Randomized trial of breast self-examination in Shanghai: methodology and preliminary results.

BACKGROUND: The efficacy of breast self-examination in helping to reduce mortality from breast cancer has not been rigorously demonstrated. PURPOSE: To assess efficacy, a large, randomized trial was initiated in Shanghai, China. METHODS: From October 1989 to October 1991, 267040 current and retired female employees associated with 520 factories in the Shanghai Textile Industry Bureau were randomly assigned on the basis of factory to either a self-examination instruction group (133375 women) or a control group (133665 women). The women were born within the period from 1925 through 1958. Women in the instruction group were given intensive training in breast self-examination, including the use of silicone breast models and personalized instruction, plus two subsequent reinforcement sessions and multiple reminders to practice the technique. Women in the control group were asked to attend training sessions on the prevention of low back pain. All women have been followed for the development of breast diseases and for death from breast cancer. RESULTS: A high level of participation during the first 4-5 years of the trial was documented among women in the instruction group. Randomly sampled women in this group demonstrated greater proficiency in detecting lumps in breast models than did randomly sampled women in the control group. Approximately equal numbers of breast cancers were detected in the two groups (331 in the instruction group and 322 in the control group) through 1994, which is the last year for which case-finding efforts have been completed. The breast cancers detected in the instruction group were not diagnosed at an appreciably earlier stage or smaller size than those in the control group. More benign breast lesions were detected in the instruction group than in the control group (1457 versus 623, respectively), suggesting a higher index of suspicion for women who received training. Cumulative breast cancer mortality rates through 5 years from entry into the study were nearly equivalent for the two groups. CONCLUSIONS: Breast self-examination has not led to a reduction in mortality from breast cancer in this study cohort in the first several years since the trial began. A shift toward the diagnosis of disease at a less advanced stage in women given instruction has also not been demonstrated. Longer follow-up of participants in this trial is required before final assessment can be made of the efficacy of breast self-examination. IMPLICATIONS: At this time, there is insufficient evidence to recommend for or against the teaching of breast self-examination.

Heritability of quasi-continuous skeletal traits in a randombred population of house mice.

Heritabilities of 11 quasi-continuous skeletal traits were estimated in randombred house mice of three separate ages (1, 3, and 5 months). Three separate methods--regression, maximum likelihood correlation, and Falconer's Method--were used to obtain heritabilities for each of the separate age groups. Significant differences in the incidences of seven of the skeletal traits were found among ages, but they did not affect the heritability estimates, these estimates being pooled over ages. Heritabilities calcuated from female parents were consistently higher (by about 13%) than those from male parents, indicating the presence of maternal effects. Mid-parent estimates made by all three methods gave very similar mean levels (0.17--0.20). Although low, this level compared favorably with that expected on the basis of previously estimated rates of accumulation of genetic variance. Maternal effects estimated from full sib correlations averaged 0.08.

Risk of recurrence after ductal carcinoma in situ of the breast.

A cohort study was conducted to estimate the risk of breast cancer recurrence among women diagnosed with ductal carcinoma in situ (DCIS) and to identify tumor or patient characteristics that influence that risk. A population-based cancer registry was used to identify a cohort of 709 female residents of western Washington who were diagnosed with DCIS between January 1980 and June 1992 and were treated with breast-conserving surgery. Information about breast cancer recurrences, treatment, and several patient characteristics and exposures was obtained from postal questionnaires. Recurrences were confirmed using information from the cancer registry or hospital pathology reports. Approximately 15% of women experienced a recurrence within the first 5 years after diagnosis [95% confidence interval (CI), 12-18%]; 31% had a recurrence within 10 years (95% CI, 24-38%). There was a suggestion that risk was slightly elevated for women with larger tumors (> or =1.5 cm) and tumors of comedo subtype. Relative risks (RRs) were elevated for women who were premenopausal at diagnosis of DCIS (RR = 2.3; 95% CI, 1.1-5.0). Women in the upper decile of body mass index were at twice the risk of a recurrence as those women in the lower four deciles (RR = 2.3; 95% CI, 1.1-4.8). There was also a suggestion that women who used menopausal hormones for at least 2 years after their diagnosis of DCIS were at increased risk of recurrence compared to nonusers of menopausal hormones (RR = 1.8; 95% CI, 0.7-5.0). Our results suggest that the risk of recurrence may be related to some tumor characteristics as well as the hormonal milieu of the patient at or after her diagnosis of DCIS. However, larger studies are needed to more clearly document predictors of disease recurrence after DCIS.

