Percentage of Stalk Section Lobe Parenchyma or Scar Lobe Occupied by Infarct or Scar 1 (M.L.) 47 F 30 hours + 216-2 21-8 194-4 90 29-9 2 (J.T.) 57 M 2 days + 127-7 25-4 102 3 80 9.5 3 (A.L.) 61 F 6 days 188-5 40 7 147-8 78 32-2 4 (T.McN) 36 M 7 days + 95-8 83-3 12-5 13 39-4 5 (M.K.) 61 F 11 days - 199-5 102-0 97-5 49 24-1 6 (V.B.) 49 F 15 days + Pituitary gland not measured 7 (R.B.) 37 F 19 days + Pituitary gland not measured 8 (S.) F 3 weeks Pituitary gland not measured 9 (B.L.) 45 F 3j weeks + Pituitary gland not measured 10 (J.McK) 57 M 4 weeks + 116-9 33-6 83-3 71 6-8 11 (G.B.) 52 F 6 weeks - 138-3 99-2 39-1 28 23-3 12 (J.H.) 53 M 21 months + 43-7 17-1 26-6 70 17-6 13 (M.W.) 62 F 3 months + 68-3 52 3 16-0 23 18-1 14 (A.W.) 40 M 3 months + 111-1 31-8 79-3 71 19-9 15 (R.A.) 63 M 3i months + 140-8 108-4 32-4 23 18-3 16 (C.H.) 67 M 4 months + 46-1 39-8 6-3 14 15-2 17 (A.D.) 49 F 4 months + 113-1 91-7 21-4 19 11-0 18 (J.V.) 45 F 6 months + 103-9 67-9 36-0 35 19-5 guest. Protected by copyright. 19 (J.McE) 52 F 11 months + 89-0 87-3 1-7 2 13-3 20 (E.J.) 45 F 11 months - 167-3 164-0 3-3 2 15-9 21 (M.H.) 57 F 11 months + 157-3 79-6 77-7 49 27-2

I+means that an impermeable plate had been inserted at operation between the cut ends of the stalk. -means no plate: a blank means no information available.

lagenous scar within the lobe, and the posterior lobe scopical appearances varied according to the length were measured by a method in which Simpson's rule is of survival. First the cytoplasm of the epithelial applied to serial paraffin sections (Adams, Daniel, cells became swollen and homogeneous and the Prichard, and Venables, 1963). Using a projection nuclei became pyknotic. Then the cell boundaries microscope, the requisite areas were measured on disappeared and the nuclei became fragmented. graph paper, and from these the volumes of the various within the infarct contained parts of the gland could be calculated. Many of the sinusoids It would have been of interest to have been able to red blood cells. Immediate reactive changes were compare the volumes of the anterior and posterior lobes restricted to a mild peripheral infiltration of poly- of the pituitary in these stalk-sectioned patients with the morphonuclear leucocytes, but within about a week volumes of these lobes in the normal subject. However, macrophages could be seen within the infarct and the difficulty in obtaining absolutely normal glands from there was an early mild peripheral fibroblastic patients of the various age groups and the great labour reaction. In the three cases that survived for less and time needed to process and measure such glands has than a week the infarct occupied from 78 % to 90 % http://jnnp.bmj.com/ of the work. deterred us from this extension present of the lobe (cases 1-3). In case 4 (survival seven days) there was a very broad band of surviving parenchyma FINDINGS IN THE PITUITARY GLANDS adjacent to the posterior lobe with the result that the anteriorly placed infarct was unusually small, The anterior lobe (pars distalis) of the pituitary gland occupying only 13 % of the lobe. In case 5 (survival was abnormal in every case. In the patients of short 11 days), where there was also an unusually broad survival (cases 1-5 in Table I) the typical finding was band of surviving epithelial cells posteriorly (Fig. a large infarct occupying the greater part of the 2c), the infarct occupied 49 % of the anterior lobe. on October 1, 2021 by lobe. The necrotic tissue was always much paler than In the cases of longer survival there was a col- the surviving epithelial cells (Fig. 3) which were lagenous scar in the central and anterior parts of the usually restricted to a narrow subcapsular band and anterior lobe (Figs. 4-6). There was a sharp line of to a broader zone adjacent to the posterior lobe demarcation between the scar and the surviving (Fig. 2 a-c). Within the necrotic area the micro- parenchyma. Most of the surviving parenchyma J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.29.6.545 on 1 December 1966. Downloaded from Transection ofthe pituitary stalk in man: anatomical changes in the pituitary glands of21 patients 547

