- Email: [email protected]
ŒSTROGEN RECEPTORS IN HUMAN MALIGNANT MELANOMA
337 mahties of lactate extraction and haemodynamics. Boujoulas et a1.lo found that some of their patients with angina and normal coronary arteries produced lactate juring pacing but that left ventricular function was normal. The lack of consistent evidence in these patients may, however, be a reflection of the insensitivity of methods for detecting ischaemia. Even in patients with proven coronary-artery disease, metabolic and hxmoabnormalities cannot always be demonstrated."-" Several explanations have been advanced for angina in the presence of normal coronary arteriesincluding small-vessel disease of the myocardium, spasm of the coronary arteries, oxyhaemoglobin dissociation defects, misinterpretation of the arteriograms, cardiomyopathy, and psychosomatic factors.4 All remain unsubstantiated. Herman et al .4 claim that small-vessel disease of the myocardium is accompanied by systemic disease and involvement of other organs. They report also that conduction abnormalities are frequent, but these features were not present in our patients. Furthermore small-vessel disease would be expected to produce slow clearing of dye to the coronary sinus during angiography, which we did not observe. The few post-mortem studies of patients with angina and normal coronary arteries have yielded no evidence of small-vessel disease." Spasm of the coronary arteries may account for some episodes of pain but, in view of the generally poor response to nitroglycerin, is very unlikely to be a major factor. No abnormalities of oxyhaemoglobin dissociation were found in patients investigated by Vokonas et a},l4 Consistent misinterpretation of coronary arteriograms by many experienced investigators using modern techniques is improbable. Psychosomatic factors are clearly present in some patients but do not explain cases where the E.c.G. is abnormal. It is possible that these patients have an unknown metabolic defect, resulting in a form of myopathy. Our patients however, had none of the features of generally recognised cardiomyopathies. Many questions concerning these patients remain unanswered; but clearly they have an excellent prognosis, with myocardial infarction an uncommon event. The majority have a decrease in angina.36 In the present series there were no myocardial infarctions during the follow-up period, and 32 (73%) either improved or lost their anginal pain during this period. Treatment is difficult and response often poor, but we have found long-acting nitroglycerin and beta sympathetic blocking agents to be the most successful combination.
dynamic
Requests for reprints should be addressed
to
E.S.
REFERENCES
1 Likuff, W, Segal, B. L., Kasparian, H. New Engl. J. Med. 1967, 276,
1068.
2 Dwyer, F. M., Weiner, J. M., Cox, W. J. Am. J. Cardiol. 1969, 23, 639. 3 Kemp, H G., Vokonas, P. S., Cohn, P. F., Gorlin, R. Am. J. Med. 1973, 54, 740
4 Herman, J., M Cohn, P. F., Gorlin, R. Ann. intern. Med. 1973, 79, 5 Hurst,J W., Logue, R. B. The Heart. New York, 1970. 6 Bemiller, C. R., Pepine, C. J., Rogers, A. K. Circulation, 1973, 47, 36.
445.
7 Howitz, L.D. J Am. med. Ass. 1974, 229, 1196. 8 Fowler,N O.Circulation, 1972, 46, 1079. 9 Arbogast, R, Bourassa, M. G. Am. J. Cardiol. 1973, 32, 257. 10 Boudoulas, H., Cobb, T. C., Leighton, R. F., Wilt, S. M. ibid. 1974, 34, 501. 11 Hafant, R H., Forrester, J. S., Hampton, J. R., Haft, J. I., Kemp, H. G., Gorlin, R. Circulation, 1970, 42, 601. 12 Dwyer, E M. Jr. ibid. p. 1111. 13 Eliot, R S, Bratt, G.Am.J. Cardiol. 1969, 23, 633. 14 Vokonas,P S., Cohn, P. F., Klein, M. D. ibid. 1970, 26, 664.
ŒSTROGEN RECEPTORS IN HUMAN MALIGNANT MELANOMA RICHARD I. FISHER JAMES P. NEIFELD MARC E. LIPPMAN Medicine and Surgery Branches, National Cancer Institute Bethesda, Maryland 20014, U.S.A.
