(Received 22nd May 1974)

THE EFFECT OF ON TESTICULAR ANDROGEN PRODUCTION IN ADULT MEN WITH CHRONIC RENAL FAILURE K. RAGER, H. BUNDSCHU and D. GUPTA Department of Diagnostic Endocrinology, University Children's Hospital, Tübingen, and *Medizinische Universitäts-Klinik, Tübingen, Germany (Received 22nd May 1974) Summary. In sixteen male patients undergoing regular haemodialysis or peritoneal dialysis for chronic renal failure, the testosterone levels were studied before and after treatment with. Testosterone values before (186 ng/100 ml) and after (456 ng/100 ml) were significantly lower than those for normal healthy males. The mean plasma dihydrotestosterone (5\g=a\-DHT) value of 11\m=.\5 ng/100 ml for the treated patients was significantly lower than that of 64 ng/100 ml for the control patients. With, increments in DHT were recorded in three patients, and definite increments in testosterone were observed in six patients. The percentage binding and combining affinity showed little change due to stimulation. Achieving ejaculation was a serious problem for the patients in whom the lowest increases in testosterone and 5\g=a\-DHTlevels were observed after stimulation. Low plasma testosterone values were also muscle fibres. associated with a low mean area for Type-2 INTRODUCTION Following the observation that chronic renal failure in man is associated with decreased plasma androgen levels (Gupta & Bundschu, 1972), we undertook in the present study to examine the functional capacity of the testes in such patients treated with. We also attempted to find if a decrease in the libido of these patients was associated with a change in their plasma androgen levels, as well as with the muscular atrophy often observed. MATERIALS AND METHODS Sixteen adult male patients with chronic renal failure constituted an experi mental group, while twenty soldiers, free of apparent endocrinological dis orders, formed a volunteer control group. Of the experimental group, nine patients underwent haemodialysis, six peritoneal dialysis and one (No. 16), who was in compensated renal failure, was treated by diet alone. The clinical details for the experimental group are listed in Table 1. 113

114 K. Rager et al. All blood samples were collected under standardized hospital conditions at the same time each day. The blood was centrifuged, and the plasma im mediately deep-frozen until the assays could be performed. The response was tested by administering 1000 i.u. (Primogonyl: Schering AG, Berlin) daily for 12 to 14 days. The normal testosterone values after administration were taken from published observations (Nieschlag, Rohr, Wombacher & Overzier, 1971; Littmann & Gerdes, 1972). Plasma testosterone and 5a-dihydrotestosterone (5a-DHT) were estimated simultaneously by a competitive protein-binding assay developed in our laboratory (Gupta, McCafferty & Rager, 1972). This was achieved by using Table 1. Clinical data of sixteen male patients with chronic renal failure Patient 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 No. of treatments by HDorPD already under gone by patient 119 390 104 9 140 376 44 50 10 25 8 31 4 11 6 Body weight (kg) 58 75 62 71 65 67 58 64 62 59 61 54 72 64 67 Age (years) 22 36 28 60 28 33 36 36 50 41 54 42 53 48 30 Serum protein* (g/looml) 6-8 7-7 6-8 6-6 6-6 6-6 6-8 6-4 6-4 6-4 50 4-6 60 6-4 6-6 Sexual life + + + + Diagnosis Chronic P.N. Chronic P.N. Chronic P.N. Chronic P.N. 16 58 33 6-6 + Patients 1 to 9 were undergoing haemodialysis (HD); Patients 10 to 15 were undergoing peritoneal dialysis (PD) ; Patient 16 was being treated by diet alone. G.N. = glomerulonephritis; P.N. pyelonephritis. Sexual life: normal; = = + = reduction; = severe reduction; 4- = impotence. * Protein content was estimated before dialysis. a two-step chromatography in two different systems after extraction; the first step was thin-layer chromatography using benzene : ethyl acetate (3:2, v/v) and the second step was descending paper chromatography using a light petroleum : methanol : water (5:4:1, by vol.) eluant. The steroids (such as dehydroepiandrosterone, androstenedione and androsterone), which may be present in large amounts on the chromatograms and can contaminate the testosterone and 5oc-DHT zones, do not, however, compete with these sub stances. The 5a-DHT isolated and pure 5a-DHT showed identical mass spectra (Gupta et al., 1972). With a recovery rate of testosterone and 5a-DHT of 52-5% and a blank value of 0-18 ng/sample, a plasma level of 4-3 ng /100 ml