3 weeks S a g t ta was four weeks or longer, the posterior lobe was obviously shrunken and showed a marked increase in the density of its cells (Figs. 10 and 11). The mean volume of the lobe in cases surviving 11 days or less FIG. 1. Tracings of (excluding case 2) was 28-9 mm.3, while in those sections ofhuman surviving from four to 11 months it was 17-2 mm.3,

a pituitary glands i.e., a reduction of just over 40 %. after transection ofthe pituitary DISCUSSION stalk (three weeks andfour months 4 months Horizontal respectively after To understand the consequences of cutting the operation). To show pituitary stalk it is essential to know the anatomy the generalpicture and especially the vascular anatomy of the stalk as seen in (a) the and the pituitary gland (Xuereb, Prichard, and familiar sagittal Daniel, 1954a, b). plane and (b) the horizontalplane ANATOMICAL CONSIDERATIONS In man the pituitary, usedfor most ofour or the hypophysis cerebri, comprises two main parts, material (cf. Figs. 2 the adenohypophysis, consisting of epithelial cells, to 7). Both sections and the neurohypophysis consisting of neural tissue are through the middle portion ofthe (Fig. 12). The adenohypophysis is made up of the

gland, but the anterior lobe (pars distalis) and an upward extension guest. Protected by copyright. sagittal section is of epithelial cells (pars tuberalis) which forms a thin somewhat to one side incomplete sleeve around the stalk and contains 1 cm ofthe midline. prominent thin-walled blood vessels. Beneath this sleeve the stalk consists of neural tissue, part of the neurohypophysis, a term used to include the median Anterior lobe: living parenchyma eminence of the hypothalamus (not prominent in man), the infundibular stem and the posterior lobe E Anterior lobe: scar replacing infacct or infundibular process of the pituitary gland. The infundibular stem consists largely of a nerve tract Posterior lobe running from the hypothalamus to the posterior lobe and can be regarded as having two portions, occupied the posterior part of the anterior lobe the upper infundibular stem which lies in the free (Figs. 1 and 2) although occasionally some epithelial part of the stalk, and the lower infundibular stem cells had survived in other regions (as in Fig. 2e). which lies within the body of the gland (for nomen- In the patients who had lived for from four weeks to clature see Xuereb et al., 1954a). The transection of six months after stalk section (cases 10-18) the this nerve tract at the operation of stalk section leads scar occupied from 14% to 71 % of the lobe. In two to the immediate onset of diabetes insipidus and, of the patients who lived for 11 months (cases 19 later, to the severe atrophy ofthe posterior lobe which and 20) the scars were very small (Fig. 7), but in the is regularly found in these cases (Fig. 10). third (case 21) the scar occupied 4900 of the lobe The blood supply of the pituitary gland and stalk (Fig. 2i). is derived from the inferior and superior hypo- http://jnnp.bmj.com/ The epithelial cells in the living parenchyma in the physial arteries. These are bilateral vessels and posterior part of the anterior lobe were much smaller spring from the internal carotid artery on each side than in the normal pituitary; they also showed some at different levels, the inferior arteries arising below loss of intracytoplasmic granules (Figs. 8 and 9). and the superior arteries above the diaphragma In the patients who died in the first week or so sellae. The posterior lobe is supplied from the after operation the posterior lobe (infundibular inferior hypophysial arteries and the blood traverses process) showed no characteristic abnormalities. In the lobe through a conventional system of arterioles, on October 1, 2021 by case 2 the lobe was abnormally small and the nuclei capillaries, and efferent veins (Fig. 13). The anterior of its cells were closely packed together, but in view lobe, however, has a unique pattern of circulation; of the short survival of this patient (two days) essentially it has no arterial blood supply and all the these abnormalities cannot be attributed to tran- blood which it receives comes from the hypophysial section of the stalk, but must have been due to a portal vessels, so called because they carry blood pre-existing degenerative process. When survival from a first (or primary) to a second capillary bed. J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.29.6.545 on 1 December 1966. Downloaded from 548 J. Hume Adams, Peter M. Daniel, and Marjorie M. L. Prichard

2 days 6 days 11 days

a b c 3 weeks 6 weeks 3 months guest. Protected by copyright.

d e f 3 months 6 months 11 months

9 h i http://jnnp.bmj.com/

1 -cm

lobe: Anterior 3..:..T Anterior lobe: Posterior lobe

living parenchyma infarct or scar on October 1, 2021 by

FIG. 2. Pituitary glands ofnine patients who survivedfor variousperiods after transection ofthepituitary stalk, showing the areas ofliving and infarcted (or scarred) anterior lobe tissuefound in the mid-horizontalplane ofthegland. The extent of the surviving tissue is not constant, but in each case the parenchyma adjacent to the posterior lobe has survived. This is because the blood flow through the short portal vessels (see Fig. 13) is not interrupted by transection of the stalk (tracings of sections all projected at the same magnification). J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.29.6.545 on 1 December 1966. Downloaded from