16 of 35 patients (46%) with malignant melanoma were found to have cytoplasmic œstrogen-receptor activity in biopsy specimens. Œtrogen-receptor activity was detected in primary lesions, lymph-node metastases, skin metastases, and visceral metastases. Equal percentages of males and females had positive assays. Scatchard analysis of binding in one case was consistent with a single class of highaffinity receptor sites.
Summary
Introduction THE natural history of human malignant melanoma suggests that steroid hormones may affect the biological behaviour of the tumour. Malignant melanoma is extremely rare before puberty. The peak incidence of melanomas in females coincides with the childbearing years and the menopause. Several workers2have found rapid dissemination of melanoma during pregnancy. Some workers have also reported spontaneous regression of melanoma following parturition.4 It is notable that the prognosis for women regardless of stage or treatment is consistently better than that for men.5 Finally, occasional patients with disseminated melanoma have had objective responses following various endocrine manipulations including treatment with oestrogens or androgens and hypophysectomy.1 Studies on the mechanism of action of steroid hormones have shown that cytoplasmic steroid-receptors are a necessary, if not sufficient, requirement for steroid-hormone responsivesness in target tissues.66 Therefore, we have analysed biopsy tumour specimens from patients with malignant melanoma for the presence of cytoplasmic hormone-receptors for oestradiol. Patients and Methods
Biopsies were obtained from 35 patients (19 male, 16 female) with malignant melanoma. Sites of tumour biopsies were as follows: 3 from the primary lesion, 18 from lymphnode metastases, 12 from skin metastases, and 2 from visceral metastases.
Tumour samples obtained at surgery were trimmed of surrounding normal tissue, immediately frozen in liquid nitrogen, and then stored at -70’ C. Before freezing, a portion of the biopsy specimen was always examined histopathologically to confirm the diagnosis. Assays were generally performed within two weeks of sample collection. Samples for assay were weighed in pretared 15 ml corex tubes chilled in liquid nitrogen. The sample was then powdered in a tissue pulveriser (Thermovac Industries). All subsequent steps were performed at 0-4 °C. The powered tissue was homogenised with three tensecond bursts in a PT-10 homogeniser (Brinckmann Industries) separated by two twenty-second cooling periods in a buffer consisting of 10 mmol/1 "ths"-HCl, 1-55 MMOVI E.D.T.A., and 1 mmol/1 dithiothreitol. The homogenate was centrifuged at 104 000 x g at 40C for one hour. The supernatant was diluted to yield a protein concentration of 2-4 mg/ml. The supernatant cytosol was assayed for oestrogen-receptor activity using modifications7 of a previously described competitive protein-binding assay8 in which dextran-coated charcoal is used to
338 separate unbound HJcestradiol-17 from protein associated radioactivity. When sufficient sample permitted, multiple concentrations of [3H]cestradiol-17&bgr; were used for construction of Scatchard plots of the binding data. On smaller samples, triplicate or quadruplicate determinations using a concentration of cestradiol-17&bgr; estimated to be about 3 times the dissociation constant were used.
Results
Cytoplasmic receptors for oestrogen were detected in 16 of the 35biopsies. We considered a positive assay to be greater than 5 fmol/mg cytoplasmic protein. Operationally no analysis yielded a Scatchard plot which suggested specific high affinity binding with fewer receptor sites. 47% (9/19) of males and 43% (7/16) of females had receptor activity. When analysed according to the site of tumour biopsy, oestrogen receptors were present in 2/3 of primary melanomas, 9/18 of lymph-node metastases, 4/12 of skin metastases, and 1/2 of visceral metastases.
