Search

Insulin treated patients frequently develop insulin-binding antibody. Antisera from some patients, treated with bovine or bovine/porcine insulin mixtures, react differently with bovine and porcine insulin while for others there is no difference: when there is a difference there is greater avidity for bovine than porcine insulin. We studied antisera from 16 patients who had previously been treated with bovine insulin and then were changed to porcine insulin. Dissociation rate constants and association rates were measured with bovine and porcine insulin. Significant differences were found in association rates, which correlated well with binding at equilibrium (r=0.91), but not in dissociation rates. For 11 subjects who stayed on highly purified porcine insulin, long term reduction in insulin requirement correlated with the magnitude of the difference in antibody reaction to bovine and porcine insulin (r=0.74).

Infusion of growth-hormone release inhibiting hormone (G.H.-R.I.H.) in four fasting subjects reduced plasma glucagon and insulin concentrations to undetectable levels and this was associated with a highly significant decline in plasma-glucose (28±S.E.M. 3 mg. per 100 ml. in 1 hour, p

Clinical, laboratory and radiological findings were evaluated in twenty-nine men who had raised serum prolactin concentrations and pituitary tumours. Twenty-one had functionless pituitary tumours ('prolactinomas') and eight had acromegaly. Supraseller extension was detected in twenty of the twenty-six men who had lumbar airencephalography. Three patients were studied before, sixteen before and after and ten only after pituitary ablative therapy. Seventeen of these men complained of complete lack of libido and impotence and six had impaired libido and sexual potency; only six patients in this series denied reproductive symptoms. Thirteen of the impotent subjects had small soft testes, ten reduced facial and body hair and three had marked gynaecomastia. No features of hypogonadism were noted in the six patients without reproductive symptoms and none of the patients had galactorrhoea. Serum prolactin concentrations were higher and serum testosterone concentrations lower in the impotent men compared with those with normal sexual potency. Serum LH and FSH (both basal and in response to LHRH) oestradiol and oestrone concentrations were not different between the two groups and, except in those with post-operative hypopituitarism, were within the normal range. Following successful lowering of prolactin concentrations by surgery or bromocripitine or both, serum testosterone rose and potency returned; by contrast failure to lower prolactin concentrations was associated with persistent impotence and hypogonadism. The endocrine profile of low serum testosterone concentrations with gonadotrophins which had not risen into the range usually seen in primary hypogonadism (together with the parallel increase of LH and testosterone in one patient studied sequentially during treatment which suppressed prolactin levels to normal), suggested that the impaired gonadal function was caused by a prolactin-mediated disturbance of hypothalamic-pituitary function.

Hypoglycaemic and growth hormone responses were studied at different steady-state plasma insulin concentrations during a graded infusion of monocomponent human insulin. The control group consisted of ten volunteer subjects. The other groups studied included women taking oral contraceptives and patients with obesity, thyrotoxicosis, myxoedema, acromegaly, diabetes mellitus (moderate and severe) and liver disease. The hypoglycaemic response was measured in two ways: (i) the percentage reduction in plasma glucose below basal, and (ii) the rate of fall of plasma glucose (Kg-%/min). Insulin sensitivity was greatest in the normal subjects and in the other groups decreased in the order thyrotoxicosis greater than oral contraceptive greater than obesity greater than myxoedema greater than acromegaly greater than liver disease. Insulin sensitivity was difficult to assess in the diabetic patients because basal plasma glucose concentrations were elevated. At any given insulin concentration, the diabetics metabolized approximately the same amount of glucose as the normal subjects but the fact that this rate of glucose turnover occurred at higher plasma glucose concentrations probably indicated insulin resistance. Within each group Kg at each dose level of insulin correlated with the steady state plasma insulin concentration during the same infusion period. Diminishing sensitivity to insulin was reflected in an increasing fasting plasma insulin and insulin/glucose ratio except in patients with diabetes. GH responses to insulin infusion in normal subjects reflected the pattern of fall of plasma glucose. In the diabetic patients GH secretion appeared to be related to the infusion of insulin and occurred before plasma glucose had fallen to hypoglycaemic levels. GH secretory patterns were within normal limits in women taking oral contraceptives and in seven of eleven patients with liver disease but were impaired in three of seven patients with thyrotoxicosis and four of five patients with myxoedema. Four obese patients had a markedly delayed but eventually normal GH response.