Transection ofthe pituitary stalk in man: anatomical changes in the pituitary glands of21 patients 549

,/i S

i[(.. .tJ. FIG. 4. Case 10. Horizontal section through mid-part ofpituitary gland of a patient who diedfour weeks afterpituitary

stalk section. (Glands ofthis shape are not guest. Protected by copyright. uncommon.) The large scar (pale, S) indicates the very large amount ofanterior lobe tissue which had been destroyed. The parenchyma (P) adjacent to the posterior lobe has survived as usual, although the epithelial cells here are abnormally small (see Fig. 8). The posterior lobe (PL) is greatly shrunken. At A is a small adenoma. Haematoxylin and eosin x 6.

PL r|.

-.11. I b. FIG. 3. Case 1. Horizontal section at two levels (a and

b) through the pituitary gland ofpatient who died 30 hours http://jnnp.bmj.com/ after transection of the pituitary stalk. The posterior lobe (PL) andpart ofthe lower infundibular stem (LIS) are seen above. There is massive necrosis of the anterior lobe, the infarct (pale area, I) involving 90% of the lobe. The small Iq *t areas of surviving parenchyma (P) seen posteriorly (dark) g,.2. have been preserved by the unimpaired bloodflow through the shortportal vessels which arise in the lower infundibular stem (see Figs. 12 and 13). Haematoxylin and eosin x 4-5. on October 1, 2021 by FIG. 5. Case 12. Pituitary gland of a patient who survived for two and half months after operation. The distribution of the scar (S) and the livingparenchyma (P) in the anterior lobe is typical. PL, posterior lobe. Haematoxylin and eosin x 6. J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.29.6.545 on 1 December 1966. Downloaded from

i PL

I I(. K, FIG. 6. Case 17. Pituitary gland of a patient who livedforfour months after operation. In this case there is a considerable amount of living anterior lobe parenchyma (P) which suggests that either the initial infarct had been small or that some regeneration ofparenchymal cells had occurred. 5, scar in anterior lobe; PL, posterior lobe. Haematoxylin and eosin x 6. FIG. 7. (Case 20). Pituitary gland ofa patient who died 1I months afterpituitary stalk section. Apartfrom smallpatches ofscar tissue (at arrows) the whole ofthe anterior lobe consists oflivingparenchynma (P). PL posterior lobe. Haematoxy- lin and eosin x 6. guest. Protected by copyright.

f ' A $~~~~~~~~~1 U,tv 3 u .t

.,b'X i~~~~~~~~~~~~~~~~~~~~~~~~~~~~q

t tt t ~.9is / )w,, ^o.@:C::,; ...... ,, #..:t'; WW http://jnnp.bmj.com/

V:ZhW ''aW+tWil41 e~~~~~~~ ;*~~~ ~ ~ rt:i ' J ) ~~~~~~~~~~~~~~~~~~~~..t on October 1, 2021 by S. A114. 93 FIG. 8. Pituitary four weeks after stalk section (case 10). Living anterior lobe cells in area preserved by the short portal vessels. Note the small size of the cells compared with the normal (Fig. 9). Haematoxylin and eosin x 485. FIG. 9. Normal human pituitary. Same area as in Fig. 8 to show size and appearance ofnormal anterior lobe cells in this region. Haematoxylin and eosin x 485. Transection ofthe pituitary stalk in man: anatomical changes in the pituitary glands of21 patients 551 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.29.6.545 on 1 December 1966. Downloaded from