A typical binding curve is shown in the accompanying figure in which specific (competitive) binding is plotted as a function of increasing H]cestradiol-17 concentration. The same binding data is replotted in the inset of this figure using the Scatchard technique.9 The straight line obtained (r=-0919) suggests that under our assay conditions oestradiol is binding to a class of receptor sites of uniform affinity and limited capacity. From the slope of this line we estimate the dissociation constant to be 49x 10-9 mol/l. The "X" intercept pro-
vides
an
estimation of the total number of recptor sites in this case 24.7
saturating concentration of steroid, fmol/mg cytoplasmic protein. at
Discussion The present study was initiated because of evidence suggesting that the biological behaviour of human malignant melanoma might be affected by steroid hormones, Malignant melanoma is extremely rare before puberty. The incidence of melanoma in southern England is 3.5 per 100 000 population per year. Yet in a population of 3 million, Bodenham1 reported only 2 cases before the age of 15. SadofflO found that the peak incidence of melanoma in females occurred during the childbearing period (third decade) and menopause (fifth decade). There has been considerable controversy concerning the possible deleterious effect of pregnancy on the course of melanoma. Early studies stressed the rapidly fatal outcome observed in pregnant patients with melanoma while other studies have failed to substantiate this." Shiu et al.12 found that of all female stage II patients, those who were either pregnant or experienced an activation of their tumour during previous pregnancies had a shorter survival. Allen4 described a patient whose melanoma metastasised widely during pregnancy and then underwent spontaneous complete regression within eleven weeks of delivery. Despite these factors, when male and female patients with melanoma are compared, the survival of females regardless of stage or treatment has been significantly longer. 5 13 14 Although most attempts at endocrine manipulation6 have been unsuccessful in malignant melanoma," occasional patients have responded.Johnson et al." reported an 11% objective response rate in metastatic malignant melanoma treated with 6o(-methylpregn-4ene-3, 11, 20-trione, a drug with glucocorticoid and
anti-oestrogen activity. Studies in human breast
cancer
have shown that
analysis of cestrogen receptors in tumour specimens correlates significantly with the likelihood of an objective response following hormonal manipulation. If a breast tumour is cestrogen-receptor-positive, the response rate to a variety of additive or ablative therapies range from 50 to 70%.6 However, for patients lacking oestrogen receptor, the response rate is generally less than 10%. Some studies in breast cancer have also suggested that both androgens and progesterone’9 receptors can increase the likelihood of predicting clinical responsiveness. Similar studies have suggested that glucocorticoid receptor analyses may aid in selecting patients with leukaemia who have an increased likelihood of respond ing to steroid therapy.20 21 Further studies will be neces sary to determine whether a subset of patients witl malignant melanoma, who have a higher likelihood o response to endocrine manipulation, can be selected b hormone-receptor analysis. Requests for reprints should be addressed to R.I.F., Building 10 Room 12N226, National Cancer Institute, Bethesda, Maryland 20014 U.S.A. REFERENCES
Bodenham, D. C., Hale, B. in Endocrine Therapy in Malignant Disease (edited by B. A. Stoll); p. 377. London, 1972. 2. Pack, G. T., Scharnagel, L. M. Cancer, 1951, 4, 324. 3. Conybeare, R. C. Obstet. Gynec., N.Y. 1964, 24, 451. 4. Allen, E. P. Br. med. J. 1955, ii, 1067. 5. DeVita, V. T., Fisher, R. I. Ca. Treat. Rep. 1976, 60, 153. 6. McGuire, W. H., Carbone, P. P., Sears, M. E. in Estrogen Receptors Human Breast Cancer; p. 1. New York, 1975. 7. Lippman, M. E., Huff, K. K. Cancer, (in the press). 8. McGuire, W. L., DeLaGarza, M. I. J. clin. Endocr. Metab. 197., 36, 548. 9. Scatchard, G. Ann. N.Y. Acad Sci. 1949, 51, 660. 1.
Binding of PH] oestradiol (upper graph).
to
receptor in human malignant melanoma
The binding data in the upper graph are replotted in graph using the Scatchard technique. Kd=Equilibrium dissociation constant, 4.9x 10-9 mmol/l. B/F=Bound divided by free [H]oestradiol concentration.
the lower
339
COTRIMOXAZOLE AND FOLATE METABOLISM M. C. BATESON*
Gordon Hospital and Department of Therapeutics, Westminster
Hospital, London P. PENDHARKAR
J. P. L. A. HAYES
Department of Hœmatology Westminster Medical School, London
Cotrimoxazole 4 tablets daily (1 tablet = trimethoprim 80 mg and sulpha400 methoxazole mg) was given for a period of six to fourteen days to 13 inpatients, and serum-folate levels were measured before and one day after the course of treatment. The results were compared with those from 8 patients not receiving antibacterial therapy, tested on admission and one week later. Two assay techniques were used, one employing Lactobacillus casei and the other 125I-labelled folate isotope dilution. The microbiological technique showed a significant decline in folicacid levels in the serum after cotrimoxazole, and this decline was not seen in controls. By contrast, the radioisotope technique showed no significant alteration in serum-folate levels compared with controls. This suggests that cotrimoxazole does not depress true serum-folate and that many low microbiological results obtained during cotrimoxazole therapy reflect interference with the assay organism. There is insufficient evidence to incriminate cotrimoxazole as a significant cause of blood dyscrasias in excess of those which might occur on sulphonamide alone or even with other antibacterials.
