Search

Summary Geographically dispersed patients, inconsistent treatment tracking, and limited infrastructure slow research for many orphan diseases. We assess the feasibility of a patient-powered study design to overcome these challenges for Castleman disease, a rare hematologic disorder. Here, we report initial results from the ACCELERATE natural history registry. ACCELERATE includes a traditional physician-reported arm and a patient-powered arm, which enables patients to directly contribute medical data and biospecimens. This study design enables successful enrollment, with the 5-year minimum enrollment goal being met in 2 years. A median of 683 clinical, laboratory, and imaging data elements are captured per patient in the patient-powered arm compared with 37 in the physician-reported arm. These data reveal subgrouping characteristics, identify off-label treatments, support treatment guidelines, and are used in 17 clinical and translational studies. This feasibility study demonstrates that the direct-to-patient design is effective for collecting natural history data and biospecimens, tracking therapies, and providing critical research infrastructure., Graphical Abstract, Highlights Partnership with the patient community supports recruitment and results dissemination A patient-powered design enables high enrollment from a rare disease population Extensive clinical data reveal >40 off-label treatments used in Castleman disease De-identified linkage with a biobank supports translational research discoveries, Pierson et al. describe the feasibility of a patient-powered natural history registry for studying Castleman disease. They pair a traditional registry with a patient-powered approach, in which patients self-enroll and data collection is centralized. Clinical insights support treatment guidelines, and de-identified linkage to a biobank enables translational discoveries.

Background: Given the persistent disparity in the advancement of women compared with men faculty in academic medicine, it is critical to develop effective interventions to enhance women's careers. We carried out a cluster-randomized, multifaceted intervention to improve the success of women assistant professors at a research-intensive medical school. Materials and Methods: Twenty-seven departments/divisions were randomly assigned to intervention or control groups. The three-tiered intervention included components that were aimed at (1) the professional development of women assistant professors, (2) changes at the department/division level through faculty-led task forces, and (3) engagement of institutional leaders. Generalized linear models were used to test associations between assignment and outcomes, adjusting for correlations induced by the clustered design. Results: Academic productivity and work self-efficacy improved significantly over the 3-year trial in both intervention and control groups, but the improvements did not differ between the groups. Average hours worked per week declined significantly more for faculty in the intervention group as compared with the control group (−3.82 vs. −1.39 hours, respectively, p = 0.006). The PhD faculty in the intervention group published significantly more than PhD controls; however, no differences were observed between MDs in the intervention group and MDs in the control group. Conclusions: Significant improvements in academic productivity and work self-efficacy occurred in both intervention and control groups, potentially due to school-wide intervention effects. A greater decline in work hours in the intervention group despite similar increases in academic productivity may reflect learning to “work smarter” or reveal efficiencies brought about as a result of the multifaceted intervention. The intervention appeared to benefit the academic productivity of faculty with PhDs, but not MDs, suggesting that interventions should be more intense or tailored to specific faculty groups.

A sensitive and specific double antibody radio- immunoassay for the major apolipoprotein (apoB) of rhesus (Macaca muIatta) serum very low density lipoprotein (VLDL) and low density lipoprotein (LDL) is described. The anti- serum was raised to LDL (d 1.030-1.040 g/ml) and the LDL, (d 1.020-1.050 g/ml) was labeled with '=I by the chloramine-T or iodine monochloride method. The assay, which was sensitive to 0.02-0.5 pg of LDL2, had an inter- assay coefficient of variation of 4.5%. This assay was suc- cessfully used to measure apoB in the whole serum and low density lipoproteins of control monkeys maintained on a standard Purina monkey chow (PMC) diet and of three groups of monkeys fed atherogenic diets: an "average American diet," a 25% peanut oil and 2% cholesterol-sup- plemented PMC diet, and a 25% coconut oil and 2% choles- terol-supplemented PMC diet. The control monkeys (n = 13) had a serum cholesterol of 146 f 28 mg/dl and an apoB of 50 f 18 mg/dl. In the monkeys maintained on the atherogenic diets the serum apoB was elevated: 103 +- 28 mg/dl (American), 102 f 35 mg/dl (peanut oil), and 312 f 88 mg/dl (coconut oil). The values for serum total choles- terol were 333 f 65 mg/dl (American), 606 f 212 mg/dl (peanut oil), and 864 f 233 mg/dl (coconut oil) and were elevated relative to controls (P < 0.001). For each of the diets, total serum cholesterol correlated with serum apoB (E' < 0.001). The slopes of the regression lines of serum apoB vs. cholesterol for the monkeys on the PMC, Ameri- can, and coconut oil diets were similar (m = 0.531, 0.401, and 0.359, respectively), but differed from that of monkeys on the peanut oil diet (m = 0.12 1). The immunoreactivities of rhesus and human LDL were compared using specific antisera raised against these antigens. In homologous assay systems, monkey and human LDL exhibited unique im- munological determinants. The same results were obtained with the delipidated preparations of the two LDLs using antisera raised against either monkey or human apoB. Crossover studies using a heterologous tracer with each anti- serum resulted in the selection of a specific population of antibodies directed against antigenic sites shared by these two LDL species.

