Treatment of malignant carcinoid tumors with recombinant interferon alfa-2b: development of neutralizing interferon antibodies and possible loss of antitumor activity.

Twenty patients with malignant carcinoid tumors were treated for 6 months with recombinant interferon alfa-2b (IFN alpha-2b; Intron-A; Schering Corp., Bloomfield, NJ) at a mean dose of 5.9 megaunits three times per week. Eleven of the 20 patients (55%) had a greater than 50% reduction of tumor markers (urinary 5-hydroxyindoleacetic acid or plasma neuropeptide K), showing objective tumor response. Six patients (30%) had stable disease with no significant change in tumor markers or tumor size, and three (15%) had progressive disease with an increase in tumor markers and size. These results are similar to those reported earlier for treatment with natural leukocyte IFN in patients with carcinoid tumors. Only two patients (35%) had a slight reduction of tumor size after 6 months of treatment. Three patients developed neutralizing antibodies to IFN alpha-2b. Two of these patients initially showed an objective response, which lasted until IFN antibodies developed. In one of these patients, a change to human leukocyte IFN resulted in normalization of antibody titers within 3 months, and the patient had a second objective clinical response. There was no correlation between development of IFN antibodies and development of autoimmune phenomena such as increased titers of antinuclear antibodies or thyroid autoantibodies. IFN alpha-2b seems to be as potent as human leukocyte IFN in the treatment of patients with malignant carcinoid tumors, but it is important to recognize that antibodies neutralizing IFN may develop in some patients, with concomitant loss of antitumor effects. A change to natural leukocyte IFN might be beneficial in these patients.

Isolation of human pituitary prolactin.

A process developed earlier for the extraction of human follitropin, lutropin, thyrotropin and growth hormone from homogenized frozen pituitaries provided a residue utilized for the isolation of prolactin. The isolation procedure involved extraction at pH 9.8, molecular sieve chromatography on Sepharose CL-6B, hydrophobic interaction chromatography on phenyl-Sepharose CL-4B, molecular sieve chromatography on Sephadex G-100 Superfine, and ion-exchange chromatography on DEAE-Sepharose CL-6B using a convex gradient. The progressive purification was guided by radioimmunoassays. The final product was obtained in yields of 31 microgram/gland, and was equipotent with a pituitary preparation (VLS-3) supplied by the National Pituitary Agency (NIH, Bethesda, U.S.A.). Contamination hormones negligible (less than 0.05%). No heterogeneity of the isolated prolactin was observed by sedimentation-equilibrium analysis in the ultracentrifuge, by SDS electrophoresis in polyacrylamide gel or by molecular sieve chromatography in 6 M guanidine hydrochloride. These different techniques gave values in the range of 21 000-23 000 for the molecular weight of prolactin. In free zone electrophoresis, and also in polyacrylamide gel electrophoresis the prolactin preparation was, however, heterogeneous and resolved at alkaline pH into three distinct components. The former technique permitted isolation and assay of the components, indicating that they were all fully active.

Human pituitary thyrotropin. Isolation and recombination of subunit isoforms.

The alpha and beta subunits of human pituitary thyrotropin were prepared by a process developed earlier. Each subunit preparation contained four isoforms which were isolated by preparative agarose-suspension electrophoresis. The alpha subunit isoforms were homogeneous upon examination by analytical electrophoresis in polyacrylamide gel. Contrarily, the isolated beta subunit isoforms were always to some extent contaminated by the adjacent more acidic component. This was considered to show a transformation of these forms favoured by hydroxyl ions. Amino acid analysis revealed no differences between the alpha isoforms nor between the beta isoforms, and the amino acid compositions were in very good agreement with those earlier obtained for multi-alpha and multi-beta preparations. Attempts to recombine different alpha and beta isoforms were made at pH 7.0. The activities of recombined thyrotropin (in terms of MRC Research Standard A) ranged from 1.1 unit/mg (recombination of two minor components) to 3.5 units/mg (recombination of two major components). The activities of native and recombined thyrotropins as well as the subunit activities were followed by radioimmunoassays. The activity of native thyrotropin as a function of time was studied under different storage conditions. The subunit activities were also measured. All conditions resulted in a decrease in the thytrotropin activity (about 40% in 10 weeks). This decrease was not due to dissociation as the subunit activities remained approximately constant.