Summary
Introduction MUCH has been written on the interference by cotrimoxazole with bone-marrow function and folate metabolism. This work has taken the forms of case-reports of blood dyscrasias,1-13 the assay of serum-folate in volun68 9 marrow biopteers and patients on cotrimoxazole,5 sies in patients on cotrimoxazole,2-59 and reports to th Committee for Safety of Medicines. Cases where standard doses of cotrimoxazole have been administered and some subsequent disturbance of folate metabolism or hxmatological change reported have included only one certain case of megaloblastic anaemia without other likelv caused This occurred after a year’s course of the drug *Present address
Department
of Medicine, Dundee
University.
and the serum-folate (bacterial) method was normal. Various changes such as anaemia, neutropenia, thrombocytopenia, and megaloblastosis of the marrow have been reported in patients on cotrimoxazole. Some workers have suggested that serum-folate levels fall on a sulphonamide/trimethoprim combination, 14-16 others that they remain normal. 17 Against reports in several patients of megaloblastic changes on cotrimoxazole must be set the finding in a controlled thirteen-week trial that 11 patients on cotrimoxazole did not have megaloblastosis, whereas 1 on placebo did.4We have assessed the validity of the microbiological technique of folate assay after the exhibition of cotrimoxazole by using the isotope dilution technique. 18-20 Methods
patients on standard courses of cotrimoxazole for urinary trior lower respiratory infections 4 tablets daily (1 tablet methoprim 80 mg and sulphamethoxazole 400 mg) for six to fourteen days had the following investigations immediately before and on the day after completing therapy: (a) Hb, white blood-cells, mean corpuscular volume (Coulter S), (b) platelets (Coulter ’Thrombocounter’) (c) Serum-vitamin-B12 (Euglena gracilis21) (d) Serum-folate (Lactobacillus casei) (e) red-bloodcell (R.B.c.) folate (L. casei) (j9 Serum folate (isotope dilution Bio Rad ’Quanta-Count’ Kit’9). 8 controls on no antibacterial or antifolate agents had the same investigations at a seven-day interval. Patients were generally studied immediately after hospital admission. To normal serum of a known folate activity were added serial dilutions of trimethoprim. The apparent serum-folate activity was then re-estimated. 13
=
Differences in ’folate levels were submitted to Student’s t tests, and correlation coefficients calculated (table i). 5 of the 13 trial patients had a final serum-folate (L. casei assay, twelve to twenty-four hours after the last dose of cotrimoxazole) less then lg/1; this was calculated as a value of 1 p.g/1. R.B.C. folate levels did not change.
paired
Results theradioisotope dilution technique there moderate fall in values for serum-folate in both the control patients and the patients on cotrimoxazole. A similar fall, of 25-30%, was seen in the microbiogically estimated results for the control patients. The initial radiofolate assay values correlate very well with the microbiological ones for both controls (r=0-76; and cotrimoxazole patients P<0-001) (r=0’75 P<0.001). The paired results for the radiofolate assay before and after therapy also correlate well for control(r=0.76, p<0-05) and cotrimoxazole patients (r=0.69
Judged by
.was a
p
results in the control pa well correlated as well (r=0.78, p<0-001) However, the paired microbiological results for the pa tients
10 Sadoff, L, Winkley, J. Tyson, S. Oncology, 1973, 27,
244.
11 George, P. A., Fortner, J. G., Pack, G. T. Cancer, 1960, 13, 854.
were
TABLE I-SERUM-FOLATE LEVELS
12 Shiu, W. H., Schottenfeld, D., Maclean, B., Fortner, J. G. ibid. 1976, 37,
COTRIMOXAZOLE AND
181.
