Charles E. Antle

Aromatase activity in primary and metastatic human breast cancer

Aromatase activity was measured in 104 primary and 24 metastatic breast cancer patients. The assay employed quantitates the production of 3H water release from 1β‐[3H] androstenedione after aromatization. Of 104 human primary breast tumors studied, 64 contained measurable aromatase activity, ranging from 5–70.5 pmol estrone formed/g protein/hour. In primary breast cancers there was no difference in levels of aromatase activity when analyzed by menstrual status or age by decade. Aromatase activity was similar in small and large primary tumors. The median aromatase activity of primary breast tumors (8.6 pmol/g/h) was similar to that found in metastatic breast cancer deposits (12.0 pmol/g/h). Aromatase activity did not correlate with either estrogen (ER) or progesterone (PR) receptor concentration in the tissues assayed. In this regard there were 33 ER− PR− tumor biopsies. Twelve of these 33 tumors contained aromatase activity greater than 10 pmol/g/hour. Cancer 59:779‐782, 1987.

Discrimination Between the Log-Normal and the Weibull Distributions

The log-normal and the Weibull are often considered for situations in which a skewed distribution for a non-negative random variable is needed. The ratio of maximized likelihoods provides a good test for selecting one of these. A table of the necesqary critical values is given. The table may also be used for discriminnting between the normal and the type 1 extreme value distributions.

Maximum Likelihood Estimation with the Weibull Model

Abstract Some of the questions concerning the uniqueness of the maximum likelihood estimates for the parameters in the Weibull distribution are considered for both censored and noncensored samples. In some cases answers previously known are reviewed, while new results are presented for some other cases. In particular, it is shown that with the shape parameter known, the maximum likelihood estimates of the location and scale parameters exist and are unique. Also, the case in which all three parameters are unknown is examined in detail.

Likelihood Ratio Test for DiscriminaGon Between Two Models with Unknown Location and Scale Parameters

The problem of selecting a model from two models with unknown location and scale parameters is considered. It is noted that the distribution of the ratio of maximum likelihoods does not depend upon the values of the nuisance location and scale parameters. Consequently, this ratio provides a convenient (and hopefully powerful) test for discriminating between two location and scale parameter models when the parameters are unknown. Tables for discriminating between the normal and the Cauchy; the normal and the exponential; or the normal and the double exponential are presented. An empirical comparison of the power of this test with the power of the UMP invariant test for discriminating between the normal and the Cauchy is given. Some further comparisons with the chi-square goodness of fit and the Kolmogorov-Smirnov tests for normality are also given.

Statistical Inference From Censored Weihull Samples

In life testing experiments it is a fairly common practice to terminate the experiment before all items have failed. The Weibull distribution is often used as a model for the observations and when a computer is available maximum likelihood estimation of the parameters is to be recommended. The tables presented in this paper enable one to set confidence limits on the parameters and the reliability based on the maximum likelihood estimates for selected censoring and sample sizes. It is also observed that, as in the case with no censoring, the maximum likelihood estimator of the reliability is very nearly unbiased and its variance is near the Cramer-Rao lower bound, Unbiasing factors for the maximum likelihood estimator of the shape parameter are given.

Estimation of Parameters in the Weibull Distribution

A new method of estimation is used to obtain two simple estimators of the parameters in a Weibull distribution. These estimators are similar to the estimators given by Gumbel, Miller and Freund, and Menon. Some useful properties of these estimators are developed which make it practical to use Monte Carlo methods to determine the variances and biases of the estimators for various sample sizes. Comparisons of the estimators can be made and unbiasing factors calculated in some cases. The variances of the estimators are also compared to the Cramer-Rao lower bounds for regular unbiased estimators.

Corynebacterium parvum versus bacille Calmette-Guérin adjuvant immunotherapy of stage II malignant melanoma.

Two separate studies have been reported comparing Corynebacterium parvum and bacille Calmette-Guérin (BCG) as adjuvant immunotherapy for stage II melanoma patients (The Milton S. Hershey Medical Center, 48 patients; Southeastern Cancer Study Group [SECSG], 162 patients). As the criteria for patient selection and drugs used were similar, we have pooled the data to analyze the effects of these two treatments. Both studies used BCG (Tice, Chicago, IL) 3 x 10(8) live organisms per treatment by Tine technique and C parvum (Burroughs-Wellcome, Triangle Park, NC) subcutaneous at a dose of 4 mg/m2 (SECSG) or 5 micrograms/m2 (Hershey) per treatment. The only difference in these studies was the frequency of immunization, with patients in Hershey receiving 22 doses and the SECSG patients receiving 55 doses during the 2-year period of treatment. Kaplan-Meier life-table analysis for the 210 patients shows a prolonged disease-free interval for patients treated with C parvum (P = .02, two-sided Mantel procedure). In similar fashion, patients treated with C parvum had an improved survival rate (from all causes) when compared with BCG-treated patients (P = .012). An analysis of the results for the 170 patients for which the number of positive nodes was available was performed using Coxs model, with nodes as a stratification variable and with covariates of place, treatment, age, and sex. In this analysis, an observed benefit for C parvum on the disease-free interval had a P value of .37 while the benefit of C parvum on the survival times (from all causes) had a P value of .04. When the same analysis was performed using only patients aged younger than 60 years, the observed benefit of C parvum on disease-free interval had a P value of .08 and the benefit of C parvum on survival times (from all causes) had a P value of .008.

A Property of Maximum Likelihood Estimators of Location and Scale Parameters

It is shown that the maximum likelihood estimators (m.l.e.s)

Reliability Estimation for the Generalized Gamma Distribution and Robustness of the Weibull Model

\hat \alpha

Upper Bounds for the Power of Invariant Tests for the Exponential Distribution with Weibull Alternatives

and