Reversible thioridazine-induced magnetic resonance imaging-documented pituitary enlargement associated with hyperprolactinemia.

OBJECTIVE: To document a case of reversible thioridazine-induced pituitary enlargement associated with hyperprolactinemia in a patient with schizophrenia. METHODS: We describe a 19-year-old woman with paranoid schizophrenia who was taking thioridazine (a phenothiazine), in whom hyperprolactinemia, galactorrhea, oligomenorrhea, and a reversible anatomic pituitary abnormality developed. Serial magnetic resonance imaging (MRI) studies were used to assess the status of the pituitary gland during various pharmaceutical therapies. RESULTS: Laboratory evaluation revealed increased serum prolactin (PRL) levels that ranged from 76 to 135 mg/L. Results of thyroid function tests were normal, and gonadotropins and estradiol levels were low, consistent with hyperprolactinemia. MRI revealed asymmetric enlargement of the right side of the pituitary gland. Discontinuing the thioridazine therapy resulted in normalization of the serum PRL and resolution of the pituitary abnormality. Subsequent worsening of the patient's psychiatric condition necessitated a course of electroconvulsive therapy and initiation of treatment with clozapine, a D4 dopamine receptor antagonist. At 1-year follow-up, at which time the patient was maintained on clozapine and was not taking thioridazine, both serum PRL levels and MRI findings remained normal. CONCLUSION: Our patient was shown to have asymmetric pituitary enlargement associated with thioridazine-induced hyperprolactinemia, which reversed when use of the drug was discontinued. In patients with serum PRL levels in excess of 100 mg/L during antipsychotic drug therapy, evaluation for a prolactinoma is warranted.

Mapping diabetic sensory neuropathy by current perception threshold testing.

Detailed clinical neurological examinations were conducted on 44 nondiabetic volunteers and 59 diabetic subjects. The examinations focused particularly on sensory symptomatic and physical evaluation. Standardized assessment of symptoms and physical testing of light touch, pain, vibratory, and thermal sensation was performed at the hand, wrist, elbow, foot, ankle, and knee. A total symptom score and physical score were defined by summing test scores at each site. Current perception threshold (CPT) testing that used constant sine-wave-alternating current was conducted at the same anatomic sites. CPT correlations with the physical score gave r values of .55 for 5 Hz, .60 for 250 Hz, and .62 for 2000 Hz (n = 618). Correlations with the symptom score were not as strong: r = .45 for 5 Hz, .46 for 250 Hz, and .51 for 2000 Hz. The correlation with symptom score was due primarily to a strong relationship for the symptom of numbness (r = .53 for all 3 frequencies). Correlations with pain and paresthesia were much lower. CPTs for diabetic subjects at the three frequencies were higher at most locations than for the nondiabetic volunteers. However, CPTs were no different from normal values in diabetic subjects without evidence of neuropathy. CPT testing appears to be a useful technique for assessment of diabetic sensory neuropathy.

Microvascular blood flow, volume, and velocity measured by laser Doppler techniques in IDDM.

A laser Doppler device with the capability to simultaneously measure skin blood flow, microvascular volume, and erythrocyte velocity was used to assess blood flow changes in 35 insulin-dependent diabetes mellitus (IDDM) subjects, mean age 33 +/- 1 yr, with average duration of diabetes 14 +/- 1 yr, and in a nondiabetic control group. Blood flow was determined at 35 and 44 degree C at several sites on the upper and lower extremities with a temperature-regulated probe. Blood flow was highest at both temperatures on the pulps of the index finger and the first toe, regions of high density of arteriovenous anastomoses. There was significantly greater blood flow at most locations for the nondiabetic than the diabetic group at 35 degree C, and the differences between the two groups were substantially larger at 44 degree C. At 44 degree C, blood flow in the control group was approximately 40% greater in the upper extremity and 50% greater in the lower extremity than it was in the diabetic subjects. The differences were attributed to decreases of both microvascular volume and velocity in the diabetic group. In the upper extremity, volumes in the diabetic patients were 10-15% lower and velocities 10-40% lower than in the nondiabetic subjects. In the lower extremity, volumes were 20-25% lower and velocities 40-50% lower. We conclude that laser Doppler techniques can be used to assess microvascular changes in the skin of diabetic patients. This approach may be useful to evaluate and model diabetic microangiopathy.