(jJLg/L) 8
IN
13 PATIENTS
ON
CONTROLS
13 White, L P New Engl. J. Med. 1955, 260, 789. 14 Einhorn, L. H., Burgess, M A., Vallejos, C., Bodey, G. P., Gutterman, J., Mavligit, G., Hersch, E. M., Luce, J. K., Frei, E., Freireich, E. J., Gottlieb, J A. Cancer Res. 1974, 34, 1995. 15 Nathanson, L., Hall, T. C., Farber, S. Cancer, 1967, 20, 650. Meyer, H W, Gumport, S. L. Ann. Surg. 1953, 138, 643. 17 Johnson, R. 0, Bisel, H. Andrews, N., Wilson, W. Rochhn, A., Segaloff, A., Krementz, E., Aust. J., Ansfield, F. Ca. Chemother. Rep. 1966, 50, 671. 18 Englesman. E, Korsten, C. B., Persign, J. R., Cleton, F. I. Br. J. Cancer, 16
1975, 30, 177 19 Horowitz, E. R, McGuire, W. L. Steroids, 1975, 25, 497. 20
Lippman, M. E., Halterman, R., Leverthal, B. G., Perry, S., Thompson,
E. B J clin Invest. 1973, 52, 1715. 21 Lippman, M. E., Perry, S., Thompson, E. B.
Am. J.
Med.
1975, 59, 224.
Mean
sampling interval in both groups=7, days
N.s.Not
significant.
dynamic
Requests for reprints should be addressed
to
E.S.
REFERENCES
1 Likuff, W, Segal, B. L., Kasparian, H. New Engl. J. Med. 1967, 276,
1068.
2 Dwyer, F. M., Weiner, J. M., Cox, W. J. Am. J. Cardiol. 1969, 23, 639. 3 Kemp, H G., Vokonas, P. S., Cohn, P. F., Gorlin, R. Am. J. Med. 1973, 54, 740
4 Herman, J., M Cohn, P. F., Gorlin, R. Ann. intern. Med. 1973, 79, 5 Hurst,J W., Logue, R. B. The Heart. New York, 1970. 6 Bemiller, C. R., Pepine, C. J., Rogers, A. K. Circulation, 1973, 47, 36.
445.
7 Howitz, L.D. J Am. med. Ass. 1974, 229, 1196. 8 Fowler,N O.Circulation, 1972, 46, 1079. 9 Arbogast, R, Bourassa, M. G. Am. J. Cardiol. 1973, 32, 257. 10 Boudoulas, H., Cobb, T. C., Leighton, R. F., Wilt, S. M. ibid. 1974, 34, 501. 11 Hafant, R H., Forrester, J. S., Hampton, J. R., Haft, J. I., Kemp, H. G., Gorlin, R. Circulation, 1970, 42, 601. 12 Dwyer, E M. Jr. ibid. p. 1111. 13 Eliot, R S, Bratt, G.Am.J. Cardiol. 1969, 23, 633. 14 Vokonas,P S., Cohn, P. F., Klein, M. D. ibid. 1970, 26, 664.
ŒSTROGEN RECEPTORS IN HUMAN MALIGNANT MELANOMA RICHARD I. FISHER JAMES P. NEIFELD MARC E. LIPPMAN Medicine and Surgery Branches, National Cancer Institute Bethesda, Maryland 20014, U.S.A.
16 of 35 patients (46%) with malignant melanoma were found to have cytoplasmic œstrogen-receptor activity in biopsy specimens. Œtrogen-receptor activity was detected in primary lesions, lymph-node metastases, skin metastases, and visceral metastases. Equal percentages of males and females had positive assays. Scatchard analysis of binding in one case was consistent with a single class of highaffinity receptor sites.
Summary
Introduction THE natural history of human malignant melanoma suggests that steroid hormones may affect the biological behaviour of the tumour. Malignant melanoma is extremely rare before puberty. The peak incidence of melanomas in females coincides with the childbearing years and the menopause. Several workers2have found rapid dissemination of melanoma during pregnancy. Some workers have also reported spontaneous regression of melanoma following parturition.4 It is notable that the prognosis for women regardless of stage or treatment is consistently better than that for men.5 Finally, occasional patients with disseminated melanoma have had objective responses following various endocrine manipulations including treatment with oestrogens or androgens and hypophysectomy.1 Studies on the mechanism of action of steroid hormones have shown that cytoplasmic steroid-receptors are a necessary, if not sufficient, requirement for steroid-hormone responsivesness in target tissues.66 Therefore, we have analysed biopsy tumour specimens from patients with malignant melanoma for the presence of cytoplasmic hormone-receptors for oestradiol. Patients and Methods
Biopsies were obtained from 35 patients (19 male, 16 female) with malignant melanoma. Sites of tumour biopsies were as follows: 3 from the primary lesion, 18 from lymphnode metastases, 12 from skin metastases, and 2 from visceral metastases.