Human pituitary prolactin: isolation and characterization of three isohormones with different bioassay and radioimmunoassay activities.

Three forms of human pituitary prolactin, separable at alkaline pH in a highly purified preparation, were isolated by means of column electrophoresis in agarose suspension. The most acidic component showed a significantly lower radioimmunological activity but a higher bioactivity than the other two components, which were approximately equipotent in both assays. Consequently, in both assays the most acidic component differed markedly from the other components. Amino acid analysis indicated close similarity between the three components and no size heterogeneity was observed by sodium dodecyl sulphate electrophoresis in polyacrylamide gel. A high degree of purity of the isolated components was demonstrated by analytical electrophoresis in polyacrylamide gel at alkaline pH. Runs in the same medium also revealed that the difference in electrophoretic migration velocity remained after reduction and alkylation of the isohormones. A comparison of the migrations of the S-carboxymethylated and the S-carbamidomethylated derivatives with those of the unmodified components indicated that the diversity in electrophoretical behaviour of two adjacent isohormones was consistent with a difference in one single net charge. By analytical isoelectric focusing the component of intermediate migration velocity was resolved into two distinct bands proposed to contain isohormones differing only in the exchange of one residue of aspartic acid for one of glutamic acid. The isoelectric points of the prolactin isomers were all in the range of 5.7--5.9.

Isolation of five active thyrotropin components from human pituitary gland.

A procedure is described for the isolation of human pituitary thyroid-stimulating hormone (thyrotropin). The starting material was a side-fraction provided by the earlier developed process for the purification of growth hormone from whole frozen pituitaries. This fraction was further purified by successive chromatography on Bio-Gel P-150, Bio-Gel HT hydroxyapatite, and SP-Sephadex C-50. The resulting preparation was obtained in yields of 10 mg/kg of pituitary tissue and had a thyrotropin potency of 11 units human Research Standard A/mg as measured by a specific radioimmunoassay. Contamination by other pituitary hormone activities was low. In the ultracentrifuge a single sedimenting boundary was registered with an s20,w value of 2.7 S. The molecular weight as determined by sedimentation-equilibrium experiments was 34 000 in phosphate buffer, pH 7.0, and 17 700 in 1 M propionic acid. This thyrotropin preparation was, however, electrophoretically heterogenous. Following preparative polyacrylamide gel electrophoresis five different components associated with thyrotropin activity were isolated. Isolation on a preparative scale of electrophoretically homogeneous human thyrotropin has not earlier been reported. One of the thyrotropin components was characterized with respect to molecular weight and amino acid composition. The data were consistent with a molecular weight of 33 000 from sedimentation-equilibrium analysis at pH 7 and with 268 amino acid residues per molecule.

Human pituitary thyrotropin. Characterization of five glycoproteins with thyrotropin activity.

Human pituitary thyrotropin prepared by chromatography on hydroxyapatite or on SP-Sephadex was fractionated into five active components by preparative poly-acrylamide gel electrophoresis. The potency of the five components was 4-9 units human Research Standard A/mg. Examination of the components by analytical electrophoresis and by immunological methods revealed no heterogeneity. Ultracentrifugaiton of the three major components showed homogeneity with sedimentaiton coefficinets in the range of 2.4-3.0 S and a value for the molecular weight of about 33 000. Amino acid and carbohydrate analyses indicated close similarites between the five components.

Human pituitary thyrotropin. Isolation of alpha and beta subunits by hydrophobic interaction chromatography.