The development of a radioimmunoassay for apolipoprotein A-II (apo A-II) is described. Initial studies revealed a lack of immunological identity between purified apo A-II used as the standard and serum or HDL. Extensive testing of different buffers, standards, antisera, tracers, utilization of a detergent, and heating of sera failed to resolve the problem. Gel filtration of iodinated and non-iodinated apo A-II on Sephadex G-100 columns showed that apo A-II, in dilute solution, elutes in a higher molecular zone than expected with a broad, assymetrical profile. The use of a subfraction of the tracer in the assay resulted in parallelism in the serum and standard dilution curves. The apo A-II assay was sensitive, specific, and reproducible. Apo A-II added to sera was fully recovered and delipidation did not affect the immunoreactivity of either serum or HDL. Apo A-II contributed approximately 20% to the protein mass of HDL. Comparison of these results with those obtained by radial immunodiffusion, and with previously reported data, indicates that the reactivity of apo A-II in its native and delipidated forms may be markedly influenced by different immunologic methodologies and their specific reagents. Caution should thus be shown at present in assigning absolute concentrations to apo A-II in serum or HDL.

A sensitive and reproducible method has been developed for separation of the major serum lipoproteins from 1 ml or less of human serum by isopycnic density gradient ultracentrifugation. The serum, applied to a step gradient (total volume 12.8 ml), was spun for 48 hr at 38,000 rpm at 10°C and, in each of the fractions, apoli- poproteins B and A-I were quantified by the respective radioimmunoassays. The markers for lipid distribution used were (4-I4C )cholesterol and (U-'4C)lecithin, each in- cubated with an aliquot of serum at 20°C for 75 min prior to ultracentrifugation. In control sera, three main frac- tions, very low density (VLDL), low density (LDL), and high density (HDL) lipoproteins were clearly separated from a bottom fraction. Their flotational, electrophoretic, and chemical properties were in good agreement with those reported for the corresponding lipoproteins separated by conventional ultracentrifugation. Both apo B and apo A-I were fully recovered. Essentially all of the apo B was found in VLDL (9.3 2 3.5%) and LDL (87 & 4.6%); of the apo A-I, 81.0 rt 5.7% was in HDL and the remainder (17.0 +- 5.8%) was in the bottom fraction. The peak activities of (14C)cholesterol coincided with the peak of apo B in both LDL and VLDL, and with the peak of apo A-I in HDL. The results with the radiolabeled cholesterol were in good agreement with those obtained by chemical analyses. Carbon 14-labeled lecithin, although fully recovered, was not an accurate marker of phospholipid distribution be- cause, under our experimental conditions, a significant amount of the lecithin was converted into its lyso deriva- tive. The mechanism of the conversion was not established; it appeared to be unrelated to the activities of either lecithin-cholesterol acyl transferase or a Caz+-dependent phospholipase. Besides its validity in the study of control sera, our method also proved successful in the separation of the serum lipoproteins of the few patients with dyslipo- proteinemia (abetalipoproteinemia and familial hyper- cholesterolemia) who were examined. However, the ap- plicability of the method to all dyslipoproteinemias was not assessed. Taken together, the results indicate that the single-spin method could be useful in clinical studies as a complement to other established techniques.

Isolated livers from rhesus monkeys (Macaca mulatta) were perfused in order to asses the nature of newly synthesized hepatic lipoprotein. Perfusate containing [3H]leucine was recirculated for 1.5 hr, followed by an additional 2.5-hr perfusion with fresh perfusate. Equilibrium density gradient ultracentrifugation clearly separated VLDL from LDL. The apoprotein composition of VLDL secreted by the liver was similar to that of serum VLDL. The perfusate LDL contained some poorly radiolabeled, apoB-rich material, which appeared to be contaminating serum LDL. There was also some material of an LDL-like density, which was rich in radiolabeled apoE. Rate zonal density gradient ultracentrifugation fractionated HDL. All perfusate HDL fractions had a decreased cholesteryl ester/unesterified cholesterol ratio, compared to serum HDL. Serum HDL distributed in one symmetric peak near the middle of the gradient, with coincident peaks of apoA-I and apoA-II. The least dense fractions of the perfusate gradient were rich in radiolabeled apoE. The middle of the perfusate gradient contained particles rich in radiolabeled apoA-I and apoA-II. The peak of apoA-I was offset from the apoA-II peak towards the denser end of the gradient. The dense end of the HDL gradient contained lipoprotein-free apoA-I, apoE, and small amounts of apoA-II, probably resulting from the relative instability of nascent lipoprotein compared to serum lipoprotein. Perfusate HDL apoA-I isoforms were more basic than serum apoA-I isoforms. Preliminary experiments, using noncentrifugal methods, suggest that some hepatic apoA-I is secreted in a lipoprotein-free form. In conclusion, the isolated rhesus monkey liver produces VLDL similar to serum VLDL, but produces LDL and HDL which differ in several important aspects from serum LDL and HDL.