Tumour samples obtained at surgery were trimmed of surrounding normal tissue, immediately frozen in liquid nitrogen, and then stored at -70’ C. Before freezing, a portion of the biopsy specimen was always examined histopathologically to confirm the diagnosis. Assays were generally performed within two weeks of sample collection. Samples for assay were weighed in pretared 15 ml corex tubes chilled in liquid nitrogen. The sample was then powdered in a tissue pulveriser (Thermovac Industries). All subsequent steps were performed at 0-4 °C. The powered tissue was homogenised with three tensecond bursts in a PT-10 homogeniser (Brinckmann Industries) separated by two twenty-second cooling periods in a buffer consisting of 10 mmol/1 "ths"-HCl, 1-55 MMOVI E.D.T.A., and 1 mmol/1 dithiothreitol. The homogenate was centrifuged at 104 000 x g at 40C for one hour. The supernatant was diluted to yield a protein concentration of 2-4 mg/ml. The supernatant cytosol was assayed for oestrogen-receptor activity using modifications7 of a previously described competitive protein-binding assay8 in which dextran-coated charcoal is used to
338 separate unbound HJcestradiol-17 from protein associated radioactivity. When sufficient sample permitted, multiple concentrations of [3H]cestradiol-17&bgr; were used for construction of Scatchard plots of the binding data. On smaller samples, triplicate or quadruplicate determinations using a concentration of cestradiol-17&bgr; estimated to be about 3 times the dissociation constant were used.
Results
Cytoplasmic receptors for oestrogen were detected in 16 of the 35biopsies. We considered a positive assay to be greater than 5 fmol/mg cytoplasmic protein. Operationally no analysis yielded a Scatchard plot which suggested specific high affinity binding with fewer receptor sites. 47% (9/19) of males and 43% (7/16) of females had receptor activity. When analysed according to the site of tumour biopsy, oestrogen receptors were present in 2/3 of primary melanomas, 9/18 of lymph-node metastases, 4/12 of skin metastases, and 1/2 of visceral metastases.
A typical binding curve is shown in the accompanying figure in which specific (competitive) binding is plotted as a function of increasing H]cestradiol-17 concentration. The same binding data is replotted in the inset of this figure using the Scatchard technique.9 The straight line obtained (r=-0919) suggests that under our assay conditions oestradiol is binding to a class of receptor sites of uniform affinity and limited capacity. From the slope of this line we estimate the dissociation constant to be 49x 10-9 mol/l. The "X" intercept pro-
vides
an
estimation of the total number of recptor sites in this case 24.7
saturating concentration of steroid, fmol/mg cytoplasmic protein. at
Discussion The present study was initiated because of evidence suggesting that the biological behaviour of human malignant melanoma might be affected by steroid hormones, Malignant melanoma is extremely rare before puberty. The incidence of melanoma in southern England is 3.5 per 100 000 population per year. Yet in a population of 3 million, Bodenham1 reported only 2 cases before the age of 15. SadofflO found that the peak incidence of melanoma in females occurred during the childbearing period (third decade) and menopause (fifth decade). There has been considerable controversy concerning the possible deleterious effect of pregnancy on the course of melanoma. Early studies stressed the rapidly fatal outcome observed in pregnant patients with melanoma while other studies have failed to substantiate this." Shiu et al.12 found that of all female stage II patients, those who were either pregnant or experienced an activation of their tumour during previous pregnancies had a shorter survival. Allen4 described a patient whose melanoma metastasised widely during pregnancy and then underwent spontaneous complete regression within eleven weeks of delivery. Despite these factors, when male and female patients with melanoma are compared, the survival of females regardless of stage or treatment has been significantly longer. 5 13 14 Although most attempts at endocrine manipulation6 have been unsuccessful in malignant melanoma," occasional patients have responded.Johnson et al." reported an 11% objective response rate in metastatic malignant melanoma treated with 6o(-methylpregn-4ene-3, 11, 20-trione, a drug with glucocorticoid and
