Impaired Androgen Production in Female Adolescents and Young Adults After Total Body Irradiation Prior to BMT in Childhood

Home , Androgen, Androstenedione

Impaired androgen production in female adolescents and young adults after total body irradiation prior to BMT in childhood

L Hovi1, P Tapanainen2, UM Saarinen-Pihkala1 and MA Siimes1

1Children’s Hospital, University of Helsinki, Helsinki; and 2Children’s Hospital, University of Oulu, Oulu, Finland

Summary: The ovary is particularly sensitive to the effects of irradiation. A radiation dose of less than 4 Gy may cause Pubertal development and androgen production were the death of half of the oocytes.1 Ovarian failure has fre- evaluated 1–10 years after bone marrow transplantation quently been noted among girls and women with hemato- (BMT) in 15 females aged 14–23 (mean 17) years. logic malignancies after total body irradiation (TBI) Before BMT, these patients had received combination included in the preparative regimens for bone marrow chemotherapy for hematologic malignancy, and all had transplantation (BMT).2–4 Accordingly, women who have had a transplant program including total body been subjected to allogeneic BMT suffer from secondary irradiation (TBI). Of the nine patients who were pre- amenorrhea and infertility. In addition, the majority of these menarcheal at BMT, two had subsequently experienced women may have vasomotor and genitourinary symptoms spontaneous menarche at 11.5 and 13.3 years of age. Six and also difficulties with intercourse.5,6 These symptoms were post-menarcheal, but became amenorrheic after have been reduced by hormonal replacement therapy with BMT. Menstruation subsequently started spontaneously estrogen and progesterone. However, late ovarian recovery in one of them 6 years after BMT. At the time of the may occur in some young survivors several years after study, three patients were early to mid-pubertal and 12 BMT.7 The actuarial chance of having spontaneous men- late pubertal or post-pubertal. Twelve patients were strual periods 10 years post-BMT may be as high as 100% receiving sex steroid substitution therapy. Serum con- if the patient was transplanted before the age of 18 years.6 centrations of testosterone, androstenedione, dehydro- There are no good ways to predict ovarian recovery. More epiandrosterone (DHEA) and DHEA sulfate (DHEAS) accurate methods would allow better advice regarding hor- were determined. Androgen levels of late pubertal and mone substitution, contraception and family planning. In post-pubertal transplanted patients were compared contrast to BMT, conventional chemotherapy for leukemia with 19 post-menarcheal patients aged 14–21 (mean 17) with or without prophylactic central nervous system years who had been treated for hematologic malignancy irradiation rarely results in any evidence of ovarian failure.8 with conventional chemotherapy. Testosterone levels of The diagnosis of ovarian dysfunction is usually based on 52 healthy post-menarcheal females aged 14–29 (mean elevated gonadotropin and low estradiol levels. The ovarian 19) years were measured as controls. Androgen levels theca cells produce androgens, mainly androstenedione and of the BMT patients were lower than those of the con- testosterone, and these are converted into estrogen by the ventionally treated patients. Differences in testosterone, enzyme aromatase which is highly active in the ovarian androstenedione and DHEA levels were significant. granulosa cells. Androgens are also important for the devel- Three spontaneously menstruating BMT patients had opment of normal female secondary sexual characteristics normal androgen levels. Testosterone levels of the con- and for the activity of striated muscle. In addition, they may ventionally treated patients and healthy controls were influence libido and the quality of sex life. All these fea- similar. Subnormal androgen production might be one tures are often impaired in females after TBI.5,9 In addition factor behind the problems in pubertal development to the ovary, androgens are also produced by the adrenal and sex life experienced by females after BMT. The use cortex which contribute most of the circulating dehydro- of these hormone levels for follow-up purposes and the epiandrosterone (DHEA) and especially DHEA sulfate potential value of androgen replacement therapy in (DHEAS) in females. Only minute amounts