Judith Bromberg

Declining prevalence of cervicovaginal human papillomavirus infection with age is independent of other risk factors.

Background and Objectives: Human papillomavirus (HPV) infection of the female genital tract is the most common sexually transmitted disease. Although the prevalence of HPV in women without detectable cervical disease has been shown to decline with increasing age, the relationship to sexual behavior has not been investigated. Goal: To identify risk factors for, and associated with the age‐dependent decline in, genital HPV infection in women. Study Design: The prevalence of HPV was determined in a cohort of 439 sexually active inner‐city women between the ages of 18 and 50 years recruited in Brooklyn, New York. Cervicovaginal cells were collected by lavage, and HPV was detected by low‐stringent Southern blot hybridization. Results: The prevalence of HPV infection ranged from 36% in women younger than 25 years of age to 2.8% in women 45 years or older. Logistic regression analysis identified an increased risk for cervical HPV infection to be independently associated with number of sex partners in the past year (odds ratio [OR], 1.04 per yearly increase in age; 95% confidence interval [CI], 1.00 to 1.08), younger age (OR, 0.92 per year increase in age; 95% CI, 0.88 to 0.95), and not living with partner (OR, 2.28; 95% CI, 1.40 to 4.22). Conclusions: The lower prevalence of HPV infection in older women compared to younger women was found to be independent of sexual behavior. These results suggest that a biologic effect, such as HPV immunity acquired over time and with multiple exposures, may mediate the inverse relationship between age and HPV prevalence.

Chemotherapy and radiation therapy before transhiatal esophagectomy for esophageal carcinoma.

Recent efforts to improve survival in patients with esophageal carcinoma have combined both systemic and local therapy. From October 1985 to October 1987, 43 patients with local-regional esophageal cancer (adenocarcinoma in 21, squamous cell in 22) were treated with cisplatin, vinblastine, and 5-fluorouracil chemotherapy concurrent with 4,500 cGy radiation therapy for 21 days before transhiatal esophagectomy 3 weeks later. Two patients died of chemotherapy/radiation therapy toxicity. Forty-one completed preoperative chemotherapy/radiation therapy. At operation, 2 patients had incurable metastatic disease; 39 underwent transhiatal esophagectomy. Eleven patients had no residual tumor in the resected specimen for a 27% (11 of 41) pathological complete response rate. Preoperative chemotherapy/radiation therapy resulted in no increased perioperative morbidity as compared with our historical controls. One patient died postoperatively of an unrecognized brain metastasis (2% operative morbidity). At a median follow-up of 27 months, 20 patients (47%) are alive and clinically disease-free and 21 have died, 19 from progression of their carcinoma. The median survival time for all 43 patients is 29 months (Kaplan-Meier estimate), and cumulative survival is 72% at 12 months, 60% at 24 months, and 46% at 36 months. All 11 patients with a complete response are alive at a median follow-up of 36 months, and all are disease-free. The 2-year survival of 60% of this group as compared with 32% in our earlier patients treated with transhiatal esophagectomy alone suggests that intensive combined modality therapy improves survival in these patients. A randomized prospective trial is now in progress.

Concurrent preoperative chemotherapy and radiation therapy in localized esophageal adenocarcinoma

Twenty‐four patients with localized, potentially resectable adenocarcinoma of the esophagus were enrolled in this study to evaluate the use of preoperative chemotherapy and radiation therapy, followed by transhiatal esophagectomy. The patients were newly diagnosed and had received no prior treatment. Radiation therapy consisted of 4900 cGy, administered as 350‐cGy fractions 5 days a week for 14 fractions. The chemotherapy consisted of 5‐fluorouracil 300 mg/m2/day administered as a continuous 24‐hour intravenous infusion for 96 hours each week, concomitantly with the radiation therapy. After a 3‐week rest, patients underwent transhiatal esophagectomy. Twenty‐two patients could be observed for their responses to the chemotherapy and radiation regimen. Radiographically, 41% showed improvement, 36% had stable disease, and 23% had progression. Nineteen patients underwent surgery; all patients had total gross removal of disease, and two patients had a complete histologic response.

Procarbazine chemotherapy in the treatment of recurrent malignant astrocytomas after radiation and nitrosourea failure

The Brain Tumor Study Group has shown procarbazine (PCB) to be as effective an adjuvant treatment as 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU). We treated 35 patients with recurrent malignant astrocytomas after radiation and nitrosourea failure with successive courses of PCB 150 mg/m2/d for 28 days every 8 weeks. After 2 courses, 2 patients had complete responses, 7 had partial responses, 11 had stable disease, and 15 had progression. Significantly more patients receiving PCB had complete or partial responses or stable disease than a similar group of patients in a previous trial who received intraarterial (IA) cisplatin (DDP). There is a significant advantage in time to disease progression for those receiving PCB compared with those receiving IA diaziquone (AZQ). Our results suggest that PCB is a more effective 2nd agent than IA DDP or AZQ following radiation and nitrosourea failure.