anti-oestrogen activity. Studies in human breast
cancer
have shown that
analysis of cestrogen receptors in tumour specimens correlates significantly with the likelihood of an objective response following hormonal manipulation. If a breast tumour is cestrogen-receptor-positive, the response rate to a variety of additive or ablative therapies range from 50 to 70%.6 However, for patients lacking oestrogen receptor, the response rate is generally less than 10%. Some studies in breast cancer have also suggested that both androgens and progesterone’9 receptors can increase the likelihood of predicting clinical responsiveness. Similar studies have suggested that glucocorticoid receptor analyses may aid in selecting patients with leukaemia who have an increased likelihood of respond ing to steroid therapy.20 21 Further studies will be neces sary to determine whether a subset of patients witl malignant melanoma, who have a higher likelihood o response to endocrine manipulation, can be selected b hormone-receptor analysis. Requests for reprints should be addressed to R.I.F., Building 10 Room 12N226, National Cancer Institute, Bethesda, Maryland 20014 U.S.A. REFERENCES
Bodenham, D. C., Hale, B. in Endocrine Therapy in Malignant Disease (edited by B. A. Stoll); p. 377. London, 1972. 2. Pack, G. T., Scharnagel, L. M. Cancer, 1951, 4, 324. 3. Conybeare, R. C. Obstet. Gynec., N.Y. 1964, 24, 451. 4. Allen, E. P. Br. med. J. 1955, ii, 1067. 5. DeVita, V. T., Fisher, R. I. Ca. Treat. Rep. 1976, 60, 153. 6. McGuire, W. H., Carbone, P. P., Sears, M. E. in Estrogen Receptors Human Breast Cancer; p. 1. New York, 1975. 7. Lippman, M. E., Huff, K. K. Cancer, (in the press). 8. McGuire, W. L., DeLaGarza, M. I. J. clin. Endocr. Metab. 197., 36, 548. 9. Scatchard, G. Ann. N.Y. Acad Sci. 1949, 51, 660. 1.
Binding of PH] oestradiol (upper graph).
to
receptor in human malignant melanoma
The binding data in the upper graph are replotted in graph using the Scatchard technique. Kd=Equilibrium dissociation constant, 4.9x 10-9 mmol/l. B/F=Bound divided by free [H]oestradiol concentration.
the lower
339
COTRIMOXAZOLE AND FOLATE METABOLISM M. C. BATESON*
Gordon Hospital and Department of Therapeutics, Westminster
Hospital, London P. PENDHARKAR
J. P. L. A. HAYES
Department of Hœmatology Westminster Medical School, London
Cotrimoxazole 4 tablets daily (1 tablet = trimethoprim 80 mg and sulpha400 methoxazole mg) was given for a period of six to fourteen days to 13 inpatients, and serum-folate levels were measured before and one day after the course of treatment. The results were compared with those from 8 patients not receiving antibacterial therapy, tested on admission and one week later. Two assay techniques were used, one employing Lactobacillus casei and the other 125I-labelled folate isotope dilution. The microbiological technique showed a significant decline in folicacid levels in the serum after cotrimoxazole, and this decline was not seen in controls. By contrast, the radioisotope technique showed no significant alteration in serum-folate levels compared with controls. This suggests that cotrimoxazole does not depress true serum-folate and that many low microbiological results obtained during cotrimoxazole therapy reflect interference with the assay organism. There is insufficient evidence to incriminate cotrimoxazole as a significant cause of blood dyscrasias in excess of those which might occur on sulphonamide alone or even with other antibacterials.
Summary
Introduction MUCH has been written on the interference by cotrimoxazole with bone-marrow function and folate metabolism. This work has taken the forms of case-reports of blood dyscrasias,1-13 the assay of serum-folate in volun68 9 marrow biopteers and patients on cotrimoxazole,5 sies in patients on cotrimoxazole,2-59 and reports to th Committee for Safety of Medicines. Cases where standard doses of cotrimoxazole have been administered and some subsequent disturbance of folate metabolism or hxmatological change reported have included only one certain case of megaloblastic anaemia without other likelv caused This occurred after a year’s course of the drug *Present address
Department
of Medicine, Dundee
University.