Insulin receptors in acute infection: a study of factors conferring insulin resistance.

Acute infections are accompanied by tissue insulin resistance, as manifested by worsening of metabolic control in diabetic patients and decreased glucose tolerance in non-diabetic subjects. To clarify the potential role of altered insulin receptor status in this phenomenon, we studied [125I]insulin binding to monocytes in 7 otherwise healthy subjects during acute bacterial and viral infections of moderate severity. The values were compared to those obtained after convalescence (five patients) and those of 24 normal subjects. Insulin binding during infection, at a time when insulin resistance was demonstrable, was indistinguishable from convalescent or normal values. Plasma glucose and insulin levels, the insulin to glucose ratio, as well as plasma GH, cortisol, and FFA were significantly elevated during infection, while plasma glucagon, epinephrine, and norepinephrine levels were normal. We conclude that insofar as monocyte receptors are representative of other tissues, insulin resistance in infection is mediated at the postreceptor level.

Enhanced efficacy of U-500 insulin in the treatment of insulin resistance caused by target tissue insensitivity.

This report describes a 55-year-old diabetic patient with severe insulin resistance due to obesity and hepatic cirrhosis. Anti-insulin antibodies were responsible for only a minor part of the insulin resistance. Insulin resistance was mediated primarily at the target tissue level by a combination of insulin receptor deficiency and a postreceptor defect. When the patient was treated with U-500 regular pork insulin subcutaneously, her insulin requirement was only one third to one half of that during U-100 NPH or regular insulin treatment. The reasons for the greater efficacy of U-500 insulin are unknown. U-500 insulin may have a place in the optimization of therapy for patients with insulin resistance of nonimmunologic origin.

Circadian variation of basal plasma growth hormone in man.

Human GH (hGH) circulating during periods of basal pituitary secretion is frequently undetectable in human plasma by RIA. Such basal periods predominate throughout most of the day. Knowledge of the concentration of basal hGH is important, since hGH plays a role in the homeostasis of several metabolic fuels and ions. We have used an immunoadsorbent technique to extract and concentrate hGH from large plasma samples before RIA in order to precisely measure basal hGH levels [immunoextracted RIA (IERIA)]. Twenty milliliters of blood were drawn from five normal volunteers every 2 h over a 24-h period. The obtained plasma (10-13 ml) was passed over an anti-hGH antibody Sepharose column. The column-bound hGH was eluted and measured by RIA. Recovery of hGH in the plasma extracts was measured using [131I]hGH as an internal standard. Recovery averaged 54 +/- 12% (mean +/- SD). Theoretically, the extraction procedure expanded the useful range of the RIA about 50-fold. In practice, the detection limit was 40 pg/ml plasma. Basal hGH (defined as less than 1 ng/ml by conventional RIA) was found to range from less than 40 to 746 pg/ml by IERIA. Above 2 ng/ml, conventional RIA and IERIA yielded comparable results (r = 0.92; P less than 0.001), but below 2 ng/ml, the two assays did not correlate (r = 0.01; P greater than 0.5). In general, conventional RIA overestimated plasma hGH in that range. As a result, the ratio between nocturnal and daytime hGH secretion, as assessed by IERIA, is higher than previously appreciated. We conclude that basal plasma hGH is highly variable below 1 ng/ml, and may on occasion be as low as less than 40 pg/ml. The episodic nature of hGH release appears to persist in the basal state. The conventional definition for basal hGH (less than 5 ng/ml) should probably be revised downward to less than 0.8 ng/ml.

The orientation of insulin receptors in the red cell membrane.