Assessment of variance components models on pedigrees using cholesterol, low-density, and high-density lipoprotein measurements.

Plasma total cholesterol, low-density lipoprotein (LDL), and high-density lipoprotein (HDL) measurements on 402 individuals in 62 randomly selected families from the Columbia Medical Plan population were used to select the "best" model among a series of multifactorial models using the maximum likelihood method described by Lange et al [1976]. These models included both genetic and nongenetic components of variance. The most parsimonious model for each trait was selected and examined using a goodness-of-fit statistic designed by Hopper and Mathews [1982] to test the assumptions of this technique. A simple additive genetic model was the most plausible for all three measurements, suggesting a strong role for genetic factors in determining lipid and lipoprotein levels in these data. Goodness-of-fit statistics for these models were examined and showed little evidence of deviation from the assumption of multivariate normality within pedigrees. This approach of selecting the most parsimonious model among a series of competing models and then assessing its goodness-of-fit has many applications in studying familial aggregation of quantitative traits.

Monitoring cyclosporin concentrations in marrow transplant recipients: comparison of two assay methods.

Renal dysfunction is the major dose-limiting toxicity associated with cyclosporin therapy. We have previously shown in patients undergoing allogeneic bone marrow transplantation that serum cyclosporin concentrations, as measured by radioimmunoassay (RIA), correlate significantly with the development of renal dysfunction. However, since the RIA measures both parent drug and metabolites, the relative role of each in the development of nephrotoxicity could not be determined. Therefore, we re-measured cyclosporin concentrations in the same serum samples by high-performance liquid chromatography (h.p.l.c.). Serum cyclosporin concentrations of less than 50, 50-100 and greater than 100 ng/ml, as measured by h.p.l.c., were considered equivalent to cyclosporin concentrations of less than 150, 150-250 and greater than 250 ng/ml, as measured by RIA. Contrary to results by RIA, cyclosporin concentrations measured by h.p.l.c. did not significantly correlate with renal dysfunction, which suggests that measurement of serum cyclosporin concentrations by h.p.l.c. provides no clinical advantage to RIA for monitoring cyclosporin concentrations to prevent renal dysfunction.

Minimal risk of chronic renal dysfunction in marrow transplant recipients treated with cyclosporine for 6 months.

To determine whether 6 months of cyclosporine therapy is associated with chronic renal dysfunction, we evaluated serum creatinine concentrations 1 year post-transplant in 82 marrow transplant recipients randomized to receive either cyclosporine (n = 40) or methotrexate (n = 42) as graft-versus-host disease (GVHD) prophylaxis. Nine patients in the methotrexate group were later given cyclosporine as treatment for acute or chronic GVHD (methotrexate----cyclosporine). Cyclosporine prophylaxis was started on the day before marrow infusion, given at full doses until day 50, then gradually tapered and discontinued by day 180. Methotrexate prophylaxis was started on day 1 and given intermittently until day 102. Patients in the cyclosporine and methotrexate----cyclosporine groups had significantly higher mean serum creatinine values during the first 100 days post-transplant than methotrexate-treated patients, but by 1 year mean serum creatinine values were not significantly different between the three groups. Serum creatinine values at 1 year were also not significantly different from baseline values in each of the groups. None of the patients who had their cyclosporine discontinued by day 180 developed chronic renal dysfunction, defined as a doubling of the baseline serum creatinine at 1 year. We conclude that chronic renal dysfunction occurs rarely in marrow transplant recipients treated with 6 months of cyclosporine and when it does occur, it appears to have minimal clinical significance.

In situ and invasive cervical carcinoma in a cohort of infertile women.

