The practice of causal inference in cancer epidemiology.

Causal inference is an important link between the practice of cancer epidemiology and effective cancer prevention. Although many papers and epidemiology textbooks have vigorously debated theoretical issues in causal inference, almost no attention has been paid to the issue of how causal inference is practiced. In this paper, we review two series of review papers published between 1985 and 1994 to find answers to the following questions: which studies and prior review papers were cited, which causal criteria were used, and what causal conclusions and public health recommendations ensued. Fourteen published reviews on alcohol and breast cancer and 6 published reviews on vasectomy and prostate cancer were examined. For both series of reviews, nearly all available published studies were cited except for ecological studies and prior reviews. Sources of causal criteria were often not provided. When they appeared, all citations were either the 1964 Surgeon General's report or works of Austin Bradford Hill. Reviews often excluded and sometimes altered criteria without giving reasons for these changes. The criteria of consistency and strength of association were almost always used accompanied by dose-response and biological plausibility in a majority of reviews. The criterion of temporality, considered by many methodologists to be a necessary causal condition, was infrequently used. Confounding and bias were often added considerations. Public health recommendations were not discussed in nearly one-half of the reviews.

Androgen receptors in biopsy specimens of prostate adenocarcinoma. Heterogeneity of distribution and relation to prognostic significance of receptor measurements for survival of advanced cancer patients.

Needle biopsy specimens of primary adenocarcinoma and surgical specimens of carcinomatous nodal tissue were obtained from previously untreated clinical D stage prostatic adenocarcinoma patients. Assessment of the relation between specimen androgen receptor site content and survival using either scatterplots or Kaplan-Meier analyses showed specimen receptor content was a poor prognostic P greater than 0.1, of survival subsequent to orchiectomy or diethylstilbestrol (DES) therapy. The possibility that heterogeneity of specimen androgen receptor site content contributed to this finding was evaluated by comparing receptor content of multiple small or large tissue specimens from the same prostate gland of patients with benign prostatic hyperplasia or nonmetastatic prostatic cancer. This evaluation showed significant microheterogeneity of human prostate androgen receptor site content which was substantially masked in large tissue specimens. We conclude that microheterogeneity of human prostate androgen receptor site content compromises the use of biopsy specimen androgen receptor measurements as a prognostic of patient survival subsequent to initiation of hormonal therapy.

Immunocytochemical assay of estrogen receptors in endometrial carcinoma with monoclonal antibodies. Comparison with biochemical assay.

Specimens of endometrial adenocarcinoma, surgically obtained from 18 women, were analyzed for distribution of estrogen receptors by an immunocytochemical assay, employing monoclonal anti-estrophilin antibodies and the peroxidase-antiperoxidase technique. Results were compared with biochemical receptor analyses, and were in concordance in 83% of them. Marked tumor cell and tissue receptor heterogeneity were apparent with the immunocytochemical method, and a variety of patterns of nuclear staining in positive tissue samples were revealed. These results indicate that the immunohistologic method will provide a number of entirely new variables that may eventually be correlated with both clinical and pathologic features of this malignancy, and may prove to be of value in the prediction of clinical endocrine response.

Biochemical and morphological characterization of clonal AXC rat prostate cancer cells.

We used three heterogeneous parental cultures of LSC-AXC rat prostate cancer cells: LSC-AXC-C/O, cells maintained on culture medium; LSC-AXC-D/O, cells maintained on culture medium containing 10(-7) M 5 alpha-dihydrotestosterone; and LSC-AXC-T/O, cells maintained on culture medium containing 10(-7) M testosterone, to isolate clonally derived cell lines. Eleven of 15 clonal cell lines were tumorigenic when inoculated into intact male AXC rats. Eight tumorigenic clonal cell lines were selected for further evaluation, and all were found to possess features characteristic of secretory epithelium, as judged by light and electron microscopy. All parental cell lines and the eight selected clonal cell lines contained cytoplasmic and nuclear androgen receptors. Total receptor content was 131 +/- 61 (S.D.), 43 +/- 32, and 274 +/- 96 fmol/100 micrograms of DNA, respectively, for C-, D-, and T-cells. The differences were significant (p less than 0.05). Androgen receptor content of young mature or senescent AXC rat ventral prostate, respectively, is 518 +/- 58 and 266 +/- 40 fmol/100 micrograms of DNA. Since chromosomal analysis established that LSC-AXC prostate cancer cells are hypotriploid, androgen receptor content per cell in C- and T-cells is indicated to be either greater than or equal to that of senescent AXC rat ventral prostate, the tissue in which the original adenocarcinoma arose. Parental and clonal cell lines contained 5 alpha-reductase activity. There were significant differences (p less than 0.05) in both total reductase activity and metabolite distribution. Consequently, the intracellular content of testosterone metabolites was cell line specific. All characterized cell lines contained a higher concentration (p less than 0.05) of APase activity than did young mature or senescent AXC rat ventral prostate. In 6 of 11 cell lines, prostate-secretory APase concentration exceeded (p less than 0.05) that of AXC rat ventral prostate. However, the relative content of secretory APase compared to total APase in carcinoma cells consistently was less (p less than 0.05) than that of AXC rat ventral prostate. These studies document the establishment of clonal AXC rat prostate adenocarcinoma cell lines which have retained important morphological and phenotypic markers characteristic of differentiated prostate epithelium. Since these cells are tumorigenic and represent a spectrum of retained differentiated phenotypic markers, they should be particularly useful for in vivo and in vitro studies of hormonal regulation of prostate cancer cell behavior.