Comparison between BCNU and procarbazine chemotherapy for treatment of gliomas

SummaryWe compared sequential single-agent BCNU and procarbazine (PCB) chemotherapy in 31 patients with gliomas [grade IV (10), grade III (15), grade II (6)]. Patients had failed surgical biopsy ± resection and radiation therapy. All patients were treated initially with BCNU 150-300mg/m2 by intra-arterial or intravenous route every 6 weeks. After CT evidence of tumor progression, all patients received PCB 150mg/m2/day for 28 days every 8 weeks. Patient responses to BCNU were CR (0), PR (7), SD (12), progression (12), and to PCB CR (2), PR (9), SD (6), and progression (14). Kaplan-Meier estimates of median time to failure for all patients were shorter for BCNU, 5.0 months (range 1.5–20), than for PCB, 6.0 months (range 2–50+). There was a statistically significant difference (Mantel-Cox test, p=0.02) in the distribution of time to disease progression between the two drugs, especially for grade III tumors (p= 0.02). The cumulative proportion of patients without disease progression at 6 months was 26% while on BCNU, compared to 48% while on PCB; at 12 months the cumulative proportions were 3% for BCNU compared to 35% for PCB. Although there was no formal washout period between administration of the two drugs, no carryover effect was evident. These data provide further evidence that PCB has significant activity against malignant glioma and may, in fact, be more effective than BCNU.

Health insurance coverage of the unemployed.

Of 1,332 unemployed individuals in the Detroit area interviewed in late 1983, 51% did not have health insurance. Lack of insurance was directly related to length of unemployment. Of those unemployed 3 months or less, 31% had no insurance, as compared with 56% of those unemployed more than 3 years. For the most part, these were not the chronically uninsured: 78% of them were insured when they were employed. Three fourths of those without insurance were not covered by Medicaid either. These findings suggest that during the latest economic recession, the problem of health insurance loss due to losing ones job was more severe than had been assumed by most policymakers.

Radiosensitization with carotid intra‐arterial bromodeoxyuridine ± 5‐fluorouracil biomodulation for malignant gliomas

Bromodeoxyuridine (BUdR), a nonhypoxic radiosensitizing drug, is a halogenated pyrimidine analog that is incorporated into the DNA of dividing cells in a competitive process with thymidine. BUdR sensitizes cells to radiation therapy. 5-Fluorouracil (5-FU) inhibits the endogenous synthesis of thymidine, resulting in increased incorporation of the BUdR. Neurons and glial cells have a very low mitotic rate; they will not incorporate BUdR and will not be sensitized. BUdR and 5-FU are best delivered intra-arterially (IA) because of their regional advantage. We infused BUdR ±5-FU over 8½ weeks, before and during 59.4-Gy focal conformal external beam radiation therapy, through a permanently implanted pump with a catheter placed retrograde through the external carotid artery to the carotid bifurcation. Sixty-two patients with grades III or IV glioma were entered into one of two trials, with 23 patients receiving BUdR alone and 39 patients receiving BUdR + 5-FU. The maximum tolerated dose (MTD) of BUdR alone was 400 mg/m2/d for 834 weeks. The Kaplan-Meier median survival (KMS) was 20 months. In the BUdR + 5-FU trial, the MTD of BUdR was also 400 mg/m2/d and 5-FU was 5 mg/m2/d with a KMS of 17 months. The KMS of all 62 patients in both trials 1 and 2 was 18 months. Pathologic grading used both the original World Health Organization (WHO) and 1993 modified WHO systems. The KMS of grade IV patients was 13.8 months (48 patients) with the original system and 17 months (58 patients) with the modified system. The 2-year survival rate was 21% with the original and 28% with the modified grading system. The dose-limiting toxicity was a reversible unilateral focal forehead dermatitis, blepharitis, and conjunctivitis. Continuous IA halopyrimidine infusion may enhance the effectiveness of radiation in the treatment of malignant glioma.

An introduction to logistic regression with an application to the analysis of language recovery following a stroke

The aim of a statistical model is to present a simplified representation of the underlying structure in a data set by separating systematic features from random variation. Sometimes the purpose of a statistical model is to provide a simple descriptive summary of the data and sometimes it is to use the data for comparative or inferential purposes. In practice, the specification of a statistical model requires a thorough understanding of the substantive area of application, an assessment of the validity of the assumptions of the model, and an evaluation of the fit of the model to the data. In this paper, as an illustration of these aspects of the statistical modeling of data, we consider the specification, application, and interpretation of a logistic regression model for the investigation of relationships between binary response data and a collection of explanatory variables. We illustrate applications of the methodology using data from a prospective study of spontaneous language recovery following a stroke (Holland, Greenhouse, Fromm, & Swindell, 1989).

R-wave amplitude changes during exercise in adolescents with left ventricular pressure and volume overload

Abstract To determine the diagnostic value of exerciseinduced R-wave changes in adolescents with congenital heart disease, the responses of 50 adolescents without significant heart disease were compared with those of 72 patients with either a left ventricular (LV) pressure or volume overload lesion. Among the pressure overload group, 24 patients had valvular aortic stenosis (AS) and 27 had coarctation of the aorta. The volume overload group included 12 patients with mitral regurgitation (MR) and 9 with aortic regurgitation (AR). Severity of the cardiac lesion was assessed using cardiac catheterization in patients with AS, physical examination in patients with coarctation of the aorta and clinical or angiographiec criteria, or both, in patients with valvular regurgitation. The R wave was measured in 10 consecutive QRS complexes in leads II, aVF and V 5 at rest, maximal exercise and 1-minute recovery. At maximal exercise, control subjects had a mean decrease in amplitude (ΔR) of −3.6 mm (p

A normed measure of variability among proportions

A measure of variability among a set of proportions is developed. There are no distributional assumptions, so the measure, H, is applicable in a wide variety of situations. H is scaled so that its range is zero (all proportions are equal) to one (maximum variability among the proportions, given the weighted average). The measure can be interpreted as a distance: the value of H indicates the position of the proportions relative to the possible extremes of no variability or maximum variability. Any set of constant weights can be applied to the proportions: the weights are used to compute the weighted average proportion and are also used to determine the extent to which each proportion affects the variability measure. Comparisons are made to other measures of variability and a numerical example is given.