OBJECTIVE: To assess the risk of cervical neoplasia associated with the use of ovulation-inducing agents such as clomiphene citrate (CC) DESIGN: Case-cohort study. SETTING: Infertility clinics in Seattle, Washington. PATIENTS: A cohort of 3,837 women evaluated for infertility at some time during 1974-1985. MAIN OUTCOME MEASURE: Computer linkage with a population-based tumor registry was used to identify women diagnosed with cervical cancer before January 1, 1992. Data regarding infertility testing and treatment were abstracted from medical records for women who developed cancer and a randomly selected subcohort. RESULTS: Thirty-six women in the cohort developed in situ or invasive cervical cancer in comparison with an expected number of 67.8 cases (standardized incidence ratio = 0.5, 95% confidence interval [CI] 0.4 to 0.7). Infertile women with fallopian tube abnormalities were at an increased risk of cervical cancer relative to women whose infertility was believed to be due to other causes. The risk among women who had taken CC was reduced relative to infertile women who had not used this drug (relative risk = 0.4, 95% CI 0.2 to 0.8). This association was present both in women with and without tubal abnormalities. However, the size of the reduction in risk was not influenced by duration of use. CONCLUSIONS: The hypothesis that use of antiestrogenic agents, such as CC, can lead to a reduced risk of cervical neoplasia warrants testing in other studies.

Risk of cutaneous melanoma in a cohort of infertile women.

We assessed the risk of cutaneous malignant melanoma associated with the presence of ovulatory abnormalities and with the use of ovulation-inducing agents (such as clomiphene citrate) in a cohort of 3,837 women evaluated at infertility clinics in Seattle, WA, between 1974 and 1985. Computer linkage with a population-based tumour registry was used to identify women diagnosed with melanoma before 1992. Data regarding infertility testing and treatment were abstracted from the infertility clinic medical records for women who developed cancer and a randomly selected subcohort. Twelve women in the cohort developed cutaneous malignant melanoma, in comparison with an expected number of 6.8 cases (standardized incidence ratio = 1.8; 95% confidence interval (CI) 0.9-3.1). Within the cohort, risk was increased among women who had used clomiphene during 12 or more menstrual cycles (relative risk = 2.2; 95% CI 0.5-10.2). All four of the women with this duration of clomiphene use who developed melanoma had ovulatory abnormalities, and three had also used human chorionic gonadotropin (HCG). No elevation in risk associated with the presence of ovulatory abnormalities was observed in the absence of at least 12 cycles of clomiphene exposure; also, there was no increased risk associated with long-term use of clomiphene among women without ovulatory abnormalities, but the number of such women was very small. Thus, it is not certain to what extent the observed increased risk of melanoma in this cohort (if not due to chance) may be attributable to the use of clomiphene or HCG, or is a reflection of some underlying hormonal abnormality for which the drug was administered.

Design and analysis of cancer screening trials.

This article reviews approaches to the design and analysis of cancer screening trials. After summarizing some basic screening concepts and potential pitfalls, we introduce several possible screening trial designs with examples from the literature. We review in detail methods for analyzing screening trial data, including testing for a significant difference in disease-specific mortality between the control and intervention groups, estimating the mortality differential if one exists, and evaluating the programme lead time, the screen sensitivity and the role of stage shifting. We consider Overall mortality analyses, which are based on the experience of the trial population, and Limited mortality analyses, which are based on the experience of comparable groups of cases in the control and intervention groups. We discuss methods for selecting candidate comparable case groups and confirming that they are in fact comparable. We conclude by showing how the principles discussed have been used in the planning and design of a current screening trial for multiple cancers.

Early HIV infection in vivo: branching-process model for studying timing of immune responses and drug therapy.

We propose a stochastic, branching-process model of early events in vivo in human or simian immunodeficiency virus (HIV or SIV) infection and study the influence that the time of appearance of virus-specific antibodies or cytotoxic cells, or of administration of antiretroviral drugs, has on the probability of progression to a chronic infection. In some biological scenarios, our model predicts that a few days' delay in response or intervention would make little difference, while in others it would be highly deleterious. We show that prophylactic efficacy does not require perfect efficiency at neutralizing infectious virus. Data from a trial of PMPA, a potent antiretroviral drug, as post-exposure therapy for SIV infection in macaques, reported by C.-C. Tsai, P. Emau, K.E. Follis, T.W. Beck, R. E. Beneveniste, N. Bischofberger, J.D. Lifson, W.R. Morton (J. Virol. 72 (1998) 4265), provides a test of the model. We show that their observations are consistent with a branching-process without invoking supplementary viral- or host-variability. Finally, most animal trials of antiviral drugs or vaccines use very high viral inoculums; our model demonstrates that in such experiments we risk greatly underestimating the efficacy of these agents.

