SWOG Secondary Analyses Grab Headlines

All this data is fodder for secondary analyses, i.e., fresh looks at existing quantitative or qualitative data. No one at SWOG does secondary analysis better than Dr. Joe Unger, a SWOG health services researcher and biostatistician based at Fred Hutch and a key leader on our statistics and data management center team.

Today, at the 2019 Quality Care Symposium in San Diego, Unger will give an oral presentation on SWOG results demonstrating that cancer patients who enroll in a clinical trial and have no health insurance, or are enrolled in Medicaid, the insurance program for low-income Americans, may not get the same benefits of successful trial treatments that other cancer patients do.

Unger wanted to determine whether age, sex, race and ethnicity, and insurance status had any influence on the treatment effects from positive trials. Because SWOG captures all this demographic information – and more – from trial participants, Unger could answer this vital question. To do so, he dove into data from 19 SWOG trials enrolling 11,026 patients – all large randomized phase III trials for which the experimental therapy improved overall survival.

In a separate poster presentation at the ASCO meeting, Unger will share results that show that patients from socioeconomically deprived areas appear to have worse outcomes than their counterparts, even when treated on clinical trials with access to guideline-based cancer care. To conduct the research, Unger and his team examined survival for patients enrolled in phase III SWOG clinical trials conducted between 1985 and 2012. Using participant zip codes, and the Area Deprivation Index, a measure of 17 demographic indicators, the team was able to provide evidence of this disparity in outcomes for people with cancer.

Both studies suggest that socioeconomically disadvantaged cancer patients see fewer clinical trial benefits. For details on the ASCO Quality Care work, read our press release.

This work is part of Unger’s growing secondary analysis portfolio. At the ASCO annual meeting this year , Unger shared the fruits of mining three decades of SWOG trial data that included 36,397 patients, 297 trials, and more than 500,000 adverse events. He discovered that women experience more adverse events during cancer trial treatment, particularly when given immunotherapies. In the last two years, Unger has also mined SWOG data – often in conjunction with Medicare and other large data sets – to study the effects of comorbidities on cancer trial enrollment, the survival rates of rural and urban trial participants , and the durable benefits of finasteride for reducing prostate cancer diagnoses.

Unger is, of course, not the only SWOG investigator doing amazing secondary analyses. Earlier this year, a team led by Drs. Ian Thompson (GU Chair) and Drs. Cathy Tangen and Phyllis Goodman (from the SWOG stats center) published a remarkable bit of research in the New England Journal of Medicine.  The team took on the painstaking task of mining trial data on the 18,882 men who enrolled in the Prostate Cancer Prevention Trial (PCPT) and matching participants to the National Death Index, a centralized database of death record information managed by the U.S. Centers for Disease Control and Prevention. They showed, once and for all, that there was no increased risk of prostate cancer death with finasteride. In fact, they found that more men died from prostate cancer on the placebo arm of the PCPT versus those who took finasteride.

Given SWOG’s longevity, we’re blessed with a deluge of trial data. I am proud we’re making the most of that data by reexamining it with fresh eyes, seeking answers to important questions, and expanding the way we do meaningful research.