Various conditions for the dissociation of highly purified human pituitary thyrotropin into subunits have been investigated. Dissociation on a preparative scale was accomplished by treatment with 1 M propionic acid at 32 degrees C for 16 h. The isolation of one alpha and two beta subunits was achieved by hydrophobic interaction chromatography on pentyl-Sepharose-4B. Radioimmunological technique was utilized to classify the subunits in accordance with current nomenclature and also to express their activities. The activities of the subunits overlapped insignificantly (less than or equal to 0.3%) and their content of intact thyrotropin was negligible (less than 0.05%). The characterization of the subunits included determination of their amino acid compositions. Analytical polyacrylamide gel electrophoresis of the subunits at acid and alkaline pH values revealed heterogeneity. By free-zone electrophoresis at alkaline pH it was possible to isolate four discrete iso-forms of both the alpha and the beta subunit. All these eight individual subunits had activities consistent with those of their immediate precursor fractions. Isolation of electrophoretically homogeneous thyrotropin subunits has not been reported previously.

Human pituitary luteinizing hormone. Isolation and characterization of four glycoproteins with luteinizing activity.

Two major and two minor components of human luteinizing hormone (lutropin) were isolated from whole frozen pituitaries by a procedure involving extraction of homogenized pituitaries, (NH4)2SO4 fractionation, chromatography on DEAE-cellulose, Sephadex G-100, and SE-Sephadex C-50 and electrophoresis in polyacrylamide gel. The isolation procedure was monitored by both bioassays and radioimmunoassays. Contamination of the final products by other pituitary hormone activities was very low. The four lutropin components were all homogeneous by polyacrylamide gel electrophoresis (a sieving medium) and by free zone electrophoresis (a non-sieving medium). No heterogeneity was observed when the components were studied in the ultracentrifuge by sedimentation-equilibrium technique. The molecular weights of the components were in the range of 34 000-40 000. Sedimentation velocity experiments with the two major components revealed in each case one boundary with S20,W values of 3.2 S and 3.5 S. Further evidence for the homogeneity of the components was the observation of only one precipitin line for each component upon immunodiffusion against a rabbit anti-human lutropin serum. Amino acid and carbohydrate analyses indicated close similarity among the four components. From the analysis data the molecular weights of the components were calculated to be 31 000-33 000.

More basic forms of both human follicle-stimulating hormone and luteinizing hormone in serum at midcycle compared with the follicular or luteal phase.

The purpose of this study was to compare the heterogeneity and median charge of FSH and LH in serum at different phases of the normal human menstrual cycle. Serum specimens were obtained during the follicular phase [9.0 +/- 1.9 (+/- SD) days before the midcycle], at the midcycle LH peak, and during the luteal phase (9.75 +/- 1.6 days after the midcycle) in 16 women with normal menstrual cycles. The 48 serum specimens were subjected to electrophoresis in 0.17% agarose suspension in 0.075 M veronal buffer at pH 8.6, using a column suitable for measuring the median charge of the isoforms of FSH and LH, expressed as median electrophoretic mobility. Four sera were also analyzed by use of a larger column, giving a high resolution. The FSH and LH activities were measured with a time-resolved sandwich fluoroimmunoassay. The number of isoforms of both FSH and LH in each serum specimen analyzed by electrophoresis with high resolution was between 20-30. The median charge of the isoforms of FSH was less negative (P < 0.001) at the midcycle than in the follicular or luteal phase in all 16 women. The same was found for LH in 14 of 15 women. The median charge of FSH or LH in the follicular phase was not significantly different from that in the luteal phase of the cycle. We conclude that at least 20-30 isoforms of both FSH and LH circulate in blood during the menstrual cycle. More basic isoforms of both hormones appear in serum at midcycle than in the follicular or luteal phase. The difference is most likely due to a selective secretion from the pituitary of more basic forms of FSH and LH at midcycle.

Higher plasma disappearance rate in the mouse for pituitary follicle-stimulating hormone of young women compared to that of men and elderly women.