Basal serum prolactin and growth hormone (GH) concentrations were measured by radioimmunoassay in forty patients with acromegaly. GH concentrations were elevated in all patients studied before treatment and prolactin levels were raised in seven of twenty-six patients. Of the thirty-two patients reviewed after treatment (which in most cases was transsphenoidal hypophysectomy) twenty-five had GH concentrations below 5 ng/ml and twenty-nine had normal prolactin levels. In eighteen patients hormone measurements were made both before and after hypophysectomy: though GH levels fell in all but one, prolactin fell in only six patients. They were not significantly changed in eleven patients. There was no correlation of GH and prolactin either before or after surgery. Seven patients had greatly elevated prolactin levels and in four of these there was evidence of upward extension of a pituitary tumour on air encephalogram (AEG). Only one patient with a normal prolactin level had an abnormal AEG. Two patients with elevated prolactin concentrations and normal AEGs had a parallel fall of prolactin and GH in response to surgery. Four of the five hyperprolactinaemic men complained of loss of libido: in three gonadotrophin concentrations before and after treatment were normal. We conclude that there is no overall correlation of GH and prolactin levels in patients with acromegaly. Seven of twenty-six untreated patients (27%) had hyperprolactinaemia. We suggest that in these patients a raised prolactin level may be due either to interference with the normal inhibitory control mechanism of prolactin by suprasellar extension or, more rarely, to secretion of both GH and prolactin by the tumour itself. A high prolactin concentration may be the cause of the impotence of which some patients with acromegaly complain.

SUMMARY The response to the subcutaneous injection of glucagon (1 mg) has been studied as a test of the ability of the pituitary to secrete ACTH and growth hormone (GH). Tests have been performed on fifteen normal subjects or patients with unrelated conditions, nine patients with pituitary tumours and seventeen with acromegaly. It is suggested that a normal response is a peak level of 5 ng/ml or more of GH (standard NIH GH 1216C) and either a rise of plasma cortisol measured by competitive protein binding of at least 4 μg/100 ml or a peak level of 7 μg/100 ml. In seventeen patients the responses were compared with those obtained in the insulin hypoglycaemia test and there was good agreement. Measurement of plasma ACTH showed that the response is at pituitary and not adrenal cortical level. Peak responses of ACTH and GH were obtrained at 150 min and of cortisol at 180 min. Plasma glucagon concentrations were falling steadily by this time although still above the control level. The mechanism of action of glucagon is not clear but does not appear to be related to fall of blood sugar. The subcutaneous injection of glucagon is not without side effects; about half the patients were nauseated and nine vomited. Nevertheless we believe that a subcutaneous glucagon test is of value for testing the pituitary reserve of ACTH and GH in situations where insulin hypoglycaemia cannot be used.

Highly purified insulins offer the possibility of reducing insulin antibody levels and insulin requirement. Those likely to respond cannot be predicted on clinical grounds and a simple laboratory test is recommended for this purpose. This is based on insulin binding capacity (IBC) of plasma and has been used to follow a group of 47 patients over six months. 47 patients previously treated with British soluble and isophane insulins were changed to highly purified Leo Neutral and Leo Retard. 36 showed a reduction of insulin requirement and two groups could be identified. An IBC of greater than 40 muU/ml was associated in 94% with a reduction of insulin dose; if the IBC was less the response was unpredictable. The initial IBC was related to the initial insulin requirement and to the eventual percentage reduction. Serial measurements in patients with high initial IBC showed a steady fall. Measurement of the IBC is a simple investigation; if the level is greater than 40 muU/ml changing to a highly purified insulin is likely to be associated with reduction of insulin requirement.

A group of 14 patients with skin reactions to insulin have been treated with highly purified insulin over a period of 5--14 months while continuing to inject into the abnormal areas. Fat atrophy improved in all cases, coincident with a reduction in insulin requirement and insulin binding capacity. The response of patients with fat hypertrophy was less predictable, but there was an improvement in 3 out of 6 cases. Measurements of insulin binding capacity of the plasma suggested that fat atrophy and hypertrophy are immunologically distinct and some explanations for this are discussed. It is concluded that highly purified insulin benefits some patients with skin reactions to insulin.