Joint inferences on vaccine efficacy against infection and disease with application to the first HIV vaccine efficacy trial.

In many clinical trials, subjects are followed for two stages of outcomes, and it is of interest to compare the incidence of each outcome between two randomized groups. The outcome of the first stage may influence the outcome of the second stage. Moreover, the relative risks of the two outcomes may be linked, with the time-dependent profile of relative risk for the second outcome functionally dependent on that of the first. For example, during exposure to HIV, virologic and host factors simultaneously impact the probability of infection and the subsequent viral trajectories, and the efficacy of a tested vaccine to prevent infection and to prevent viral failure may work in concert. We address this problem by modeling the relationship between the stage two hazard function and covariates via Cox's proportional hazards model (Cox, 1972), with the stage one log-hazard ratio theta(*) at the first event time Tl, included as a covariate. With theta(*) estimated using three methods, 1) nonparametric kernel smoothing; 2) locally parametric penalized splines; and 3) fully parametric cubic linear splines, we subsequently develop inference procedures for the regression parameter in the stage two Cox model based on each of the estimator of theta(*). The inferential procedures are studied in simulations and are illustrated with application to data from the world's first preventive HIV vaccine efficacy trial.

An adaptive weighted log-rank test with application to cancer prevention and screening trials.

A class of adaptive weighted log-rank statistics is described where the vector of weights is chosen in a data-dependent way from a family of "smooth" weight vectors. A parametric family of weight vectors is identified which includes most shapes of weighting vectors that will be near optimal in many cancer prevention and screening trials. This family of weight vectors is used in an application of the proposed method to data from a breast cancer screening trial. Results from a small simulation study comparing the power of the adaptive statistic to that of the unweighted log-rank statistic are presented.

Aspects of the use of relative risk models in the design and analysis of cohort studies and prevention trials.

Relative risk regression methods provide a unifying and powerful approach to a range of problems in the design and analysis of cohort studies and prevention trials. Standard partial likelihood-based estimation procedures do not, however, encompass several features that are important in such contexts. Specifically, one may wish to relate disease rates marginally to 'recent' risk factor measurements, whereas a partial likelihood approach requires one to condition on an accumulating risk factor history. Secondly, risk factor values may be ascertained with considerable measurement error, thereby requiring specialized procedures to estimate relative risk parameters. Thirdly, analysis of raw materials to obtain desired covariate (risk factor) histories may involve considerable expense if carried out for the entire cohort. Case-control and case-cohort sampling procedures can avoid much of this expense, but once again partial likelihood estimation procedures require generalization. Such generalizations are described herein.

The effect of additional follow-up and accrual on survival time data.

A method called partial completion is proposed for predicting the gain in precision of the Kaplan-Meier survival curve associated with additional follow-up and accrual. This is accomplished by using the initial data to predict the numbers of patients who would be at risk at the observed death times by the end of the proposed second follow-up period. A consistency result ensures that the predictors will be accurate in large samples while simulation results suggest that the predictors are accurate with moderate sample sizes. The procedures are applied to a bone marrow transplant study and the Channing House data set.

On the catch-up time method for analyzing cancer screening trials.