Female mice were injected iv with extracts of different human pituitaries and the recovery of hFSH in plasma and the disappearance rates of the hormone from the circulation were measured. In the first samples taken 5 min after injection, the recovery of pituitary FSH of young women (52%) was significantly lower (P less than 0.001) than that of men (64%) and elderly women (63%). The plasma concentration of human FSH was measured in the mice up to 242 min after injection. During this period the disappearance rate of the FSH of a young woman continued to be higher than that of the man and the elderly woman. The plasma disappearance curves were multiexponential and gave a poor fit in a two-component exponential model, probably due both to distribution into more than one compartment and to heterogeneity of the material investigated. The MCR and the t1/2 of irreversible loss of human FSH from the circulation were therefore also calculated from graphic integration of the areas under the disappearance curves extrapolated to 480 min. The values obtained in this way for MCR of FSH of a young and an elderly woman and a man were 2.4, 1.4, and 1.7 ml/h, respectively, and the values for t1/2 were 19, 33, and 28 min, respectively. The differences between these values for FSH of the three individuals were highly significant (P less than 0.001). The results indicate that the relatively low in vivo biological activity reported for pituitary FSH of young women compared to that of men and elderly women is due to more rapid clearance of the hormone from the circulation of the test animal.

Serum gonadotropin isoforms become more basic after an exogenous challenge of gonadotropin-releasing hormone in children undergoing pubertal development.

Recent clinical studies have questioned whether there is a qualitative change in the circulating isoforms of LH and FSH after stimulation by GnRH. The present study investigated the median charge of serum gonadotropin isoforms before and after an exogenous challenge of 100 micrograms GnRH in 10 girls and 10 boys undergoing pubertal development. All of the children had basal serum levels of LH and FSH and responses 30 and 60 min after GnRH treatment that were considered normal for puberty. LH and FSH in serum and eluates after electrophoresis in 0.10% agarose suspension were measured with sandwich fluoroimmunoassays. The increases in serum gonadotropin concentrations were generally similar for both sexes, but girls had significantly (P < 0.05) higher LH levels at 30 and 60 min than boys and a larger (P < 0.05) relative increase in serum FSH levels after GnRH treatment. In terms of the median charge of serum isoforms, the girls had significantly less negative (i.e. more basic) isoforms of LH (P < 0.05) and FSH (P < 0.001) than the boys. There was a change to more basic isoforms of both LH and FSH in all children 30 min after GnRH administration. For LH, the charge had returned to pretreatment values by 90 min after GnRH, but for FSH, the charge was still significantly (P < 0.05) more basic at this time (n = 4/sex). When the LH isoforms were more acidic before GnRH treatment, the change in median charge was larger than when the isoforms were more basic beforehand. A similar relationship was not found for FSH. Conversely, there was for FSH, but not for LH, a significant (P < 0.001) relationship between the relative increase in serum concentrations and the change in charge of the isoforms 60 min after GnRH treatment. These findings show that the circulating forms of LH and FSH become more basic after an exogenous challenge of GnRH in children undergoing pubertal development and suggest that the differences in the responses of LH and FSH isoforms may be due to differing degrees of selective secretion and/or survival.

Qualitative difference in follicle-stimulating hormone activity in the pituitaries of young women compared to that of men and elderly women.

The possible existence of qualitative differences between FSH in pituitaries from men and women of different ages was investigated with the use of an in vitro bioassay, an in vitro bioassay, and a RIA. Aqueous extracts were made from pituitaries obtained at autopsy and frozen until extracted. The FSH activities per pituitary and the ratios of FSH activities as obtained with the three assay methods were similar for young and elderly men. The ratios of in vivo biological to in vitro biological FSH activities were similar for men and postmenopausal women and significantly higher than the corresponding ratio for FSH from young women. With the in vivo bioassay the activity in the extracts of pituitaries from men and young women were similar, whereas extracts from postmenopausal women had significantly higher activity. With the in vitro bioassay the extracts from young women and elderly women had a similar content of FSH activity, whereas the FSH level in the extracts of male pituitaries was significantly lower. The results of the RIA correlated well with those of the in vitro bioassay. In conclusion, the results show that FSH in pituitaries from young women has a biological activity that is qualitatively different from FSH of men and elderly women. The relatively low in vivo biological activity of FSH from young women compared to FSH from men and elderly women was most likely due to a more rapid clearance of the hormone from the circulation of the test animal.

Hypersecretion of an abnormal form of follicle-stimulating hormone associated with suppressed luteinizing hormone secretion in a woman with a pituitary adenoma.