J i iA*,p+,tV.. ,+'Nw,f *''n

t~~~- wt. X 9,gt. 4w >; .t.'> S~~~

<;l..~~~~4~~~~ ...A 4~~~~~~~~~~~~~~~~~~~~~~~1iA AV

t,Ut. .li guest. Protected by copyright. AV`~ ~ ~

FIG. 10. Pituitary four weeks after stalk section (case JO). Closely packed cells in the greatly shrunken posterior lobe. Compare with the normal appearance in Fig. 11. Haemtatoxylin and eosin x 350. FIG. 11. Normal human pituitary. Cells in posterior lobe, for comparison with Fig.Hae.atoxylinJO. and eosin x 350. McConnell (1953) and Stanfield (1960) believe that any arterial contribution that may be provided in the anterior lobe of the human pituitary is supplied this way must be small and be confined to a very by some small arteries as well as by portal vessels, limited region of the anterior lobe. but in our view such an arterial supply when it The hypophysial portal vessels which provide the exists at all is so small as to be insignificaiat. True, anterior lobe with its blood supply are disposed in there is an artery which enters the anterior lobe and two distinct groups: the long portal vessels which run passes through it surrounded by a thick sheath of down the stalk, and the short portal vessels which fibrous tissue, the artery of the trabecula (Fig. 13), lie below the level of the diaphragma sellae. Both but although this vessel, which links the superior groups, on entering the anterior lobe, break up into http://jnnp.bmj.com/ and inferior hypophysial arterial systems, was a sinusoids, and this network of vessels, which extends prominent feature in many of our injected pre- throughout the lobe, forms the second capillary bed parations, we never found any branches distributed of the portal system. Each group, the long and the to the parenchyma. At one point, namely, in the short portal vessels, feeds a well-defined territory 'fibrovascular tuft' (Xuereb et al., 1954a, Fig. 3), in the anterior lobe, the long portal vessels supplying the artery of the trabecula does give off a few very a large area in the anterior and central part of the small, short twigs; these do not, however, as a rule lobe, while the short portal vessels supply a much on October 1, 2021 by pass beyond the fibrous tissue of the trabecula, smaller area adjacent to the posterior lobe. The first and though it is possible that in some instances a capillary beds which these two groups of portal little blood may spill over from these vessels into vessels drain are both situated in neural tissue, but sinusoids of the immediately adjacent parenchyma, lie at different levels in the neurohypophysis (Fig. J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.29.6.545 on 1 December 1966. Downloaded from 552 J. Hume Adams, Peter M. Daniel, and Marjorie M. L. Prichard Hypothalamus 13). The long portal vessels drain a capillary bed in the median eminence and the upper infundibular stem, a bed supplied by the superior hypophysial arteries. The short portal vessels drain a bed in the lower infundibular stem and this bed is supplied mainly by a branch of the inferior hypophysial u s . artery but partly also by the artery of the trabecula. \ S t a k cut here ______::::EFFECTS OF TRANSECTION OF THE STALK When the :-: . .... :.\ ... pituitary stalk is cut immediately above the level of the diaphragma sellae (as in Fig. 12) the transection should be sufficiently low to sever all the long portal Posterior Anterior vessels, but the short portal vessels and their supply- lIo.be .. _ lobe _ing vessels will remain intact (see Fig. 13). It would be expected, therefore, that although an infarct would develop in the anterior lobe, part of the lobe would survive, namely, the territory supplied by the short portal vessels. This is indeed the case, a con- FIG. 12. Diagram ofhumanpituitary gland in mid-sagittal stant feature of the pituitaries reported here being the plane. To show the relations of the glandular (black) and survival of the posterior region of the anterior lobe neur,al (stippled) components. The level ofa low transection which is adjacent to the lower infundibular stem and of thte stalk is also indicated. UIS, upper infundibular stem; the posterior lobe (see Figs. 1-6). A superficial rim,

lower infundibular stem. guest. Protected by copyright. LIS, only a few cells thick, along the other borders of the lobe also regularly survives, preserved presumably by diffusion from the adjacent highly vascular dura mater or by intermittent venous reflux. It is important to know how much of the lobe is destroyed by infarction, since the extent of the Hypothalamus infarct will be one factor in determining the degree to which anterior pituitary function is depressed. ",*8.4 ~~~~~Muchdepends on the level at which the stalk has SHA been cut, since if the transection was not at a really low level some of the long portal vessels will have AT\Mwe\ escaped transection and the infarct produced will <4\\AT not be of the maximum possible size. In our experi- ments on animals with a pituitary blood supply Stalk ') \A9\\\.< LPV v similar to that of man, a low transection of the \\2aK\ sstalk was usually achieved and the results in goats, sheep, and monkeys were on the whole very con- sistent, the immediate infarct produced involving Posterior w i X ; <'te 1 80-95 % of the lobe (Adams et al., 1963a, b, c, 1966b). lobe Very occasionally an unexpectedly large proportion SPV/1Anterior of /K:the lobe::: escaped infarction, and in the rat the http://jnnp.bmj.com/ tJJ/:-: -}figures were more variable than in the other species (Adams et al., 1963a). We attribute these differences not only to possible variations in the level of the IHA <- transection but also to variations in the size of the territory supplied by the short portal vessels in FIG. 13. Diagram of human pituitary gland showing its individual animals. bloo,d supply. When the stalk is cut at a low level, as in Fig. From the present series of cases it is not easy to