- - Testicular androgen in men with renal failure 115 can be satisfactorily detected. The % testosterone binding was measured in plasma dilutions of 1:5 and 1:25 (v/v) by using dextran-coated charcoal to separate bound and unbound testosterone (Gupta, Huenges & Rager, 1971; Gupta & Bundschu, 1972). The 'combining affinity' (litres/g) of the plasma proteins for testosterone was calculated (Westphal, 1961) from the % testo sterone binding levels : Combining affinity = Steroid bound 1 Steroid unbound This formula takes into consideration the plasma protein content (P), in g/litre, estimated by the biuret method. For the sexual behaviour analysis, the patients were arbitrarily allocated to four groups according to their self-reported sex life before the onset of renal failure. Group 1 contained patients who were completely impotent and Group 4 those who reported normal ejaculation. The remaining two groups were between these two extremes. By using NADH-dehydrogenase it is possible to distinguish two different types of fibres in human muscle. Type 1 fibres are present in normal males in a greater amount than Type 2 and it is now established that Type 2 fibres become atrophic in patients with chronic renal failure, thereby producing a markedly reduced mean area (Bundschu, Suchenwirth & Dürr, 1970; Bund schu & Suchenwirth, 1973). 1400 r ( ) 90 r 1000 600 200 Before After Before After Men with chronic renal failure Normal men Men with chronic renal failure Text-fig. 1. Concentrations of plasma testosterone (a) and dihydrotestosterone (b) in patients with chronic renal failure and in normal men before and after treatment. The data for the normal men are taken from the published literature. Vertical lines with horizontal bars represent mean values + S.D. Normal men

116 K. Rager et al. 100 80.',. ' il 60 40 20 ( ) _L J_I_I_I_I_I_L 100 80 60 40! '! M 20 (b) J_I_I_I_I_I_ 8 10 Patient no. J_1_L 12 14 16 Text-fig. 2. Percentage testosterone binding in (a) 1:5 and (b) 1:25 dilutions of plasma from patients with chronic renal failure before (#) and after ({) treatment.

- Testicular androgen in men with renal failure 117 RESULTS The testosterone level of the patients in the experimental group was 186 ng/100 ml, which was significantly lower ( <0 01) than that of the men in the control group (456 ng/100 ml). After the administration of, the mean testo sterone level for the patients in the experimental group rose to 281 ng/100 ml, while that of the normal adult males averaged 1000 ng/100 ml (Littmann & Gerdes, 1972; Nieschlag et al., 1973). In plasma, 5a-DHT levels paralleled those of testosterone in the same 0 08 r 0 06 004 0-02 Before After Mean ± S.D. Normal men Text-fig. 3. Combining affinity before and after administration in patients with chronic renal failure. samples, yielding a mean of 11-5 ng/100 ml for the experimental group. This was significantly lower (P<0-001) than the mean of 64 ng/100 ml recorded for our own volunteer control group. Two patients (Nos 2 and 3) had testosterone values within the normal range, while an increase was detected in six patients. After stimulation, plasma 5a-DHT levels for five patients (Nos 6, 8, 11, 12 and 13) did not rise but in three patients (Nos 5, 14 and 16) a marked difference between pre- and post- stimulation was recorded (Text-fig. 1).

118 K. Rager et al. In the 1:5 (v/v) plasma dilution, a testosterone binding of 75+11 % was found for the men in the experimental group compared to a value of 65-3+ 4-8% for those in the control group. After the administration of 12,000 to 14,000 i.u., the testosterone-binding level decreased to a mean value of 68-3+13-2% in the experimental group. A normal testosterone-binding level before administration was found in five patients (Nos 2, 5, 7, 9 and 14), and in seven patients (Nos 3, 4, 6, 8, 11, 13 and 15) administration did not lead to an improvement. In the 1:25 (v/v) plasma dilution, the testosterone- Ë o o 700» 400 100 Impotence Text-fig. 4. The relationship of the sexual life of patients with chronic renal failure to changes ( ) in the plasma concentrations of testosterone and dihydrotestosterone before and after. Vertical lines represent the S.D. binding level of the patients in the experimental group was found to be 58%, and a value of 45% was reported for those in our own volunteer control group. After, a drop in the percentage binding occurred in seven patients who had normal testosterone-binding levels. Three patients (Nos 1, 5 and 14) who had elevated % testosterone-binding levels before treatment with achieved normal levels after (Text-fig. 2).