Pituitary tumors producing FSH have hitherto been reported only in males, all of whom have had normal or raised LH levels in serum. This report describes a female with a pituitary adenoma associated with supranormal serum levels of FSH. The FSH was also qualitatively abnormal when compared with FSH in the serum of other postmenopausal women, had a lower apparent molecular weight on gel chromatography, and was less negatively charged, as shown by electrophoresis. The results of LRH tests and suppression tests with ethinyl estradiol indicated autonomy of the FSH-producing adenoma. The FSH level increased concomitant with tumor enlargement and decreased after surgical removal of the pituitary adenoma or pituitary irradiation. The serum level of the glycoprotein alpha-subunit was raised about 100-fold. Any free FSH beta-subunits were not detectable in serum. The abnormal FSH had antigenic sites in common with both the alpha- and beta-subunits of FSH. The LH level was extremely low, and there was no response to LRH tests or ethinyl estradiol treatment. After gel chromatography, a small amount of LH, corresponding to 1/50th of the average for the patient's age, was detected at the position for normal LH. There was no GH response to insulin-induced hypoglycemia, while the cortisol increase was normal. Thyroid and adrenal functions were normal. The PRL level was within the normal range and increased slightly after estrogen treatment.

Change in electrophoretic mobility of human follicle-stimulating hormone in serum after administration of gonadotropin-releasing hormone.

The median electrophoretic mobility of FSH in serum (S-FSH) was determined in 14 girls, aged 9-14 yr, with Turner's syndrome before and after iv administration of GnRH. The basal S-FSH level varied almost 100-fold between patients. Less negatively charged forms of FSH appeared in serum in all patients after GnRH stimulation. The aim of the study was to determine if the change in electrophoretic mobility of S-FSH after GnRH treatment was related to the basal S-FSH level, the absolute or relative increase in the S-FSH level after GnRH, or the electrophoretic mobility of S-FSH present in the basal state. A highly significant (r = 0.94; P less than 0.001) correlation was found between the relative increase in S-FSH 60 min after GnRH treatment and the decrease in electrophoretic mobility. A selective survival of different forms of FSH in the circulation is proposed as the most likely explanation for the appearance in serum of the less negatively charged forms of FSH after GnRH stimulation.

More basic isoforms of serum gonadotropins during gonadotropin-releasing hormone agonist therapy in pubertal children.

An acute challenge of exogenous GnRH elicits rapidly increased serum gonadotropin levels with qualitative changes to more basic isoforms of both FSH and LH. Chronic GnRH agonist therapy suppresses endogenous gonadotropins, and the serum levels of FSH and LH are low and fairly constant. A possible qualitative change in the gonadotropins during GnRH agonist therapy was investigated by determination of the median charge of the gonadotropin isoforms before and during therapy in 18 pubertal children. Two different GnRH agonists were studied: buserelin, given intranasally or as a sc implant for 1.5-34 months to five girls, aged 7-10 yr, and for 5-6 months to two boys, aged 11-13 yr; and triptorelin, administered as a depot preparation for 3-6 months to four girls, aged 9-12.5 yr, and for 1-24 months to seven boys, aged 10.5-12 yr. FSH and LH in serum and eluates after electrophoresis in 0.10% agarose suspension were measured with sandwich fluoroimmunoassays. The mean serum FSH and LH levels decreased significantly (P < 0.05) in girls during triptorelin therapy, whereas only the FSH level decreased (P < 0.05) in the boys. There were no significant (P > 0.05) changes in serum gonadotropin levels during buserelin therapy. All of the children had more basic serum isoforms of LH, and all but one had more basic forms of FSH during the GnRH agonist treatments. In a girl who had more basic gonadotropin isoforms after treatment with triptorelin for 2 and 6 months, a GnRH challenge elicited the release of still more basic isoforms. The changes in mean median charge to more basic gonadotropin isoforms were highly significant for both busereline (P < 0.01) and triptorelin (P < 0.001) treatment. An increased (P < 0.001) degree of charge heterogeneity was observed for FSH after triptorelin therapy. These findings show that there is a qualitative change in the isoforms of both FSH and LH in serum during GnRH agonist therapy in pubertal children. The changes in charge to more basic gonadotropin isoforms most likely reflect a direct effect at the pituitary level, leading to the synthesis and/or selective release of less sialylated and sulfated isoforms of the gonadotropins. The observed qualitative changes in the gonadotropin isoforms in these pubertal children may be part of the clinical effects of GnRH agonist therapy, leading to an arrest or regression of puberty.