The pathogenesis of hyperglucagonemia and of the alterations in the pattern of circulating immunoreactive glucagon (IRG) associated with renal insufficiency was studied in rats in which a comparable degree of uremia was induced by three different methods, i.e., bilateral nephrectomy, bilateral ureteral ligation, and urine autoinfusion. Nephrectomized and ureteral-ligated rats were markedly hyperglucagonemic (575 +/- 95 pg/ml and 492 +/- 54 pg/ml, respectively), while IRG levels of urine autoinfused animals (208 +/- 35 pg/ml) were similar to those of control rats (180 +/- 26 pg/ml), indicating that uremia per se does not account for the hyperglucagonemia observed in renal failure. Similarly, plasma IRG composition in this group of animals was indistinguishable from that of controls, in which 88.2 +/- 5.9% of total IRG consisted of the 3,500-mol wt fraction. The same component was almost entirely responsible (82.6 +/- 4.1%) for the hyperglucagonemia observed in ligated rats, while it accounted for only 57.6 +/- 5.0% of the circulating IRG in nephrectomized animals. In the latter group, 36.8 +/- 6.6% of total IRG had a mol wt of approximately 9,000, consistent with a glucagon precursor. This peak was present in samples obtained as early as 2 h after renal ablation and its concentration continued to increase with time reaching maximal levels at 24 h. These results confirm that the kidney is a major site of glucagon metabolism and provide evidence that the renal handling of the various circulating IRG components may involve different mechanisms. Thus, the metabolism of the 3,500-mol wt fraction is dependent upon glomerular filtration, while the uptake of the 9,000-mol wt material can proceed in its absence, as long as renal tissue remains adequately perfused. This finding suggests that the 9,000-mol wt component may be handled by peritubular uptake.

Two analogs of porcine insulin with substitutions of leucine for phenylalanine in the COOH-terminal region of the insulin B chain have been prepared by a combination of solid-phase synthesis and semisynthesis. Solid-phase synthesis of the substituted octapeptides B23-B30 bearing the trifluoracetyl group on lysine-B29, enzymatic coupling of the octapeptides to bis(tertiary-butyloxycarbonyl)desoctapeptide insulin by trypsin, and deprotection of the corresponding adducts in formic acid and piperidine resulted in two insulin derivatives, one with leucine at position B24 and the other with leucine at position B25. These analogs had only about 10% and 1%, respectively, of the activity of porcine insulin in competing for the binding of [125I]iodoinsulin to both rat adipocytes and human IM-9 lymphocytes. The relative potencies of the analogs in stimulating glucose oxidation by rat adipocytes decreased in the order porcine insulin > [LeuB24]insulin > [LeuB25]insulin. However, at high concentrations both analogs had full agonists activity. Experiments in which the semisynthetic insulins were mixed with the native hormone showed that [LeuB24]insulin, but not [LeuB25]insulin, was an active antagonist of insulin action. These results suggest that the antagonistic activity of a human insulin variant having leucine at position B24 or B25 can be assigned to the molecule with the sequence Gly-Leu-Phe-Tyr (residues B23-B26) in its active site.

Both alleles of the insulin gene of a patient with mild diabetes [maturity-onset-diabetes-of-the-young (MODY)-type syndrome] associated with hyperinsulinemia have been cloned, and the sequences have been determined. One allele contained a mutation (single nucleotide transition) in the coding sequence for the B chain at position 24 (TTC leads to TCC), resulting in the loss of a restriction enzyme (Mbo II) cleavage site in the gene. This mutation results in the substitution of serine for phenylalanine in a critically important region of the insulin molecule that is intimately involved in receptor binding. Both insulin alleles were of the alpha type and, aside from a single nucleotide deletion in the 5' region of the normal allele, their sequences were identical to those previously determined.