Summary Results of treatment of 52 patients with amenorrhoea associated with hyper-prolactinaemia are presented. All patients had a detailed radiological examination of the pituitary fossa, including lateral tomography in every patient and air encephalography in those in whom a pituitary tumour was suspected. There were 17 patients with untreated pituitary tumours, 5 patients with previously treated pituitary tumours and persisting hyperprolactinaemia, and 30 patients with normal pituitary radiology. Patients with pituitary tumours were treated either by transsphenoidal or transfrontal surgical extirpation of the tumour, followed, if necessary, by external irradiation and/or bromocriptine. Four patients were treated with external irradiation as primary therapy, and three patients who did not wish to conceive were treated with bromocriptine as primary therapy. Patients with normal radiological appearances were treated with bromocriptine as primary treatment. Ovulatory menstrual cycles developed in 42 patients and there were 19 pregnancies. Those ovulating but not conceiving had adequate non-endocrine factors to account for the disparity. Failure of response was seen in 10 patients and was due to inadequate fall of prolactin in response to surgery (2 patients), external irradiation (3 patients) and bromocriptine (1 patient), and gonadotrophin deficiency which developed after surgery in 3 patients but was present pre-operatively in 1. The relative merits of treatment by surgery, external irradiation and bromocriptine are discussed and a policy of treatment outlined.

Summary A series of 26 patients diagnosed as having an insulin-secreting, islet-cell tumour of the pancreas is described. Three patients were treated medically, but in the remaining 23 the tumour was found at operation; 1 patient had a carcinoma with hepatic metastases, but in the other 22 the tumour was apparently benign and the subsequent course of these patients has confirmed this conclusion. The tumour was found at the initial operation in 19 of the 22 patients with a benign tumour treated surgically. Of the other 3, the tumour was removed at a second operation in 2 and at a third operation in 1. Two of the 22 patients died following the operation. Diazoxide was administered to 21 patients, before and/or after operation. It was given preoperatively to 17 patients, in all of whom it prevented the development of hypoglycaemic symptoms, but in 2 it caused serious fluid retention and had to be discontinued. Seven patients have been successfully treated with long term diazoxide, either as an alternative to surgery or following an unsuccessful operation, 1 of these being the patient with an islet-cell carcinoma. In these patients diazoxide has been given for up to 12 years without apparent side effects. Selective coeliac angiography was performed in 13 patients. In no patient did it provide critical information that significantly influenced the treatment of the patient. Percutaneous, transhepatic portal venous catheterization was performed in 4 patients and indicated the site of the tumour in 3, 2 of whom had had 1 previous operation and 1 had had 2 previous operations, at none of which was a tumour found: in all 3 patients the tumour was removed at a subsequent operation. The application of these findings to the management of patients with an insulinoma is discussed and a scheme of management is proposed.

Comparisons are made between the incidence, prognosis and treatment of juvenile-onset diabetes and other endocrinopathies in the young. 548 patients with insulin deficient diabetes diagnosed before 20 years of age have been reviewed. Excess mortality, especially at 35--40 years of age was found. Profiles of blood glucose and serum insulin have been studied and compared to those of normal subjects. The variation of insulin absorption and effect of insulin antibodies on the free insulin levels achieved after exogenous insulin injections have been demonstrated. The common occurrence of nocturnal subclinical hypoglycaemia following intermediate or long-acting insulin was often found to be the cause of poor diabetic control. Five out of 33 patients with 'difficult' diabetes had an unexplained resistance to high levels of free-insulin. The value of self-monitoring and HbAl measurements in the improvement of diabetic control and possibly life expectation is reviewed. The incidence of thyroid disease was found to be increased in 1779 insulin deficient diabetics of all ages and persistence of islet-cell antibodies suggests that the diabetes may be due to autoimmunity in some of these patients.

In patients with hyperosmolality as the sole cause or as an important factor in the production of the coma, therapy with hypotonic or isotonic fluids and with insulin, often in amounts less than those generally used in diabetic ketoacidosis, has proved effective in restoring consciousness in many, but not all, instances of hyperosmolar coma. The mortality, however, has been high, usually because of uremia, stroke, massive infection, pancreatitis, etc.