12, eall the long portal vessels (LPV) are severed, and this assess how much of the anterior lobe undergoes on October 1, 2021 by prodfuces the largest possible infarct in the anterior lobe. EvenF with a low cut, however, the short portal vessels infarction when the pituitary stalk is cut in man, (SPIV) and their afferent supply remain intact, so that some because it is only in the first few days after operation part of the lobe always survives (see Fig. 2). IHA, SHA, that a true estimate of this can be made. The im- infer'ior and superior hypophysial arteries; AT, artery ofthe mediate infarct is rapidly replaced by fibrous tissue; trab ecula; V, venous sinus. the latter then shrinks, so that the size of the final J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.29.6.545 on 1 December 1966. Downloaded from

Transection ofthe pituitary stalk in man: anatomical changes in the pituitary glands of21 patients 553 scar does not represent the size of the original infarct (see Adams et al., 1963a, b; 1964a; 1966a). Only four patients of the present series died within a week of operation. Of these, three had infarcts involving 78-90% of the anterior lobe, but in the fourth the infarct was small, involving only 13% of the lobe. The first report in the literature of an acute infarct after stalk section was that of Cushing's case in which the stalk was accidentally cut during an intracranial operation (Dandy and Goetsch, 1911), and this infarct was clearly of very large size. Equally extensive infarcts have been reported by Russell (1956) in three cases and in one case by Nicolaidis (1962), who, however, also reported a second case in which the acute infarct involved only 45 % of the anterior lobe. Very large infarcts have also been seen in cases in which the stalk had been ruptured as a result of head injury (Daniel, Prichard, and Treip, 1959). In the group of longer surviving cases the large IHA size of the scar seen in some of the glands indicates FIG. 14. Diagram to show the pathways by which it is that the original infarct had been very extensive. thought that neurohumours, elaborated by cells in the hypo- guest. Protected by copyright. However, in others (particularly cases 19 and 20) the thalamus, are transmitted to the pituitary gland. Note in scar was small, and this suggests that the initial particular the axons of hypothalantic cells (H) which infarct, as in one of the acute cases, had also been terminate on the primary capillary beds (P) which feed the small. It would seem, therefore, that although in portal vessels. Here the neurohumours enter the blood- theory pituitary stalk section in man should cause stream, to be carried through the long and short portal massive destruction of the anterior lobe, in practice vessels (LPV, SPV) to the parenchymal cells (C) in the anterior lobe. One axon is shown ending on capillaries this is not invariably the case. (Cap) of the posterior lobe. IHA, SHA, inferior and Several factors could explain why in some instances superior hypophysial arteries. the infarct or scar produced is smaller than expected. First, the stalk may not have been cut at a sufficiently The amount of living parenchyma that remains low level to ensure that all the long portal vessels after stalk section should not, however, be taken as are transected. Secondly, the gland may have been a true indication of residual pituitary function since, one in which there was some slight variation from if an effective barrier was inserted at operation, this the usual pattern of the pituitary blood supply. parenchyma no longer has any connexions with the For instance, the territory supplied by the short hypothalamus. It is now generally believed that the portal vessels may have been unusually large, or the functional activity of the epithelial cells of the artery of the trabecula may, exceptionally, have anterior lobe is largely controlled by the hypo- contributed a small supply to the immediately thalamus (Harris, 1955, 1960). The nerve tract which surrounding parenchyma (as suggested in Fig. runs down the pituitary stalk consists of the axons

2c and e), or an atypical vessel may have penetrated of hypothalamic cells, particularly those of the http://jnnp.bmj.com/ the anterior lobe from the dura mater, thus pre- supraoptic and paraventricular nuclei. A large serving a localized area of superficial parenchyma number of these axons pass to the posterior lobe broader than the usual narrow rim (as suggested and are concerned with the production of posterior in Fig. 2b and c). Thirdly, one cannot exclude the lobe hormones, but it is believed that many of them possibility that in patients who survive for several terminate on the loops of the primary capillary beds months some regeneration of anterior lobe paren- which feed the portal vessels (Rinne, 1960). The chyma may have taken place. In animals we found theory is that neurohumours are released from the

that although normally regeneration of parenchymal terminations of these latter axons into the primary on October 1, 2021 by cells did not occur to any appreciable extent a capillary beds, and that these neurohumours are then substantial regrowth of these cells might occur under carried in the blood through the portal vessels to the certain conditions (Adams et al., 1963b). Nicolaidis epithelial cells, causing them to secrete (or release) (1962) believed that in man regeneration of paren- their hormones (Fig. 14). When the pituitary stalk chymal cells was a factor of some importance after is cut all the nerve fibres which run down it are pituitary stalk section. severed and the short portal vessels are therefore

6 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.29.6.545 on 1 December 1966. Downloaded from

