Faculty Bibliography

The ontogenic switch from fetal to adult hemoglobin could be due to gradual replacement of fetal by adult erythroid progenitor cells, each with unique potential hemoglobin expression programs. We have studied globin synthesis in single colonies derived from the BFU-E of human newborn infants and rhesus monkey fetuses, both at the time of the switch. In the 140-day rhesus fetal colonies, beta synthesis ranged from 8% to 44% of (beta + gamma) in 47 colonies. When fetal colonies were arbitrarily defined as those with below 26% beta, and adult colonies those with above 26% beta, we predicted that there would be 12 adult colonies; 14 were observed. The rhesus fetal results were compatible with the presence of two or more classes of progenitors within the population seen at the time of the switch. In the human newborn study, the number of colonies increased fourfold from day 14 to day 21 in culture. beta synthesis also increased. These increases occurred even in cultures from which red, well-hemoglobinized colonies were completely removed as they appeared. New cohorts of late-appearing colonies were derived from immature progenitors. The early colonies showed correlation of G gamma and beta synthesis, but the late colonies did not show this correlation. We had found the latter pattern previously in adult, fetal-like colonies. Thus fetal progenitors in man have high levels of Hb F and G gamma, with correlation of these parameters at the level of single-progenitor-derived colonies; these fetal progenitors are found in fetuses and newborns. Adult, fetal-like progenitors have lower levels of Hb F and G gamma, without any correlation between the levels of each. The adult progenitors appear in newborn infants as the switch occurs, and they are the only class found in adults. Our observations in the simian and human studies support a clonal model for hemoglobin switching.

Fifty febrile severely granulocytopenic patients were given four daily transfusions of 2.2 X 10(10) normal donor granulocytes. Twenty-three (46 percent) responded clinically, although both responders and nonresponders were similar in clinical characteristics at the outset. This study examines the relation between serum opsonic activity before initiation of granulocyte administration and clinical response. Opsonic activity to three test organisms (Escherichia coli 286 and ON 2, and Staphylococcus aureus) and to 15 blood stream isolates from 14 patients was measured as serum-dependent uptake of heat-killed 14C-labeled bacteria by normal donor leukopheresis granulocytes in an in vitro assay and compared with results obtained with a standard normal serum in each assay. At a concentration of 8 percent serum, all patient groups were equivalent to standard (90 to 102 percent) for the three test organisms. When rate-limiting concentrations of serum (1 to 2 percent) were employed, opsonic activity remained similar to standard for S. aureus in all patient groups and for the two E. coli strains in responders (82 to 98 percent). In contrast, opsonins for E. coli decreased to 41 to 50 percent of standard in nonresponders (p less than 0.01). When patients with proved infection were separately analyzed, opsonin activity for E. coli 286 and ON 2 was significantly greater in responders than nonresponders (73.6 versus 34.9 percent and 124.8 versus 58.1 percent, respectively for the two strains) (p less than 0.01). Patients with opsonin activity of 50 percent or greater of standard had a greater response rate (73 versus 19 percent and 45 versus 0 percent for the two E. coli strains) (p less than 0.005 and p = 0.08, respectively). Eight of 10 patients with 75 percent or greater of standard for opsonic activity against their own blood stream isolates also responded, whereas zero of four with opsonins less than 75 percent of standard had a favorable outcome. These results indicate that serum opsonic activity may be a determinant of clinical response to granulocyte transfusion in infected granulocytopenic patients and may be predictive of outcome. We conclude that opsonic activity should be assessed in such patients before granulocyte administration and suggest a trial of plasma infusion in opsonin-deficient patients

Previously untreated patients who had anemia (hemoglobin level, less than or equal to 10 g/dL) caused by myelofibrosis (MF) (16 patients) or other myeloproliferative disorders (13 patients) were given the opportunity to participate in a prospective randomized study-to be treated either with 30 mg/day of oral fluoxymesterone and necessary transfusions or by transfusions alone. Of the 24 patients whose data could be evaluated, four (29%) of 14 responded well to fluoxymesterone therapy (hemoglobin level rise, of greater than 2 to greater than 10 g/dL and relief of symptoms of anemia), whereas, in the transfusion arm, there were no good 'responders'; one of ten patients was a partial 'remitter' (responder), with a rise in the hemoglobin level of 1 to 2 g/dL. All responders to fluoxymesterone therapy showed a 50% or more maximum uptake of injected ferrous citrate Fe 59 into RBC hemoglobin, whereas no nonresponder met this criterion. All responders had MF (marrow more than one third replaced by collagen). There was no significant difference in survival of patients in the two arms of the study