A method for the quantitative assessment of rapid changes in the delivery rate of hormones into the systemic circulation is reported. Plasma hormone levels are examined in relation to the “non-steady-state” disappearance rate measured following an intravenous infusion of the hormone. The insulin delivery rate in response to iv glucose has been measured; a biphasic insulin response has been confirmed; and the amount of insulin entering the systemic circulation in each phase can be examined.

The plasma half-life of exogenous unlabelled porcine ACTH (Organon) and porcine (synthetic 1–39) ACTH 131I has been compared with that of endogenous ACTH in the domestic pig. A sensitive radioimmune assay was used for measurement of exogenous and endogenous ACTH. With ACTH 131I, free undamaged hormone in the plasma was assessed by a technique using antibody binding and charcoal separation. The estimated half-life of porcine ACTH 131I and exogenous unlabelled porcine ACTH were found to be 13 and 28 min respectively. The half-life of endogenous ACTH measured following the abrupt termination of an insulin hypoglycaemic stimulus by intravenous glucose and dexamethasone was 7 min. Radioimmune assay of endogenous ACTH produced in physiological amounts was considered to be the best method of assessing the half-life of the hormone.

In recent years two comparatively simple methods have been described for the assessment of urinary glucocorticoid excretion. Reddy, Jenkins, and Thorn in 1952 gave details of a method for estimating urinary " 17-hydroxycorticoids." This involved extraction of the urinary glucocorticoids (cortisone and hydrocortisone) and their metabolites with n-butyl alcohol and the use of the Porter-Silber colour reaction. The second method is the measurement of urinary 17-ketogenic steroids (Norymberski, Stubbs, and West, 1953). This is done by oxidizing the glucocorticoids and their metabolites to 17-ketosteroids with sodium bismuthate and estimating the newly formed and pre-existing 17-ketosteroids by the Zimmermann reaction. If at the same time the 17-ketosteroid content of the untreated urine is measured, the increase following bismuthate oxidation is an index of the glucocorticoid content of the urine. For the investigation of patients suspected of having adrenal dysfunction, it would be of considerable value if the urinary glucocorticoid content could be assessed by a method suitable for use in a hospital biochemical laboratory. From the original descriptions of these methods either appears suitable for this purpose. We have used them in the investigation of patients with endocrine disease, after adrenalectomy and in other connexions, and have encountered certain difficulties. This has led us to modify the original Norymberski method and to adopt some of the features of the urinary 17-hydroxycorticoid estimation described by Smith, Mellinger, and Patti (1954). Using these modifications we have tried to ascertain whether these methods really are an index of urinary glucocorticoid excretion. The 17-hydroxycorticoid and 17-ketogenic steroid content of more than 400 urines have been measured and the results compared. Recoveries of cortisone and hydrocortisone added to urine have been studied and the glucocorticoid excretion of adrenalectomized patients receiving different amounts of hydrocortisone and of cortisone acetate have been measured. URINARY 17-HYDROXYCORTICOIDS The first difficulty encountered with this method was in the purification of n-butyl alcohol. The methods suggested by Reddy (1954) and Smith et al. (1954) proved disappointing. We are indebted to Mr. R. W. H. Edwards, B.Sc., of the Courtauld Institute of Biochemistry of the Middlesex Hospital, for suggesting the method described in the next section. The other difficulty was with the Porter-Silber reaction. 17, 21-Dihydroxy-20-ketosteroids produce a yellow colour with phenylhydrazine and strong sulphuric acid, but when the reaction is applied to urine extracts allowance must be made for the colour produced by strong sulphuric acid without phenylhydrazine. Like Smith et al. (1954) we found that the 62% sulphuric acid used in the original method gave very high readings; their modification using 56% sulphuric acid and allowing longer time for the colour to develop has proved more satisfactory.

Intravenous glucose and xylitol tests have been performed on normal, diabetic and obese subjects. In normal and obese subjects glucose is considerably more effective than xylitol in stimulating the first phase of insulin release. In diabetic subjects the initial response to xylitol is not reduced to the same extent as the response to glucose, and the 2 stimuli are equally effective in stimulating the reduced first phase response. Xylitol and glucose equally stimulate the second phase in all groups, but xylitol gives a less prolonged response. Neither xylitol nor glucose stimulates either phase in severely diabetic subjects. The equal stimulation of the second phase suggests xylitol and glucose may share a common stimulatory mechanism, possibly via the pentose phosphate pathway. The less prolonged xylitol response is probably due to lack of stimulus to insulin synthesis. Glucose appears to stimulate the first phase by a mechanism which is not equally available to xylitol; a deficiency of this “glu...