Calcium gluconate (10 mg Ca(++)/kg) was administered intravenously over a 2-hour period to 16 adult patients who were evaluated for hypoglycemia. In nine of ten patients with benign or malignant insulinomas (eight proven at operation, and two with positive chemical tests and angiographic localization awaiting operation), significant hypoglycemia and hyperinsulinemia occurred within 60 to 90 minutes after the start of the calcium infusion. Serum proinsulin and Cpeptide concentrations increased at the time of the calciuminduced hyperinsulinemia in several patients in whom these parameters were studied. The one individual who did not respond to the calcium infusion was found to have a benign insulinoma. His basal glucose/insulin ratio of 0.64 was the lowest of the insulinoma group and thus his failure to respond to calcium may indicate that his tumor was secreting maximally at the time of the infusion. Following successful removal of the insulinoma, calcium infusion did not result in changes in serum glucose or insulin concentrations (tested in five patients). In contrast, neither a patient with pathologically documented islet cell hyperplasia, five others with reactive, fupctional or drug-induced hypoglycemia, nor four healthy volunteers showed any changes in circulating glucose or insulin levels while receiving calcium intravenously. Calcium infusion is a safe, rapid and effective provocative test for the diagnosis of insulin-secreting, islet cell tumors of the pancreas.

In order to evaluate potential differences in the kinetics of peptide hormone-receptor interactions in hepatocytes of different species, we developed a simple procedure for the isolation of canine hepatocytes. Cells (obtained by collagenase perfusion of an extirpated dog liver lobe) were isolated with uniform high viability and yield. In addition, isolated dog hepatocytes tolerated incubation for at least 4 hr in defined medium with only a slight decrease in viability and with no change in the kinetics of [125I]iodoinsulin or [125I]iodoglucagon binding to cell-surface receptors. Comparisons of peptide hormone interactions with isolated dog and rat hepatocytes showed that (i) [125I]iodoglucagon associated with specific membrane receptors more rapidly than did [125I]iodoinsulin, for both rat and dog hepatocytes and at both 30 degrees C and 37 degrees C; (ii) the steady-state binding of [125I]iodoglucagon at 30 degrees C was greater than that of [125I]iodoinsulin in dog hepatocytes, but the reverse relationship held in rat hepatocytes; (iii) the rate of dissociation of [125I]iodoinsulin from hepatocytes of both species was enhanced by the presence of the unlabeled hormone, whereas the rate of dissociation of receptor-bound [125I]iodoglucagon was enhanced by the presence of unlabeled glucagon only in hepatocytes derived from the dog; and (iv) [125I]iodopancreatic polypeptide bound to neither rat nor dog hepatocytes, although the [125I]iodotyrosylated peptide bound to rat hepatocytes with an unusually high apparent dissociation constant. While confirming essential findings of pancreatic hormone binding to isolated hepatocytes, this comparison suggests that both qualitative and quantitative aspects of hormone-target cell interactions can show interspecies variability.

Gel filtration of human serum extracts on Bio-Gel P-30 columns produced two peaks of material reactive with insulin antisera. The earlier eluting fraction appeared at the elution position of proinsulin (serum proinsulin-like component, PLC) while the second fraction corresponded in elution volume to insulin. In assays using porcine insulin-131I and an antiserum against porcine insulin, human pancreatic proinsulin was less reactive than human insulin. Serial dilutions of the serum PLC in the immunoassay showed immunological identity with the human proinsulin standard. Partial tryptic digestion of the serum PLC yielded products with increased immunological reactivity as estimated with insulin as the standard. With larger amounts of trypsin, all the serum PLC was converted to insulin-like components (desthreonine and desoctapeptide insulin). On the basis of these results we conclude that the earlier eluting fraction of human serum extracts is proinsulin. The fasting values of proinsulin in normal subjects ranged between 0.05 and 0.4 ng/ml, representing from 5 to 48% of the insulin concentration. In one subject the values of proinsulin were higher than those of insulin. After oral administration of 100 g of glucose, the proinsulin levels tended to rise similarly to insulin. Three obese patients with hyperinsulinemia had higher fasting levels of proinsulin and a greater increase after glucose than the normal subjects. As the high levels of proinsulin coexisted with raised insulin concentration in these obese subjects, the relative proportions of the two hormones were in the same range observed in the normal group. Thus hyperinsulinemia in these obese subjects was not accompanied by an increase in the fraction of serum proinsulin. When the values for serum proinsulin were expressed as percentage of the insulin levels, there was a decrease in the per cent proinsulin in the first hour of the glucose tolerance test. After the second hour, the per cent tended to rise towards the fasting levels.