In randomized cancer screening trials, the ratio of the mortality rate for the screened group to that for the control group is typically not constant as a function of years from randomization. This is due to an initial lag effect, but also to a dilution effect that results from the accrual of comparable cases in both groups after the end of the screening period. In order to combat the potential loss of power when applying conventional analysis tools, specifically the logrank test, Aron and Prorok (International Journal of Epidemiology 15, 36-43), have advocated analyzing the mortality experience using only the subcohort of cases ascertained within a given time period. However, it is not clear how to select an appropriate case ascertainment point, since this will depend on aspects of the natural history of the disease process which are poorly identified. Aron and Prorok suggest choosing the case ascertainment point to be the point at which the cumulative number of cases in the control group first becomes equal to that in the intervention group, that is, the "catch-up time." In this paper, we undertake a thorough evaluation of the bias and power properties of the catch-up time method. We base our study on simulated data resembling the Health Insurance Plan of Greater New York study cohort. We consider several models for postdiagnosis survival under the null hypothesis of no screening effect on mortality, and under the alternative hypothesis of an effect of screening. We show that the catch-up method can yield tests with sizeable bias. In the absence of detailed knowledge about the underlying disease process, we suggest some adaptive tests that maintain nominal size but have more attractive power properties than the standard logrank test.

A likelihood ratio test for cancer screening trials.

In randomized cancer screening trials, mortality rates for the screened group relative to those of the control group are not likely to be constant as a function of years from randomization due to the inherent lag between initiation of screening and any putative effects of screening on mortality. In this situation, a log rank test for differences in mortality between the randomization groups will not be optimal. Although optimality could potentially be recovered by use of a weighted log rank statistic, the optimal weights are difficult to specify a priori and the potential loss of power by use of poorly specified weights is great. We describe a likelihood ratio test with two degrees of freedom for use in this situation which is based on a fit of a weakly structured full model. Computation of an approximate significance level for this test is described and a large sample justification for this approximation is given. Size and power properties of the proposed statistic are compared to that of several other statistics in a small simulation study and the statistic is applied to data from the HIP Breast Cancer Screening Trial.

Linear rank tests for interval-censored data with application to PCB levels in adipose tissue of transformer repair workers.

Linear rank statistics are described for testing for differences between groups when the data are interval-censored. The statistics are closely related to those described by Prentice (1978, Biometrika 65, 167-179) for right-censored data. Problems in calculating the statistics are discussed and several approaches to computation including estimation of the efficient rank scores are described. Results from a small simulation study are presented. The methods are applied to data from a study relating tissue levels of PCBs to occupational exposure.

The Cox proportional hazards model with change point: an epidemiologic application.

In this paper, we develop the Cox proportional hazards model with special structured time-dependent covariates in the context of prospective epidemiologic studies. Our model possesses the following two features: (i) different relative risk parameters are allowed for early versus late onset of the disease of interest; (ii) an additional parameter is introduced so that specification is not required for the time (age) at which a change of the magnitude of the relative risks takes place, the so-called change point. Some difficulties with statistical inference for the proposed model are briefly discussed, and the large-sample distribution of a test for no change point is derived. As an illustration, we apply the model to a set of data gathered on a group of white male medical students of The Johns Hopkins Medical School enrolled between 1948 and 1964. We examine the hypothesis that the effect of reactivity to the cold pressor test may vary with early versus late onset of hypertension.

Measurement of junctional acetylcholine receptors in myasthenia gravis: clinical correlates.

We measured the number of acetylcholine receptors (AChRs) at neuromuscular junctions (NMJs) in motor point biopsies from 76 patients with myasthenia gravis (MG) and 49 control subjects, using an alpha-bungarotoxin binding technique. The mean number of AChRs/NMJ was significantly lower (P less than 0.001) in muscles of MG patients (deltoid 0.7 +/- 0.1 X 10(7) than in controls (2.1 X 0.2 X 10(7). None of the control muscles had fewer than 0.9 X 10(7) AChRs/NMJ, and 75% of MG muscles were below that level. Conversely, 88% of controls had 1.5 X 10(7) or more AChRs/NMJ, whereas only one of the MG muscles (2.5%) was at or above that level. The reduction of AChRs correlated approximately with the clinical severity of weakness (P less than 0.02). Patients with focal weakness had reduced numbers of AChRs/NMJ in clinically strong muscles. The only other condition in which junctional AChRs were reduced was polymyositis. These findings in a large sample of patients confirm the reduction of junctional AChRs in MG, the relationship of available junctional AChRs to clinical manifestations, and the systemic nature of the disorder even in cases with focal weakness. Measurement of AChRs can be useful diagnostically in situations where other tests are inconclusive.