- Testicular androgen in men with renal failure 119 The combining affinity in the two different dilutions, clearly differentiated the experimental group of patients (0-028 litres/g) from our own volunteer control group (0-0078 litres/g). Individually, thirteen (Nos 2 to 4, 6 to 14 and 16) of the sixteen patients had values significantly (P<0-001) higher than normal. Administration of resulted in a decrease in the combining affinity (Text-fig. 3). In relating the plasma androgen levels to sexual behaviour (Text-fig. 4), it was noted that the smallest changes in testosterone and 5a-DHT levels were 7000 E 6000 5000 g 4000 S 3000 100 200 300 400 500 Testosterone (ng/iooml) Normal men Text-fig. 5. A comparison of the plasma testosterone concentrations and the area of Type 2 muscle fibres in seven patients with chronic renal failure. associated with impotence. A significant difference (P<0-001) was found between the steroid levels in Group 1 and Group 4. The mean area of Type 2 muscle fibres in seven patients (Nos 1 to 3, 5 and 7 to 9) of the experimental group who underwent haemodialysis was 3882+ 1143 µ 2 as compared to the control group value of 6149+838 µ 2. Because of the small number of patients in the experimental group, a statistical cor relation was not calculated. It was noted, however, that low plasma testo sterone values were associated with small mean areas of Type 2 muscle fibres, and higher testosterone levels were associated with larger areas (Text-fig. 5).

120 K. Rager et al. DISCUSSION Chronic renal failure in man is thought to be associated with the syndrome of decreased libido, decreased testicular size and muscular atrophy. We consider that our findings lend support to this hypothesis. Stimulation of the testes of the patients in the experimental group with, even in high doses, did not improve the androgen levels. The levels for % testosterone binding and com bining affinity also showed little change. These results indicate that in chronic renal failure, androgen synthesis by the testicular tissue and metabolic con version of testosterone to 5a-DHT by the prostatic cells may be impaired. The factor in patients undergoing haemodialysis or peritoneal dialysis which is responsible for the unresponsiveness of the testes to, however, remains undefined. REFERENCES Bundschu, H. D. & Suchenwirth, R. (1973) Primare und sekundäre Myopathien aus enzymhistologischer Sicht. Fortschr. Neurol. Psychiat. 41, 419-449. Bundschu, H. D., Suchenwirth, R. & Dürr, F. (1970) Enzymhistorische Befunde an der Skelett muskulatur des Menschen. II. Muskelschwund bei Niereninsuffizienz. Klin. Wschr. 48, 1380 1397. Gupta, D. & Bundschu, H. D. (1972) Testosterone and its binding in the plasma of male subjects with chronic renal failure. Clin. chim. Acta, 36, 479 486. Gupta, D., Huenges, R. & Rager, K. (1971) Studies on testosterone-protein reactions in children in relation to their sexual maturation. Ada endocr., Copenh., Suppl. 152, 24, Abstr. Gupta, D., McCafferty, E. & Rager, K. (1972) Plasma 5a-dihydrotestosterone ( 17/9-hydroxy-5aandrostan-3-one) in adolescent males at different stages of sexual maturation. Steroids, 19, 411-431. Littmann,. P. & Gerdes, H. (1972) Testosteron tagesrhythmik und -Stimulation bei akuten und chronischen Leberkrankungen. Klin. Wschr. 50, 1123-1126. Nieschlag, E., Rohr, M., Wombacher, H. & Overzier, C. (1971) Leydig-Zell-Funktionstest. Bestimmung von Plasmatestosteron durch kompetitive Proteinbindung vor und nach - Belastung. Klin. Wschr. 49, 91-100. Westphal, U. (1961) Interactions between steroids and proteins. In Mechanism of Action of Steroid Hormones, pp. 33-89. Eds. C. A. Villee and L. L. Engel. Pergamon Press, Oxford.