High-affinity binding of insulin to receptors in human erythrocyte membranes occurred at the external surface, but not at the cytoplasmic surface of the plasma membrane, as assessed by insulin binding to right-side-out and inside-out membrane vesicles. Even after prolonged (3 h) incubation at 22 degrees C, binding at the cytoplasmic membrane aspect remained negligible. The data indicate that the insulin receptor displays its hormone-binding site exclusively toward the extracellular space and that transmembrane mobility ("flip-flo") of the receptor from one to the other membrane leaflet is severely restricted.

Insulin resistance and insulin receptors in hepatic cirrhosis.

To explore the influence of extrahepatic factors in the pathogenesis of insulin resistance in hepatic cirrhosis, we studied 125I-insulin binding to erythrocytes and monocytes of 14 clinically stable cirrhotic individuals and compared the results with a normal control group. All patients had fasting normoglycemia at the time of the study but abnormal glucose tolerance was detected in 7 of 9 cirrhotic patients after an oral glucose load. Seven patients (group N) had normal fasting serum insulin levels, and 7 patients (group H) manifested fasting hyperinsulinemia. However, all patients had elevated insulin levels after oral glucose. Insulin binding to erythrocytes was significantly decreased in both cirrhotic subgroups; monocyte studies in 5 hyperinsulinemic patients revealed a similar decrease in binding. Scatchard analysis in monocytes suggests that this decreased binding is secondary to a decrease in the receptor number per cell. No correlation between insulin binding and fasting plasma insulin, glucagon, or growth hormone levels was seen. Sera from 4 patients were examined for the presence of a non-specific inhibitor of insulin binding, but no evidence for such a factor was found. We conclude that the decrease in insulin binding is mediated in the monocyte by a reduction of receptor concentration; in the erythrocyte the mechanism for decreased binding could not be clearly delineated. The insulin resistance seen in cirrhosis may result in part from decreased binding of insulin to target tissues; an additional postreceptor defect cannot be excluded in hyperinsulinemic individuals.

Insulin receptors and insulin resistance in human pregnancy: evidence for a postreceptor defect in insulin action.

Pregnancy is accompanied, in its later stages, by physiological resistance to the action of insulin. We studied the potential contribution of altered insulin receptors to this phenomenon in 12 healthy pregnant women during their third trimester. For comparison, we studied the same women again several weeks postpartum. We also used a group of randomly chosen nonpregnant subjects as a control population. Women in this control group were studied during the luteal phase of the menstrual cycle. Plasma insulin and insulin to glucose ratios were significantly higher in the pregnant women. Insulin binding to peripheral blood monocytes was higher in pregnancy than postpartum in the majority of women, and as a group, the pregnant subjects showed significantly higher insulin binding than the nongravid subjects. This appeared to be due to a greater number of receptor sites per cell. We found no correlation between plasma insulin and insulin binding in pregnancy. We conclude that the insulin resistance of pregnancy is not attended by impaired binding of insulin to cellular receptors, at least in the monocyte. Our data suggest that the defect in insulin action lies at a site distal to the receptor.

Sex differences in the size of bile acid pools.

In view of the excess prevalence of gallstones among women and the association of gallstones with diminished bile acid pool size, we measured bile acid pools in 27 male and 25 female healthy human volunteers. The average bile acid pool in the women was significantly smaller than in the men (2.25 +/- .12 g versus 2.88 +/- .16 g; p = 0.003). Chenodeoxycholic acid pool size, computed from bile acid composition data available in 43 of these subjects, was also smaller in women than men (0.94 +/- 0.06 versus 1.22 +/- 0.07 g; p = 0.004). Age, race, and body size bore no statistically significant relationship to bile acid pool size. Biliary cholesterol saturation was positively correlated with weight and obesity and showed a significant inverse correlation with chenodeoxycholic acid pool size, but not with total bile acid pool size. These findings suggest a possible mechanism for the higher prevalence of gallstones among women.