and the serum-folate (bacterial) method was normal. Various changes such as anaemia, neutropenia, thrombocytopenia, and megaloblastosis of the marrow have been reported in patients on cotrimoxazole. Some workers have suggested that serum-folate levels fall on a sulphonamide/trimethoprim combination, 14-16 others that they remain normal. 17 Against reports in several patients of megaloblastic changes on cotrimoxazole must be set the finding in a controlled thirteen-week trial that 11 patients on cotrimoxazole did not have megaloblastosis, whereas 1 on placebo did.4We have assessed the validity of the microbiological technique of folate assay after the exhibition of cotrimoxazole by using the isotope dilution technique. 18-20 Methods
patients on standard courses of cotrimoxazole for urinary trior lower respiratory infections 4 tablets daily (1 tablet methoprim 80 mg and sulphamethoxazole 400 mg) for six to fourteen days had the following investigations immediately before and on the day after completing therapy: (a) Hb, white blood-cells, mean corpuscular volume (Coulter S), (b) platelets (Coulter ’Thrombocounter’) (c) Serum-vitamin-B12 (Euglena gracilis21) (d) Serum-folate (Lactobacillus casei) (e) red-bloodcell (R.B.c.) folate (L. casei) (j9 Serum folate (isotope dilution Bio Rad ’Quanta-Count’ Kit’9). 8 controls on no antibacterial or antifolate agents had the same investigations at a seven-day interval. Patients were generally studied immediately after hospital admission. To normal serum of a known folate activity were added serial dilutions of trimethoprim. The apparent serum-folate activity was then re-estimated. 13
=
Differences in ’folate levels were submitted to Student’s t tests, and correlation coefficients calculated (table i). 5 of the 13 trial patients had a final serum-folate (L. casei assay, twelve to twenty-four hours after the last dose of cotrimoxazole) less then lg/1; this was calculated as a value of 1 p.g/1. R.B.C. folate levels did not change.
paired
Results theradioisotope dilution technique there moderate fall in values for serum-folate in both the control patients and the patients on cotrimoxazole. A similar fall, of 25-30%, was seen in the microbiogically estimated results for the control patients. The initial radiofolate assay values correlate very well with the microbiological ones for both controls (r=0-76; and cotrimoxazole patients P<0-001) (r=0’75 P<0.001). The paired results for the radiofolate assay before and after therapy also correlate well for control(r=0.76, p<0-05) and cotrimoxazole patients (r=0.69
Judged by
.was a
p
results in the control pa well correlated as well (r=0.78, p<0-001) However, the paired microbiological results for the pa tients
10 Sadoff, L, Winkley, J. Tyson, S. Oncology, 1973, 27,
244.
11 George, P. A., Fortner, J. G., Pack, G. T. Cancer, 1960, 13, 854.
were
TABLE I-SERUM-FOLATE LEVELS
12 Shiu, W. H., Schottenfeld, D., Maclean, B., Fortner, J. G. ibid. 1976, 37,
COTRIMOXAZOLE AND
181.
(jJLg/L) 8
IN
13 PATIENTS
ON
CONTROLS
13 White, L P New Engl. J. Med. 1955, 260, 789. 14 Einhorn, L. H., Burgess, M A., Vallejos, C., Bodey, G. P., Gutterman, J., Mavligit, G., Hersch, E. M., Luce, J. K., Frei, E., Freireich, E. J., Gottlieb, J A. Cancer Res. 1974, 34, 1995. 15 Nathanson, L., Hall, T. C., Farber, S. Cancer, 1967, 20, 650. Meyer, H W, Gumport, S. L. Ann. Surg. 1953, 138, 643. 17 Johnson, R. 0, Bisel, H. Andrews, N., Wilson, W. Rochhn, A., Segaloff, A., Krementz, E., Aust. J., Ansfield, F. Ca. Chemother. Rep. 1966, 50, 671. 18 Englesman. E, Korsten, C. B., Persign, J. R., Cleton, F. I. Br. J. Cancer, 16
1975, 30, 177 19 Horowitz, E. R, McGuire, W. L. Steroids, 1975, 25, 497. 20
Lippman, M. E., Halterman, R., Leverthal, B. G., Perry, S., Thompson,
E. B J clin Invest. 1973, 52, 1715. 21 Lippman, M. E., Perry, S., Thompson, E. B.
Am. J.
Med.
1975, 59, 224.
Mean
sampling interval in both groups=7, days
N.s.Not
significant.