Plasma testosterone values in the basal state were within normal range (0.03–0.20 μg/100 ml) in 40 women with idiopathic hirsutism. Corticotropin stimulation tests were done in 8 patients; in 7, plasma testosterone increased significantly, and 1 patient tested at 4-hr intervals after 1 injection of ACTH gel showed maximal increase at 4 hr. Stimulation tests resulted in elevation of plasma testosterone in 3 of 4 nonhirsute female control subjects. Dexamethasone suppression tests resulted in significant reduction in plasma testosterone in 5 of 9 patients. HCG was given to 10 patients. No response in plasma testosterone was seen in the first 24 hr after 1500 U was given to 2 patients on 3 successive days, or after 12,000 U to 1 patient on 2 successive days, the adrenals being suppressed in all 3. A single injection of 24,000 U given to 1 patient while adrenal suppression was maintained resulted in a significant rise only after 72 hr. A single injection of 12,000 U was given, without adrenal suppress...

Aldosterone secretion rates have been measured in six patients with cirrhosis of the liver using 16–3H aldosterone. The proportion of administered radioactivity excreted as pH 1 extractable conjugates and as a glucuronide has also been determined. On unrestricted diets the aldosterone secretion rates were 19, 85 and 94 μg/day in three patients with ascites and oedema, and 140, 170 and 386 μg/day in three without fluid retention. Five of these patients were restudied after at least seven days on a low sodium diet (20 – 30 meq./day) and slight increases of secretion rate observed. The levels obtained were much lower than those previously reported for normal subjects on a low sodium diet. One patient was studied on a subsequent occasion after a haematemesis (urine sodium 1 meq./day), the secretion rate was 1130 μg/day. The proportion of administered radioactivity excreted as glucuronide was below normal in at least one determination on four of the six patients and there was an increase of the proportion excreted as pH 1 extractable conjugates. There was a striking variation of the proportion of radioactivity excreted as pH 1 extractable conjugates. One patient was studied on five occasions and the range was from 6.0 to 25.2 % of the administered radioactivity. This suggests that in patients with cirrhosis of the liver the urinary pH 1 extractable aldosterone may have a variable relationship to the actual secretion rate.

Infusions of cortisol (11β,17,21-trihydroxy-pregn-4-ene-3,20-dione), prednisolone (11β,17,21-trihydroxy-pregna-1,4-diene-3,20-dione), aldosterone (11β,21-dihydroxy-3,20-dione-pregn-4-en-18-al), and adrenal cortical extract have been given to four adrenalectomized dogs. The changes of inulin, creatinine, para-aminohippuric acid clearances, urine volume, and sodium and potassium excretion have been measured. Between experiments oral replacement therapy was given to obviate the effect of slow absorption of previously injected steroids. The effects of steroid-free control infusions have been studied. Varying the rate of infusion from 1.0–2.6 ml/min, and sodium from 0–170 μeq/min has little effect on glomerular filtration rates (G. F. R.). Infusions of cortisol (1.5 and 10 mg/h), prednisolone (2 mg/h) and adrenal cortical extract (10 ml/h) raised the G. F. R. and lowered the filtration fraction. As the G. F. R. increased the ratio creatinine: inulin clearance fell. The different dogs varied in their responses. Infusions of prednisolone and cortisol (1.5 mg/h) usually increased sodium output as G. F. R. rose. With aldosterone, adrenal cortical extract and cortisol (10 mg/h) there was sodium retention in three of the four dogs, the fourth was resistant to the sodium retaining action of these steroids while the plasma level was high. All steroid infusions increased potassium excretion. Cortisol increased free water clearance independent of G. F. R. provided the infusion rate exceeded 2 ml/min, prednisolone increased both osmolar and free water clearance, aldosterone increased osmolar clearance and reduced free water clearance.