of testosterone females after TBI merit further study. and androstenedione are produced by the adrenals. There Keywords: androgens; testosterone; androstenedione; are hardly any reports on adrenal and ovarian androgen pro- dehydroepiandrosterone; females; bone marrow transplan- duction after TBI. Hinterberger-Fischer et al10 reported tation seven adult female patients with normal testosterone levels after TBI. Plasma cortisol levels remain within normal limits after TBI.5,11 We studied pubertal development and androgen production of young females surviving childhood leukemia and report here impaired androgen production in patients Correspondence: Dr L Hovi, Children’s Hospital, University of Helsinki, after a TBI-containing transplant program as compared FIN-00290 Helsinki, Finland with that in patients who had received conventional chemo- Received 14 February 1997; accepted 2 June 1997 therapy. Androgen production in transplanted females L Hovi et al 562 Patients and methods diagnoses were ALL (n = 12), AML (n = 4) and non-Hodg- kin’s lymphoma (n = 3). The therapeutic regimens used The study included 34 female patients over 14 years of age, were the standard Nordic or CCG-derived protocols for leu- treated and followed-up at the Children’s Hospital, Univer- kemia and lymphoma.12 Of these patients, nine had sity of Helsinki, for hematologic malignancy from 1980 received prophylactic central nervous system irradiation of to 1994. 24 Gy and two mediastinal irradiation of 24 Gy. The malig- nant disease had been diagnosed between 0.3 and 14 years Transplanted patients (mean 7.0) of age. The post-therapy interval was 1–14 years (mean 7). All 19 patients had experienced a normal men- Fifteen patients seen at 14–23 (mean 17) years had been arche at the ages of 11–15.2 years (mean 12.8). They con- transplanted at ages between 4 and 17 (mean 12) years and tinued to have normal menstrual periods. Three were using received TBI of 10–12 Gy in 1–6 fractions as part of the contraceptive pills of combined type at the time of this conditioning regimen. Primary diagnoses included acute study. lymphoblastic leukemia (ALL) (n = 6), acute myeloblastic leukemia (AML) (n = 6) and chronic myelogenous leuke- Reference subjects mia (CML) (n = 3). Before BMT, all patients had received combination chemotherapy. As conditioning therapy As normal healthy controls, we studied 52 females aged patients received high-dose chemotherapy with either 14–29 (mean 19) years. They were all menstruating nor- cyclophosphamide (n = 12) or cytosine arabinoside (n = 3) mally. Eighteen were using contraceptive pills of the com- prior to TBI and BMT. The post-BMT interval ranged from bined type. Of the androgens, only testosterone was meas- 1 to 10.5 (mean 5.1) years. Of the nine patients who were ured in these subjects. pre-menarcheal at BMT, two have subsequently experienced spontaneous menarche at the ages of 11.5 and 13.3 Methods years, 7.3 and 2.0 years after BMT, respectively. The other pre-menarcheally transplanted patients received sex hor- Pubertal development and ovarian function were assessed mone substitution for delayed pubertal development. Six both clinically and biochemically 1–14 years after discon- patients were post-menarcheal at BMT, but became tinuation of therapy. Pubic hair and breast development amenorrheic. Menstruation subsequently started spon- were staged according to Tanner. In addition to history and taneously in one of them 6 years after BMT; the others physical examination, serum concentrations of testosterone, received cyclic sex hormone therapy. At the time of this androstenedione, DHEA and DHEAS were determined in study, three of the transplanted patients were in early to the same laboratory, using specific radioimmunoassays. midpuberty (Tanner stages II–III) and 12 were in late Testosterone and androstenedione were quantified by RIA puberty (Tanner stages IV–V) (Table 1). after extraction with ethyl ether-ethyl acetate (70:30) and separation on a lipophilic gel (Lipidex 5000, Packard Instruments, BV Chemical Operations, Groningen, The Conventionally treated patients Netherlands) as described by Apter et al,13 except that the We also studied 19 patients (aged 14 to 21 years, mean 17 antiserum for androstenedione was obtained from Sigma. years) who had received conventional chemotherapy. Their DHEA was quantified by RIA after extraction with pet-