A change in the isoforms of human chorionic gonadotropin occurs around the 13th week of gestation.

hCG exhibits a considerable heterogeneity in blood during pregnancy. The overall charge of the isoforms of hCG has been shown to be more negative in the early than in the latter part of pregnancy. The present study analyzes the change in median charge and the charge heterogeneity as pregnancy progresses. The hCG activity in sera from 76 women from weeks 6-43 of gestation was measured with a noncompetitive time-resolved sandwich fluoroimmunoassay. The median charge and degree of charge heterogeneity of the isoforms of hCG in each serum was determined by electrophoresis in 0.10% agarose suspension. Median charge was expressed as median electrophoretic mobility. Electrophoresis with a high resolution revealed that the number of isoforms of hCG in a serum specimen from week 36 of gestation was 20-30. A change in median charge was found to occur at a limited time period of gestation, around the 13th week. All 16 sera from weeks 6-10 had isoforms of hCG with a more negative median charge than that of hCG in 21 sera from weeks 16-43 of gestation. The change in charge was accompanied by an increased degree of charge heterogeneity. There was a significant (P < 0.01) correlation between median mobility and the concentration of hCG during weeks 11-15 of gestation, but not before or after this period. There was no relationship between the sex of the fetus and the median charge of the isoforms of hCG. The data show that a change in the isoforms of hCG, revealed by a change in median charge of hCG, occurs at a limited time around the 13th week of gestation. This change occurs when the hCG concentration in blood decreases, and the placental production of estradiol and progesterone rapidly increases.

Reduced gonadotropin secretion in postmenopausal women during treatment with a stimulatory LRH analogue.

The potent and long-acting LRH agonist D-Ser(TBU)6-EA10-LRH was administered in a daily subcutaneous dose of 5 microgram to 5 postmenopausal women for a period of 10 days. The LRH analogue produced a significant decrease in both the basal FSH and LH levels and the gonadotropin responses to the agonist. The estrogen levels in serum remained unchanged during the study period. The results suggest that D-Ser(TBU)6-EA10-LRH has a direct inhibitory effect at the pituitary level.

Restored insulin sensitivity but persistently increased early insulin secretion after weight loss in obese women with polycystic ovary syndrome.

The impact of weight reduction on metabolic, endocrine, and anthropometric variables was studied in 13 obese, insulin-resistant women with the polycystic ovary syndrome (PCOS). Insulin sensitivity (euglycemic insulin clamp), insulin secretion and glucose tolerance (iv glucose tolerance test), basal sex steroid hormones, gonadotropins and free fatty acids (FFA), skin folds and waist hip ratio (WHR) were evaluated before (PCO-BD) and after (PCO-AD) diet-induced weight reduction to a weight stable level [mean (SD) diet duration 14.9 (6.2) months]. Mean weight loss was 12.4 kg (4.7; P < 0.0001), equalling a reduction from a body mass index (BMI) of 32.2 (3.7) kg/m2 to 27.6 (3.7; P < 0.0001) kg/m2. The results were compared with those of two groups of weight stable (no diet) women, 21 with PCOS (PCO-ND) and 23 normal control subjects (C), who were matched to the BMI the diet group reached after weight loss. Insulin sensitivity index (M/I) improved on average 132% (P < 0.001) and plasma FFA by 32% (P < 0.01), serum sex hormone binding globulin levels increased by 35% (P < 0.01), and the sum of truncal-abdominal skin-folds (subscapular, umbilical, and suprailiacal) were reduced by 28% (P < 0.0001), whereas the early insulin response to iv glucose, the levels of gonadotropins and androstenedione, and the femoral sc fat did not change significantly with weight loss. M/I, levels of SHBG and FFA and truncal-abdominal fat reached levels similar to the controls, whereas PCO-ND had lower M/I (P < 0.01) and SHBG levels (P < 0.0001), greater concentrations of FFA (P < 0.01) and truncal-abdominal fat (P < 0.05) than C. Among women with normal glucose tolerance, the insulin increment was higher in both PCO-AD (P < 0.05) and PCO-ND (P < 0.01) than in C. There was a strong correlation between M/I and sum of truncal-abdominal skinfolds in all groups (PCO-BD: r = 0.82; P < 0.001, PCO-AD: r = 0.68; P < 0.05, PCO-ND: r = 0.81; P < 0.0001, C: r = 0.44; P < 0.05). The variation in M/I in PCO-AD and PCO-ND (pooled) was best explained by FFA and truncal-adbominal fat (model R2 = 0.67). In conclusion, insulin resistance in obese women with PCOS was reduced by weight loss to similar levels as BMI-matched control subjects, suggesting that insulin resistance in PCOS is not a feature of PCOS per se.(ABSTRACT TRUNCATED AT 400 WORDS)