554 J. Hume Adams, Peter M. Daniel, and Marjorie M. L. Prichard now draining a denervated primary capillary bed removed. It might be expected that isolation from (see Adams et al., 1964b, 1966b). Consequently the hypothalamus would depress all anterior pituitary the blood that these vessels carry to the epithelial function, but there is some evidence that in these cells in the surviving areas of the anterior lobe circumstances the output of lactogenic hormone no longer contains the neurohumours on which these may actually be increased. For example, Ehni and cells normally depend for their functional activity. Eckles (1959) and Dugger, Van Wyk, and Newsome To ensure that the pituitary gland is permanently (1962) found that milk might be secreted by non- separated from direct hypothalamic influence it is pregnant women after stalk section. necessary at operation to insert a barrier of some In this paper we have concentrated on the ana- inert and impermeable material (e.g., dental acrylic tomical changes which occur in the human pituitary resin) between the cut ends of the stalk to prevent gland as a result of stalk section. It is important to the possible regeneration of the long portal vessels know what these changes are if one is to understand (Harris and Jacobsohn, 1952; Jacobsohn, 1954) and the rationale for the operation. However, the also to obstruct the growth of new nerve fibres important question remains as to whether stalk (Beck and Daniel, 1961) which might re-innervate section (which is not as radical an operation as the primary capillary bed of the short portal vessels. hypophysectomy and therefore has some practical The accurate placing of a barrier of adequate size is advantages) achieves the object for which it is per- of vital importance. formed. This question is beyond the scope of the In animals, when a permanent isolation of the present paper, but it is known that the beneficial gland from the hypothalamus has been achieved, the effects of stalk section include a dramatic relief of cells in the anterior lobe which survive show the pain in patients with carcinoma, regression of