Table 1 Characteristics of the transplanted patients

Patient No. Diagnosis Age at Post BMT Age at Tanner Menarche/ﬁrst Sex hormonal BMT follow-up study stage spontaneous replacement (years) (years) (years) menstruation after therapy BMT (years)

Pre-menarcheal at BMT 1 ANLL 5.2 8.8 14.0 2 + 2 ALL 12.5 1.6 14.1 2 + 3 CML 11.5 2.6 14.1 4 13.5 − 4 ANLL 4.2 10.0 14.2 5 11.5 − 5 ANLL 12.4 2.1 14.5 4 + 6 ALL 11.8 3.1 14.9 3 + 7 ANLL 9.0 7.0 16.0 4 + 8 ANLL 13.5 6.0 19.5 4 + 9 ALL 11.1 10.5 21.6 5 +

Post-menarcheal at BMT 10 ALL 16.4 1.1 17.5 4 + 11 CML 16.3 1.5 17.8 5 + 12 CML 17.7 1.0 18.7 5 + 13 ANLL 14.0 5.2 19.2 5 + 14 ALL 14.5 8.2 22.7 5 20 − 15 ALL 16.3 6.7 23.0 5 + Androgen production in transplanted females L Hovi et al 563

1.6 8 20 10

8 1.2 6 15

mol/l 6 nmol/l µ , 0.8 4 10 DHEA 4 DHEAS, Testosterone, nmol/l Testosterone,

0.4 Androstenedione, nmol/l 2 5 2

0 0 0 0 +– + – + – + – TBI Figure 1 Androgen levels in female survivors of leukemia. Androgen levels in female late pubertal or postpubertal leukemia patients treated with total body irradiation (TBI+)(n=12) or with conventional chemotherapy (TBI−)(n=19). Transplanted individuals with spontaneous menstruation are indicated by . The testosterone level (mean Ϯ s.d.) of healthy control subjects (n = 52) is shown by the dotted area. roleum ether, as described by Hammond et al.14 The anti- treated patients (P = 0.002 and 0.012, respectively) serum was obtained from Sigma, St Louis, MO, USA. (Table 2). DHEAS was measured by RIA using Coat-A-Count The DHEAS levels of the transplanted patients tended to

DHEA-SO4 (Diagnostic Products Corporation, Los Ang- be lower than those of the conventionally treated patients, eles, CA, USA). Results for the androgen levels of late but the difference did not reach statistical signiﬁcance. pubertal and post-pubertal females are reported. Two or more subnormal androgen levels were measured in 6/10 BMT patients with у2 measurements available. The three spontaneously menstruating transplanted patients Statistical methods (Nos 3, 4 and 14) had normal androgen levels (Table 3, Signiﬁcance of the differences between groups was calcu- Figure 1). lated by the two-tailed t-test or by analysis of variance.

Discussion Results Ovarian failure with amenorrhea, low circulatory estradiol Androgen levels of the female patients treated with TBI levels, elevated gonadotropin levels and infertility is seen were lower than those of the conventionally treated patients in all female patients after post-menarcheal total body (Figure 1, Table 2). The mean testosterone level of 12 of irradiation used as a preparative regimen for BMT.4,5,7 In the BMT patients was significantly lower than that of the females transplanted before menarche, total body 52 healthy controls (mean 0.81, 95% CI 0.72–0.91 nmol/l; irradiation seems to be less damaging to the ovaries.6,7 The P = 0.002) and that of the 19 patients who had had conven- low androgen levels in the BMT patients, reported for the tional chemotherapy (P Ͻ 0.001). The two latter groups did first time in this study, most likely also reflect ovarian not differ from each other. damage. The androstenedione and DHEA levels of the BMT In normal ovaries, both granulosa and theca cells are patients were also lower than those of the conventionally needed for steroid hormone synthesis. The principal steroid