Changes in the isoforms of luteinizing hormone and follicle-stimulating hormone during puberty in normal children.

Concentrations of LH and FSH are known to increase during normal pubertal development, but changes in the isoforms of the gonadotropins at this time have not been investigated in depth. We examined the median charge of serum LH and FSH using agarose suspension electrophoresis in 81 normal children at pubertal stages I-V. In pubertal girls there were no significant (P > 0.05) differences in the median charge of LH, but there was a small (P = 0.05) shift to more acidic FSH isoforms between pubertal stages I and IV. In boys there was a significant (P < 0.01) shift to more acidic isoforms for both LH and FSH by pubertal stage II. Further changes were not found later in puberty. Except for LH at pubertal stage I, where the median charge was similar (P > 0.05) for both sexes, the median charge was more basic (P < 0.001) for both LH and FSH in girls compared with boys at all five pubertal stages. The degree of charge heterogeneity of FSH, estimated as the peak width at half the peak height, was significantly (P < 0.01) larger at pubertal stage I than at pubertal stages III-V in both boys and girls. The charge heterogeneity of LH was similar for all pubertal stages in both sexes. In conclusion, there were few qualitative changes in the gonadotropins during normal female puberty, whereas in the male there was a dramatic shift to more acidic isoforms of LH and FSH early in puberty. This information may assist our understanding of normal and pathological processes during puberty and may be of clinical relevance in detecting the initiation of puberty in boys.

Thyrotropin-releasing hormone testing during antithyroid drug treatment of Graves' disease as an indicator of remission.

In 74 patients with hyperthyroid Graves' disease, TRH tests were undertaken every third month during the course of a standardized antithyroid drug and T4 treatment program. The antithyroid drug dose was reduced gradually and finally withdrawn when persistently normal TSH responses were obtained. In 46 patients (62%), such normal responses occurred and therapy was discontinued after a mean treatment period of 13 months (range, 5-24 months). In the remaining unresponsive 28 patients (38%), therapy was gradually withdrawn after 2 yr of treatment (mean treatment period, 27 months; range, 25-36 months; P = 0.0001 vs. the other group). The mean overall follow-up period after cessation of treatment was 65 months (range, 32-100 months) and did not differ between the TRH-responsive and TRH-unresponsive group. In the TRH-responsive group, 12 relapses (26%) occurred 23 months (range, 6-45 months) after discontinuation of therapy, in contrast to 20 relapses (71%) after 6 months (range, 0-12 months) in the TRH-unresponsive group. The differences in relapse rates and time duration until relapse are highly significant (P = 0.0003 and P = 0.0001, respectively). Small but significant differences in serum T3 and T4 levels were found between the groups throughout the treatment periods, emphasizing the importance of thyroid hormone levels in regulating the pituitary responsiveness to TRH. It is concluded that regular TRH tests during antithyroid drug treatment are useful in deciding the dose and duration of therapy and in predicting the likelihood of remission.