Missed Opportunities in Ovarian Cancer Causation by Talc

Thousands of tort suits have arisen around ovarian cancer causation by talcum powder. Most recently, a California jury awarded $417M to a plaintiff, finding that Johnson & Johnson failed to warn that its product cause cancer.

A veritable "battle of the experts" has erupted around the issue, with lines being drawn in both general and specific causation concerns. JuriLytics used Rebutl to peer review both plaintiff and defense expert reports and find new attacks that both sides missed.

The points raised by peer review are substantial, and illustrate the dangers of relying solely on your testifying expert team to mount an expert attack. Independent view points, especially from researchers who know the subject, can give you the winning edge.

talcum powder and finger snap

We Peer Reviewed Both Plaintiff and Defense Expert Reports

There were serious flaws that neither side exploited.

Download the full reports and peer reviews below.

Download Examples

Plaintiff Expert Report

Dr. Roberta Ness on general causation

Defense Expert Report

Dr. David Hoel on general causation

Plaintiff Peer Review

Dr. Michele Cote's evaluation

Defense Peer Review

Dr. Michele Cote's evaluation

Key Critiques (Download Full Reports Above)

Plaintiff Report Peer Review

I would continue to note the lack of data available (or adequate sample sizes to have statistical power to detect associations) once tumors are separated by subtype. This report does mention subtypes as an issue, but does not highlight the Wentzensen et al. pooled analysis showing differences in risk factors by subtype.
Failure of most studies to have sufficient statistical power address both duration and number of applications (daily or more, weekly, etc.). The studies that do are case-control, and are thus deemed less valid (by [Report #2], and by the field in general) than the findings from cohort studies.
As the strongest evidence comes from the case-control studies, recall bias is the largest concern. The Schildkraut et al. study in 2016 suggests that recall bias is an issue in that population, as the association is only seen in cases diagnosed in 2014 and beyond.
At this time, other than additional studies showing talc in the pelvic lymph nodes of other women with ovarian cancer, there is not a clear set of experiments that could be performed to support or refute the role talc plays in ovarian carcinogenesis that are scientifically feasible.
"There is one area, regarding the interpretation of Attributable Risk Percent (AR%), that should be clarified (pages 34-35). In my understanding, AR% is the proportion of women, who were exposed to talc, who would not develop ovarian cancer if talc were eliminated as an exposure. The report currently describes Population Attributable Risk Percent which is the percentage of cases that could be eliminated from the entire population if talc were eliminated."

Defense Report Peer Review

The cohort studies that might have the ability to examine duration (via updated responses on yearly questionnaires, for example) have not done so, and follow up time may not be adequate for a disease that likely takes years if not decades to become clinically apparent (although I could find no references estimating the natural history/timeline of ovarian cancer).
A key issue in this report is the failure to consider ovarian cancer histologic subtype when examining the association between talc and ovarian cancer. The cohort studies are not adequately powered to detect potential associations for non-serous subtypes. In Wentzensen et al., a key finding was the significant heterogeneity by histologic subtype when considering various risk factors.
Talc use is a very difficult exposure to assess and measure, and this impacts the validity of the study regardless of study design. Just because it is difficult to measure, and classifications may differ across studies, does not mean there is no association. An example of this would be second hand smoke (environmental tobacco smoke) and lung cancer risk. It took several decades after the risk between personal cigarette smoking and lung cancer was established to show risk extended to those exposed to secondhand smoke. Why? It is difficult to measure compared to “number of cigarettes smoked per day” and “age started/age stopped smoking”.
I agree that confounding must be taken into account, but disagree that “these behaviors [listed above], being likely confounders, need to be properly taken into account in any talc study” (page 5). One would have to show that the behaviors or factors, say BMI and cigarette smoking, are associated with ovarian cancer in that particular study population. Using the meta-analysis of Wentzensen et al. as an example, neither BMI nor cigarette smoking are associated with increased risk of ovarian cancer in the overall population. Thus, they would not be confounders.
Various studies do show some dose-response associations, particularly the case-control studies because they are much more likely to have adequate power to examine this compared to the cohort studies. For example, Cramer et al. (Epidemiology, 2016) shows a strong trend associated with years of talc use (in quartiles) among pre-menopausal women (p=0.0006) and post-menopausal women who used hormone therapy (p=0.001). The case-control study by Cramer had approximately 3x as many women with ovarian cancer (n=2041) compared to the largest cohort study to date (NIH-AARP Diet and Health Study, n=703 cases).