Table 2 Androgen concentrations in late pubertal or post-pubertal female patients treated with total body irradiation and BMT (TBI+) or with conventional chemotherapy (TBI−)

TBI+ TBI− P n = 12 n = 19

Testosterone, nmol/l 0.46 (0.32–0.56) 0.89 (0.72–1.06) Ͻ0.001 Androstenedione nmol/l 2.14 (1.33–2.95) 4.32 (3.56–5.09) 0.002 Dehydroepiandrosterone nmol/l 6.21 (3.20–9.23) 11.35 (9.30–13.39) 0.012 Dehydroepiandrosterone sulfate, ␮mol/l 2.81 (0.96–4.66) 4.10 (3.29–5.09) NS

Means and 95% conﬁdence limits are shown. Androgen production in transplanted females L Hovi et al 564 Table 3 Individual androgen levels in late pubertal or postpubertal females after total body irradiation

Patient No. Testosterone nmol/l Androstenedione nmol/l DHEA nmol/l DHEAS ␮mol/l

3a 0.70 4.5 12.0 5.5 4a 0.90 2.7 7.3 2.6 5 0.35 1.7 13.0 1.5 7 0.39 ND ND 2.9 8 0.20 0.5 1.3 0.2 9 0.42 ND ND ND 10 0.26 1.9 4.5 1.7 11 0.40 2.6 4.5 1.9 12 0.20 0.5 1.6 0.9 13 0.30 2.0 5.5 0.7 14a 0.65 2.9 ND 10.2 15 0.50 ND ND ND

Reference values Mean Ϯ s.d. 0.81 Ϯ 0.39b 4.39 Ϯ 1.74c 14.12 Ϯ 6.38c 3.54 Ϯ 1.75c Range 0.26–1.80 2.55–7.72 5.76–23.82 0.65–6.50

(Values р−1 s.d. of reference values are underlined). The patient number corresponds with that in Table 1. aMenstruating normally. bHealthy controls of this study. cLashansky et al.15 ND = not determined (samples missing).