effect of their isolation within a few weeks of metastases in some patients (see Schurr, 1966),guest. Protected by copyright. operation by becoming much smaller than normal and a frequently striking relief of symptoms in (Adams et al., 1964a; 1966b) and there is a marked patients suffering from diabetic retinopathy (Field loss of secretory granules in their cytoplasm (Daniel, et al., 1962; Cullen et al., 1965). In rats with experi- Duchen, and Prichard, 1964a, b). Unfortunately in mentally induced mammary tumours we found that most of our human cases necropsy had been delayed stalk section was highly successful in causing and autolytic changes made an assessment of cell regression of tumours in some instances, but that size difficult, but in the one case (surviving for four hypophysectomy produced more consistently fav- weeks) where accurate assessment was possible the ourable results (Daniel and Prichard, 1963). This small size of the surviving parenchymal cells made may or may not be the case in man. The question a striking contrast with the normal (Figs. 8 and 9). could probably be answered by a controlled trial On the whole, however, the loss of secretory granules of the two operations in the human subject. in the surviving cells does not seem to be as marked in man as it is in animals. We are surprised that in SUMMARY their longer surviving human cases Ehni and Eckles (1959) and Nicolaidis (1962) noted no The pituitary glands of 21 patients surviving for abnormality in the surviving cells, but they may also periods ranging from 30 hours to 11 months after have been faced with the difficulty of autolytic transection of the pituitary stalk were examined. changes confusing the picture. In 17 cases the volumes of the anterior lobe, the It may well be, therefore, that in man the actual infarct or scar within it, and the posterior lobe were size of the infarct is of only limited importance as an measured. In every case the anterior lobe was indication of depression of pituitary function. In abnormal. Infarcts or scars were present, usually of http://jnnp.bmj.com/ animals it has been found that once the anterior considerable size, but some parenchyma, in a lobe has been effectively and permanently isolated characteristic situation in the posterior part of the from the hypothalamus the functional activity of lobe, had always escaped destruction. The epithelial the cells which survive is greatly impaired (Daniel cells in this surviving anterior lobe tissue were and Prichard, 1964; Cowie, Daniel, Prichard, and abnormally small. Tindal, 1963; Cowie, Daniel, Knaggs, Prichard, The information provided by this series of glands and Tindal, 1964; Daniel et al., 1964a, c). Whether suggests that although pituitary stalk section will this is also true in man, however, is not known, usually cause extensive infarction of the anterior on October 1, 2021 by for although various tests of pituitary function have lobe, this result will not be invariable. been made on patients after stalk section the results As a result of the operation the posterior lobe of are vitiated by the hormone replacement therapy the gland shrank and showed an increased density that has to be given, and often by the fact that the of nuclei. patients have also had other endocrine glands The findings are explained in the light of the J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.29.6.545 on 1 December 1966. Downloaded from Transection of the pituitary stalk in man: anatomical changes in the pituitary glands of21 patients 555 and the vascular arrangements of the (1964c). Some effects of pituitary stalk section and of anatomy hypophysectomy on the endocrine organs and growth of rats. pituitary stalk and gland. Ibid., 49, 243-257. work reported here was assisted by grants to P.M.D. Daniel, P. M., and Prichard, M. M. L. (1958). The effects of pituitary The stalk section in the goat. Amer. J. Path., 34, 433-469. from the Nuffield Foundation, the British Empire -,- (1963). The response of experimentally induced mammary Cancer Campaign for Research, and the Research Fund tumours in rats to hypophysectomy and to pituitary stalk of the Bethlem Royal and Maudsley Hospitals. We are section. Brit. J. Cancer, 17, 446-453. - (1964). Observations on the metabolic disturbances, grateful to Mrs. J. Storms, Mr. P. H. Schurr, and Dr. growth and behaviour of young goats after transection of the Sabina Strich for their help. pituitary stalk. Acta endocr. (Kbh.), 45, 84-98. -, -, and Schurr, P. H. (1958). Extent of the infarct in the REFERENCES anterior lobe of the human pituitary gland after stalk section. Adams, J. H., Daniel, P. M., and Prichard, M. M. L. (1963a). The Lancet, 1, 1101-1103. volumes of pars distalis, pars intermedia and infundibular and Treip, C. S. (1959). Traumatic infarction of the process of the pituitary gland of the rat, with special reference anterior lobe of the pituitary gland. Ibid., 2, 927-931. to the effect ofstalk section. Quart. J. exp. Physiol., 48, 217-234. Dugger, G. S., Van Wyk, J. J., and Newsome, J. F. (1962). The effect ,- (1963b). The effect of stalk section on the volume of of pituitary-stalk section on thyroid function and gonadotropic- the pituitary gland of the sheep. Acta endocr. (Kbh)., 43, hormone excretion in women with mammary carcinoma. J. suppl. 81. Neurosurg., 19, 589-593. ,-(1963c). Volume of the infarct in the anterior lobe Ehni, G., and Eckles, N. E. (1959). Interruption of the pituitary stalk of the monkey's pituitary gland shortly after stalk section. in the patient with mammary cancer. Ibid., 16, 628-652. Nature (Lond.), 198, 1205-1206. Fager, C. A., Rees, S. B., and Bradley, R. F. (1966). Surgical ablation , (1964a). Transection of the pituitary stalk in the of the pituitary in the treatment of diabetic retinopathy. Ibid., goat, and its effect on the volume of the pituitary gland. J. Path. 24, 727-734. Bact., 87, 1-14. Field, R. A., Schepens, C. L., Sweet, W. H., and Appels, A. (1962). --, ,--(1964b). Distribution of hypophysial portal blood The effect of hypophyseal stalk section on advancing diabetic in the anterior lobe of the pituitary gland. Endocrinology, 75, retinopathy. Report of thirteen cases. Diabetes, 11, 465-469. 120-126. Greep, R. O., and Barrnett, R. J. (1951). The effect of pituitary stalk- , (1964c). Some effects of transection of the pituitary section on the reproductive organs of female rats. Endo- stalk. Brit. med. J., 2, 1619-1625. crinology, 49, 172-184. (1966a). The long-term effect of transection of the Harris, G. W. (1950). Oestrous rhythm. Pseudopregnancy and the guest. Protected by copyright. pituitary stalk on the volume of the pituitary gland of the adult pituitary stalk in the rat. J. Physiol. (Lond.), 11, 347-360. goat. Acta endocr. (Kbh.), 51, 377-390. (1955). Neural Control of the Pituitary Gland. Arnold, London. -, , (1966b). Observations on the portal circulation of -(1960). Central control of pituitary secretion. In The Handbook the pituitary gland. Neuroendocrinology, 1, 193-213. of Physiology, Section 1: Neurophysiology, vol. 2, pp. 1007- --, , -, and Schurr, P. H. (1963). The volume of the infarct 1038, edited by J. Field, H. W. Magoun, and V. E. Hall. in pars distalis of a human pituitary gland, 30 hours after American Physiological Society, Washington, D.C. transection of the pituitary stalk. J. Physiol. (Lond.), 166, -, and Jacobsohn, D. (1952). Functional grafts of the anterior 39 P-41 P. pituitary gland. Proc. Roy. Soc. B., 139, 263-276. -,-, and Venables, P. H. (1963). A method for measuring Holmes, R. L. (1961). Changes in the pituitary gland of the ferret the volume of the pituitary gland. Ibid., 166, 37 P-39 P. following stalk section. J. Endocr., 22, 7-14. Atkins, H. J. B., Falconer, M. A., Hayward, J. L., MacLean, K. S., (1962). The pituitary gland of normal and stalk-sectioned and Schurr, P. H. (1966). The timing of adrenalectomy and of monkeys with particular reference to the pars intermedia. hypophysectomy in the treatment of advanced breast cancer. Ibid., 24, 53-58. Lancet, 1, 827-830. Jacobsohn, D. (1954). Regeneration of hypophysial portal vessels and ,11 - , -,-, and Armitage P. (1960). Adrenalectomy grafts of anterior pituitary glands in rabbits. Acta endocr. and hypophysectomy for advanced cancer of the breast. Ibid., (Kbh.), 17, 187-197. 1, 1148-1153. (1966). The techniques and effects of hypophysectomy, pituitary Barrnett, R. J., and Greep, R. 0. (1951). The pituitary gonadotropic stalk section and pituitary transplantation in experimental activity ofstalk-sectioned male rats. Endocrinology, 49, 337-348. animals. In The Pituitary Gland, vol. 2, pp. 1-21, edited by Beck, E., and Daniel, P. M. (1961). Degeneration and regeneration in G. W. Harris and B. T. Donovan, Butterworths, London. the hypothalamus. In Cytology of Nervous Tissue, Proc. Laszlo, F. A., David, M. A., and Kovacs, K. (1962). Changes in the Anat. Soc. Gt. Britain and Ireland, pp. 60-63. Taylor and Francis, pituitary volume of rats following destruction of the pituitary London. stalk. Med. exp. (Basel), 7, 368-378. Benoit, J., and Assenmacher, I. (1953). Rapport entre la stimulation Le Beau, J., and Foncin, J. F. (1960). Anatomical study of anterior sexuelle pr6hypophysaire et la neuros6cr6tion chez l'oiseau. pituitary remnants after (1) stalk section, (2) hypophysectomy. Arch. Anat. micr., 42, 334-386. Acta psychiat. scand., 35, 13-26. Campbell, H. J., and Harris, G. W. (1957). The volume ofthe pituitary Luft, R., and Olivecrona, H. (1953). Experiences with hypophysectomy