Stimulation tests have been performed on 26 patients with normal adrenal function, 23 hirsute women, 13 patients with Cushing's syndrome (liypcrplasia 11, adenoma 1, carcinoma 1), and 3 patients with hormone-secreting adrenocortical carcinoma without Cushing's syndrome. Suppression tests have been performed on 11 patients with Cushing's syndrome (hyperplasia 9, adenoma 1, carcinoma 1). Corticotropin-gel was used for stimulation and α fluorohydrocortisone for suppression. Response was assessed from changes in urinary 17-hydroxycorticosteroid and 17-ketosteroid excretion. In the stimulation tests an abnormal increase of urinary 7-hydroxycortieosteroid excretion was noted in 1 obese male, 1 normal male, 15 hirsute females and 7 of the patients with Cushing's syndrome due to adrenocortical hyperplasia. There was a moderate increase of urinary 17-hydroxycorticosteroids in the patient with an adenoma and a very slight increase in the patient with Cushing's sjaidrome due to a carcinoma. In 2 of the patients with...

This study was planned to investigate the pathogenesis of diabetes in acromegaly. Standard intravenous glucose tolerance tests were done with measurements of blood sugar, plasma insulin, growth hormone and free fatty acids. Sixteen patients were studied, and 10 had more than one glucose tolerance test. Seven of the patients had impaired carbohydrate tolerance, including 2 with symptomatic diabetes. In 6 patients serial tests showed the change that may occur as carbohydrate tolerance deteriorates, and in 5 the response to surgical hypophysectomy was observed. Fasting hyperinsulinism was present in 6, and insulin response to intravenous glucose was increased in 10. The pattern of this response varied—in 10 there was a peak in the first 15 min, in 4 the pattern resembled that of maturity onset diabetes, with a gradual rise of plasma insulin concentration toward a peak at 60 min. In the 2 patients with symptomatic diabetes, there was no response to intravenous glucose although the fasting plasma insu...

Images Figs. 1-4

Early "prophylactic" oophorectomy and adrenalectomy has been performed on 12 patients with carcinoma of the breast. The patients selected were considered to have a very bad prognosis on account of axillary node involvement associated with internal mammary chain deposits (9 patients), supraclavicular nodes (2 patients) and a parasternal mass (1 patient). Five patients had evidence of spread beyond the primary lymph drainage area (axillary and internal mammary nodes), and all had died within 4 years. In 7 patients the disease was confined to the primary lymph drainage area and 4 lived for more than 10 years, 3 being alive and well at 11 to 12 years. This is a higher proportion than in a control series but does not quite reach the level of statistical significance. In the 7 patients with disease confined to the axillary and internal mammary nodes prognosis was not apparently related to malignancy determined histologically, but did have an association with the extent of invasion of the axillary nodes. Urinary oestrogen estimations performed in 4 patients did not give any evidence that outcome was related to persistence of oestrogen production. Details of the patients' management and replacement therapy are given and from prolonged personal follow-up of these patients it is concluded that women who have undergone oophorectomy and adrenalectomy are able to lead full and active lives.

1. The metabolism of unlabelled monocomponent human insulin and porcine proinsulin was studied in ten normal subjects (five males and five females) by using a priming dose-constant-infusion technique. In each subject, the metabolic clearance rate (MCR) was measured at four separate steady-state hormone concentrations averaging 16–216 μunits/ml (insulin) and 4·2–42·8 ng/ml (proinsulin). 2. For insulin the MCR fell progressively from 34 ml kg−1 min−1 at a mean fasting insulin concentration of 3·8 μunits/ml to 11·4 ml kg−1 min−1 at the highest concentration achieved (280 μunits/ml); for proinsulin MCR averaged 3·7 ml kg−1 min−1 at a mean plasma concentration of 4·2 ng/ml and fell to 2·71 ml kg−1 min−1 at 10·7 ng/ml, remaining constant thereafter at concentrations up to 71 ng/ml. 3. The half-disappearance time (T½) from the plasma, after the end of the infusion, averaged 4·3 min for insulin and 25·6 min for proinsulin. 4. The apparent distribution space (DS) was similar for both hormones (83 ml/kg of insulin and 98·9 ml/kg of proinsulin). 5. There was a direct correlation between T½ and DS for both hormones. 6. Although the higher MCR of insulin was reflected in its shorter T½ there was, for each hormone, no relationship between MCR and T½. 7. The biological potency of porcine proinsulin, as judged by its effect on plasma glucose, was approximately 5% of that of insulin. 8. The responses of serum growth hormone and Cortisol were shown to be directly related to the degree of hypoglycaemia induced.