products of the normal post-menarcheal ovary are estrogen patients. These findings suggest that during TBI some and progesterone. Theca cells produce androstenedione, patients incur damage to the androgen-producing cells of which is converted into estrogen by the granulosa cells. The the adrenal cortex. The sensitivity to irradiation of various high activity of aromatase in these ensures that most of the types of hormone-producing cells in the adrenal cortex, testosterone produced is converted into estradiol and only located in different cell layers, might differ as do, for small amounts of testosterone are released. In this study instance, those in the hypophysis where growth hormone both testosterone and androstenedione production were producing cells are more sensitive to irradiation than gona- impaired in female BMT patients. Testosterone levels were dotropin or thyrotropin producing cells.16 low as compared with those of the conventionally treated After conventional chemotherapy for leukemia, ovarian leukemia patients and healthy controls. The androstene- damage is rare both in girls and in adult females.8,17 This dione levels of the BMT patients were also low as com- finding is supported by the normal androgen production pared with those of conventionally treated leukemia demonstrated in the present study. The testosterone concen- patients or with previously published values for healthy trations in these patients were almost identical to those of pubertal-matched females.15 the healthy controls. In the conventionally treated patients, Although the gonads produce DHEA and androstene- the adrenal androgen concentration compared well with the dione as precursors of testosterone and estrogens, they only values reported for healthy pubertally matched females,15 secrete negligible amounts of DHEA and DHEAS. In which further supports the notion that antileukemia chemo- addition to the gonads, the adrenal cortex produces signifi- therapy is well tolerated. cant amounts of androgen precursors. More than 90% of Normal pubertal development and sexual well-being are the circulating DHEAS derives either from direct adrenal the outcome of complex interactions between several hor- secretion or from peripheral sulfation of DHEA secreted mones and psychologic factors. In female patients treated by the adrenals. The adrenal cortex produces only minute with TBI, delayed pubertal development or symptoms of amounts of testosterone and estradiol, having very low con- menopause are common and usually thought to be due to centrations of the enzymes needed for their production. a defect in ovarian estrogen production. In this study, most Thus, a low concentration of DHEA and especially of of the pre-menarcheally and post-menarcheally transplanted DHEAS is indicative of a defect in hormone synthesis by patients were receiving estrogen or cyclic estrogen and the adrenals. Plasma cortisol levels have been shown to be medroxyprogesterone substitution. Thirteen of 15 trans- within normal limits after TBI and the adrenal gland has planted patients had had delayed pubertal development or been suggested to be resistant to the effects of TBI. Littley amenorrhea after BMT, and ovarian insufficiency had also et al11 did not find any evidence of impaired hypothalamic- been verified by increased gonadotropin levels and/or low pituitary-adrenal function after TBI in 21 adult patients. estradiol levels (data not shown). The clinical experience Plasma cortisol levels were not measured in our patients. is that apparently normal pubertal development and men- In the present study, the mean value of DHEAS in the BMT struation are usually achieved in young BMT patients with patients did not differ from that of conventionally treated the help of hormone substitution but development of breast patients. However, half of the BMT patients had subnormal tissue may be suboptimal and pubic and axillary hair is DHEAS levels. In addition, DHEA levels were significantly often scanty as compared to conventionally treated girls lower in the BMT patients than in conventionally treated with leukemia. Androgen production in transplanted females L Hovi et al 565 The function of androgens in females is not totally clear. gonadal function of cyclophosphamide, total-body irradiation, Testosterone is needed for increased striated muscle and marrow transplantation. Transplantation 1983; 36: 252– activity in women during exercise. Healthy females with 255. higher serum testosterone levels are better able to increase 4 Sklar CA, Kim TH, Williamson JF, Ramsay NKC. Ovarian their muscle mass and strength during an intensive strength function after successful bone marrow transplantation in post- 18 menarcheal females. Med Pediatr Oncol 1983; 11: 361–364. training period. Muscle strength in female survivors of 5 Chatterjee R, Mills W, Katz M et al. Prospective study of leukemia, especially BMT patients, was poorer than in pituitary-gonadal function to evaluate short-term effects of healthy reference subjects. However, we observed no indi- ablative chemotherapy or total body irradiation with autolog- vidual correlations between muscle strength and testoster- ous or allogenic marrow transplantation in post-menarcheal one concentration.9 