and median eminence in stalk-sectioned rabbits. J. Physiol. in man. J. Neurosurg., 10, 301-316. http://jnnp.bmj.com/ (Lond.), 136, 333-343. McConnell, E. M. (1953). The arterial blood supply of the human Cowie, A. T., Daniel, P. M., Prichard, M. M. L., and Tindal, J. S. hypophysis cerebri. Anat. Rec., 115, 175-203. (1963). Hypophysectomy in pregnant goats, and section of the Nicolaidis, S. (1962). Section de la tige pituitaire chez l'homme. Le pituitary stalk in pregnant goats and sheep. J. Endocr., 28, Tendre, Caen (Calvados). 93-102. Rinne, U. K. (1960). Neurosecretory material around the hypo- -, Knaggs, G. S., Prichard, M. M. L., and Tindal, J. S. physial portal vessels in the median eminence of the rat. Acta (1964). Lactation in the goat after section of the pituitary endocr. (Kbh.), 35, suppl. 57. stalk. Ibid., 28, 253-265. Russell, D. S. (1956). Effects of dividing the pituitary stalk in man. Cullen, J. F., Harris, P., Munro, J. F., and Duncan, L. J. P. (1965). Lancet, 1, 466-468. Pituitary stalk section for advancing diabetic retinopathy. Schurr, P. H. (1966). Techniques and effects of hypophysectomy, Brit. J. Ophthal., 49, 393-404. pituitary stalk section, and pituitary transplantation in man. Dandy, W. E., and Goetsch, E. (1911). The blood supply of the In The Pituitary Gland, vol. 2, pp. 22-48, edited by G. W. on October 1, 2021 by pituitary body. Amer. J. Anat., 11, 137-150. Harrjs and B. T. Donovan. Butterworths, London. Daniel, P. M., Duchen, L. W., and Prichard, M. M. L. (1964a). Stanfield, J. P. (1960). The blood supply of the human pituitary gland. The cytology of the pituitary gland of the rhesus monkey: J. Anat. (Lond.), 94, 257-273. changes in the gland and its target organs after section of the Xuereb, G. P., Prichard, M. M. L., and Daniel, P. M. (1954a). The pituitary stalk. J. Path. Bact., 87, 385-393. arterial supply and venous drainage of the human hypophysis - (1964b). The effect of transection of the pituitary cerebri. Quart. J. exp. Physiol., 39, 199-217. stalk on the cytology of the pituitary gland of the rat. Quart. (1954b). The hypophysial portal system of vessels J. exp. Physiol., 49, 235-242. in man. Ibid., 39, 219-230.