Previous studies have reached disparate conclusions on the issue of whether or not granulomas confer any protection against recurrence of Crohn's disease after operation. In an attempt to resolve this controversy, we have examined postoperative recurrence rates in 102 patients with Crohn's disease. There were no differences in cumulative postoperative recurrence-free survival between the 53 patients with granulomas present in the resected specimens and the 49 patients without granulomas. It was concluded that the presence of granulomas in resected specimens of Crohn's disease exerts no independent influence on the rate of postoperative recurrence

Thirty-one patients with essential thrombocythemia were randomized to receive either melphalan or radioactive phosphorus as myelosuppressive therapy. Twenty-seven patients were evaluable for response. Of 13 patients treated with melphalan, 11 had a complete response (platelet count less than 450,000/mm3) at 3 and 6 months. This response rate was significantly better than the response to radioactive phosphorus. The response rates were similar at 12 months. No significant toxicity was observed with either regimen.

A total of 109 patients with advanced lung cancer, all cell types, were randomized between MACC chemotherapy only, consisting of methotrexate, doxorubicin (Adriamycin), cyclophosphamide, and lomustine (CCNU); MACC plus levamisole (LMS) orally; and MACC plus Corynebacterium parvum (CP) sc. Of these patients, 101 were evaluable, with no differences among the three treatment groups for overall response rate and survival time. Objective response rates and median survival times were 41% and 230 days for patients given MACC only, 39% and 257 days for those given MACC plus LMS, and 44% and 223 days for those given MACC plus CP, respectively. There was a significant increase in survival for patients with large cell anaplastic carcinoma receiving CP or LMS, particularly in the good-performance-status category. Pretreatment delayed cutaneous hypersensitivity to recall antigens in 50 patients had prognostic significance, but repeat tests after 2 months of treatment in 30 patients did not show different patterns of conversion among the three groups. There was no difference in hematologic toxicity among the three groups. With the possible exception of large cell anaplastic carcinoma, immunotherapy with LMS or CP as given in this trial does not appear to be therapeutically advantageous in advanced lung cancer

Two earlier studies of prognosis of coronary heart disease among men enrolled in the Health Insurance Plan in the 1960s and 1970s permitted us to examine whether prognosis had improved over this ten-year period. The new comparison involved 1,133 men aged 35 to 64 years who had survived a first acute myocardial infarction and were followed up for mortality after a baseline examination. Mortality estimates were controlled for clinical and demographic differences between the two cohorts by multivariate methods and by comparing subgroups. The analyses showed no difference in long-term prognosis between patients in the two decades. The observations in this population suggest that any contribution of improved medical care to the nationally observed secular decline in mortality from coronary heart disease in the time period studied was probably restricted to the acute stage of myocardial infarction

Thirty-four patients with polycythemia vera complicated by pruritus were treated with 900 mg of cimetidine daily for 30 days and their responses to treatment were evaluated. The conditions of 15 (44%) were improved, with 12 patients stating that pruritus completely disappeared. Nineteen patients either showed no improvement or had increasing symptoms. No toxic effects were reported. The positive responses seen are encouraging and suggest that controlled studies are indicated to further evaluate the effectiveness of H2 antagonists

The records of 239 infants born to 228 women dependent on narcotic drugs were reviewed to determine if type of drug abused and adequacy of prenatal care would affect pregnancy and fetal outcome. Seventy-nine (33%) pregnancies occurred in women in supervised methadone maintenance, 78 (32%) in women on unsupervised methadone maintenance, 49 (21%) in women on street heroin, and 33 (14%) in women who were multiple drug users. Although the presence of withdrawal symptoms did not differ with respect to type of drug abused, the outcome was significantly better in those infants born to women on supervised methadone maintenance as compared to all other groups (p less than 0.001). There was no demonstrable relationship between the number of prenatal visits to the clinic and fetal outcome. A relationship could not be demonstrated between the maintenance dose during pregnancy and the presence of withdrawal symptoms in the infants born to women on supervised methadone maintenance. The findings of the study suggest that supervised methadone maintenance is compatible with an uneventful pregnancy and delivery. Neonatal complications, with the exception of withdrawal, do not appear to differ from that seen among infants born to nondrug dependent women