Inability of cytoplasmic or nuclear androgen receptor content or distribution to distinguish benign from carcinomatous human prostate.

We used exchange saturation analysis at 15 degrees to quantitate total cytoplasmic and nuclear androgen receptor content of 70 patient specimens. Cytoplasmic androgen receptor contents (fmol/mg DNA) for eight specimens of clinically benign hyperplasia, 14 specimens of histologically hyperplastic prostate obtained at cystoprostatectomy, and carcinomatous and noncarcinomatous prostate obtained at radical prostatectomy for prostatic carcinoma, 48 specimens, respectively, were 830 +/- 165 (mean +/- S.E.), 890 +/- 445, 955 +/- 240, and 750 +/- 95. Nuclear androgen receptor contents of these same specimens, respectively, were 275 +/- 40, 235 +/- 30, 345 +/- 25, and 350 +/- 30; whereas, the values of the cytoplasmic/nuclear receptor content, respectively, were 3.25 +/- 0.55, 3.05 +/- 0.80, 2.50 +/- 0.50, and 2.80 +/- 0.40. Multiway analyses of variance of these cross-sectional data showed that there was no significant difference (p greater than 0.05) between group mean values. This result principally reflects the fact that the families of values for the four tissue groups were highly heterogenous with broad overlap. The results would not appear to be unduly influenced by carcinomatous epithelial cell content of the specimens, because cytoplasmic and nuclear androgen receptor content were not related to specimen carcinomatous epithelial cell content. Paired analyses of receptor content in carcinomatous and noncarcinomatous prostate specimens from the same prostate showed enhanced or unchanged receptor content in 58% (cytoplasmic) and 62% (nuclear) of specimens. Our studies show that cross-sectional analyses of androgen receptor content fail to distinguish carcinomatous prostate from noncarcinomatous prostate. However, paired analyses of these tissues from the same gland identify distinguishing differences. The clinical relevance of these observations remains to be examined.

Human prostate androgen receptor quantitation: effects of temperature on assay parameters.

When cytoplasmic extracts of human prostatic tissues were split to permit quantitation of total androgen receptor (RCT) content by saturation analysis at 15 degrees and 2 degrees, we observed that 30% (10 of 32) of the specimens yielded statistically increased values for RCT following incubation at 15 degrees as compared to 2 degrees. Considering only those specimens (13 of 32) showing statistically differentiated RCT yield, 77% (10 of 13) yielded greater RCT content following incubation at 15 degrees. The families of association constants (Ka) obtained for RCT determinations at 2 degrees and 15 degrees were not statistically differentiated. The increased yield of RCT content determined at 15 degrees was 95% (mean) and 20 to 350% (range). Nuclear androgen receptor content determined at 15 degrees was greater for 25% (2 of 8) of the patient specimens when compared to split determinations performed at 2 degrees. Incubation of nuclear extracts at 15 degrees resulted in a significant 3-fold reduction in receptor Ka for methyltrienolone (R1881). This did not appear to affect assay precision. These studies showed that incubation at 15 degrees is preferable to incubation at 2 degrees for quantitation of RCT and nuclear androgen receptor content by saturation analysis. Single saturating dose determinations of RCT consistently yielded underestimates. The extent of underestimate was variable from specimen to specimen and was both ligand concentration and assay temperature dependent. Our data suggest that results of single saturating dose determinations of RCT require cautious interpretation.