A method is described for the estimation of urinary testosterone excretion. After hydrolysis, the urine is fractionated on a Sephadex column, the appropriate fractions are extracted and the testosterone separated from other steroids by paper and thin-layer chromatography. The final quantitation is performed by sulpliuric acid fluorescence using a highly sensitive spectrophotofluorimeter. [14C]Testosterone is added to the urine after hydrolysis. It serves as a marker on the paper chromatogram and thin-layer plate and is used for correction of extraction losses. The specificity of the method has been confirmed by study of possible interfering steroids by bisection of the radioactive spots on the paper chromatogram and by biological data on normal men and women, hypogonad men, hirsute women and adrenalectomised, oophorectomised women. Satisfactory reproducibility and recovery of added testosterone glucuronide have been demonstrated. The method is extremely sensitive. Urine aliquots of 10 ml are used and it is possible to measure a urinary testosterone excretion of 1.0 μg per 24 h in a urine volume of one litre.

Serum growth-hormone (G.H.), insulin, and cortisol and plasma-glucose were measured serially in peripheral venous blood, before and during slow continuous infusions of mono-component human insulin in four patients with severe and nine patients with moderate uncontrolled diabetes mellitus. The results were compared with similar studies in nine healthy controls. In the controls serum G.H. and cortisol rose in direct proportion to the degree of hypoglycaemia achieved. In the diabetic patients there was a striking rise in serum G.H. and cortisol within 30 minutes of the start of the insulin infusion. In these patients the G.H. and cortisol responses did not seem to be related to a fall in plasma-glucose but to be directly related to the administration of insulin. No rise in serum-levels of thyroid-stimulating hormone was observed. The results are compatible with the hypothesis that the abnormal responses were due to direct stimulation of glucose metabolism within the pituitary gland, with subsequent release of G.H. and corticotrophin in response to pre-existing stimulation by G.H. and corticotrophin-releasing hormones. Slow infusions of small amounts of insulin were found to be highly effective in correcting the biochemical abnormalities of uncontrolled diabetes mellitus.

Serum-prolactin concentrations were measured in 111 patients who had radiological abnormalities of the pituitary but no evidence of acromegaly, Cushing's syndrome, or Nelson's syndrome. Raised prolactin levels were found in 45 of 64 (70%) patients studied before treatment and in 15 of 47 patients studied after pituitary surgery. The majority of hyperprolactinaemic patients presented with amenorrhœa or impotence; galactorrhœa was uncommon. By contrast, reproductive disorders were rare in patients with normal prolactin levels. It is concluded that pituitary tumours previously described as "functionless" are frequently associated with hypersecretion of prolactin and that such tumours usually present as reproductive disorders. In the investigation of patients with amenorrhœa or impotence serum-prolactin should be measured and skull radiology performed if the prolactin level is raised. Prolactin should be measured in all patients with abnormal pituitary X-rays both before and after pituitary surgery.

A case of factitious diabetes is reported. The patient gave herself insulin intermittently over a four year period. A high concentration of insulin binding antibody was found with the antibody showing much greater avidity towards beef than human insulin. There was resistance to intravenously administered beef insulin with low concentrations of free insulin and high concentrations of bound insulin. In contrast, in response to a glucose tolerance test, both the free insulin and C-peptide responses were normal as was glucose tolerance. The clinical difference in responsiveness to the human and beef insulins is well illustrated in vitro by insulin binding differences and is probably a function not only of different equilibrium constants of the antibody for human and beef insulins but also of different high affinity binding site concentrations.

Insulin administration often causes the development of insulin antibodies. Plasma insulin-binding capacity was assessed and free and total plasma insulin were measured in 96 insulin-dependent diabetic patients. Patients with the highest insulin binding capacity not only had the highest total insulin but also had the lowest free insulin. It is concluded that bound insulin is not unavailable to cells, that sufficient dissociation can occur for this insulin source to make a sizeable contribution to insulin delivery to the cell receptors, and that this can happen at a very low free insulin concentration.