In male survivors of leukemia, in con- female patients. Bone Marrow Transplant 1994; 13: 511–517. trast, a positive correlation has been found between serum 6 Spinelli S, Chiodi S, Bacigalupo A et al. Ovarian recovery testosterone level and muscle strength parameters.19 The after total body irradiation and allogeneic bone marrow trans- adrenal androgens are known to play an important role in plantation: long-term follow up of 79 females. Bone Marrow the partial virilization that accompanies the development of Transplant 1994; 14: 373–380. normal female secondary sexual characteristics. The devel- 7 Sanders JE, Buckner CD, Amos D et al. Ovarian function following marrow transplantation for aplastic anemia or leuke- opment of pubic and axillary hair, body odor and com- mia. J Clin Oncol 1988; 6: 813–818. edones in pubertal females is due primarily to synergism 8 Wallace WHB, Shalet SM, Tetlow LJ, Morris-Jones PH. 20 between ovarian estrogen and adrenal androgens. Ovarian function following the treatment of childhood acute Ovarian recovery has been reported to occur in 10–15% lymphoblastic leukaemia. Med Pediatr Oncol 1993; 21: of transplanted females who were post-menarcheal but 333–339. under 30 years of age at the time of BMT. The period of 9 Hovi L, Era P, Rautonen J, Siimes MA. Impaired muscle ovarian failure ranged from 2 to 7 years after transplant.6,7 strength in female adolescents and young adults surviving leu- Ovarian recovery has not been shown to be affected by sex kemia in childhood. Cancer 1993; 72: 276–281. hormone replacement therapy. To assess ovarian recovery, 10 Hinterberger-Fischer M, Kier P, Kalhs P et al. Fertility, preg- hormonal substitution is usually discontinued for 2 or more nancies and offspring complications after bone marrow transplantation. Bone Marrow Transplant 1991; 7: 5–9. months annually. On the other hand, measurement of 11 Littley MD, Shalet SM, Morgenstern GR, Deakin DP. Endo- androgen levels might be a useful supplementary way to crine and reproductive dysfunction following fractionated total follow recovery of the ovaries without discontinuation of body irradiation in adults. Q J Med 1991; 78: 265–274. estrogen substitution. In this study, the three BMT patients 12 Gustafsson G, Garwicz S, Hertz H et al. A population-based with spontaneous menstruation had the highest testosterone study of children with acute lymphoblastic leukemia in five and androstenedione levels; several serial measurements Nordic countries diagnosed July 1981 through June 1985: during a longer period are needed before the value of andro- Incidence, characteristics, and treatment results. Acta Paediatr gen levels as a follow-up parameter is established. Scand 1987; 76: 781–788. In conclusion, androgen levels were significantly lower 13 Apter D, Viinikka L, Vihko R. Hormonal pattern of adolescent in females who had been treated with TBI for leukemia menstrual cycles. J Clin Endocrinol Metab 1978; 47: 944–954. than in conventionally treated patients. This might be one 14 Hammond GL, Viinikka L, Vihko R. Automation of radioimmunoassays for some sex steroids with use of both iodinated factor behind the problems in pubertal development, muscle and tritiated ligands. Clin Chem 1977; 23: 1250–1257. function, and sexual life experienced by these patients. 15 Lashansky G, Saenger P, Fishman K et al. Normative data for Further studies are needed to evaluate the usefulness of adrenal steroidogenesis in a healthy pediatric population: age- these hormone levels for follow-up purposes and the poten- and sex-related changes after adrenocorticotropin stimulation. tial value of androgen replacement therapy in female J Clin Endocrinol Metab 1991; 73: 674–686. BMT patients. 16 Mauras N, Sabio H, Rogol AD. Neuroendocrine function in survivors of childhood acute lymphocytic leukemia and non- Hodgkins lymphoma: a study of pulsatile growth hormone and gonadotropin secretions. Am J Pediatr Hematol/Oncol 1988; Acknowledgements 10: 9–17. 17 Kreuser ED, Hetzel WD, Heit W et al. Reproductive and This study was supported by a grant from the Foundation of Nona endocrine gonadal functions in adults following multidrug and Kullervo Va¨re. chemotherapy for acute lymphoblastic or undifferentiated leukemia. J Clin Oncol 1988; 6: 588–595. 18 Ha¨kkinen K, Pakarinen A, Kyro¨laïnen H et al. Neuromuscular adaptations and serum hormones in females during prolonged References power training. Int J Sports Med 1990; 11: 91–98. 19 Talvensaari KK, Ja¨msen A, Vanharanta H, Lanning M. 1 Wallace WHB, Shalet SM, Hendry JH et al. Ovarian failure Decreased isokinetic trunk muscle strength and performance following abdominal irradiation in childhood: the radiosensi- in long-term survivors of childhood malignancies: correlation tivity of the human oocyte. Br J Radiol 1989; 62: 995–998. with hormonal defects. Arch Phys Med Rehabil 1995; 76: 2 Sanders JE, Pritchard S, Mahoney P et al. Growth and devel- 983–988. opment following marrow transplantation for leukemia. Blood 20 Miller WL, Tyrrell JB. The adrenal cortex. In: Felig P, Baxter 1986; 68: 1129–1135. JD, Frohman LA (eds). Endocrinology and Metabolism. 3 Sanders JE, Buckner CD, Leonard JM et al. Late effects on McGraw-Hill: New York, 1995, pp 555–712.