TOPLINE:
Late-stage cancer incidence may be a suitable alternative endpoint to cancer-specific mortality in lung and ovarian cancer screening trials, but not in breast, colorectal, or prostate cancer screening trials, new research showed.
METHODOLOGY:
- Randomized controlled trials of cancer screening typically set cancer-specific mortality as the primary endpoint. Using late-stage cancer as a primary endpoint instead may be useful if it provides similar results as cancer-specific mortality. The potential advantage of evaluating reductions in late-stage cancer is that it can shorten the duration of a trial.
- Researchers assessed the value of using stage III or IV cancer vs reductions in cancer-specific mortality as endpoints in a meta-analysis of 41 randomized controlled trials that looked at the benefits of screening for breast (n = 6), colorectal (n = 11), lung (n = 12), ovarian (n = 4), prostate (n = 4), and other cancers (n = 4). Trials explored a range of screening modalities including mammography for breast cancer, fecal occult blood test for colorectal cancer, prostate-specific antigen for prostate cancer, and low-dose CT for lung cancer.
- Specifically, the study assessed whether evaluating reductions in the incidence of late-stage cancer vs reductions in cancer-specific mortality would lead to similar conclusions about screening efficacy, the authors explained.
- The researchers conducted analyses combining the data overall and evaluating by cancer type; they used Pearson correlation coefficients, linear regression, and fixed-effects meta-analysis to compare the two endpoints.
TAKEAWAY:
- Overall, the correlation between the effects of cancer screening on cancer-specific mortality and stages III-IV cancer incidence varied across trials and by cancer type. Five trials (12%) found that the specific screening modality under investigation was associated with a reduction in cancer mortality and incidence of stages III-IV cancer; 24 trials (59%) did not demonstrate a reduction in either. In eight trials (20%), screening demonstrated reductions in stages III-IV cancer incidence but not cancer mortality, and in four trials (10%), the opposite occurred.
- When looking at cancer type, the researchers found that a correlation between reductions in cancer-specific mortality and stages III-IV cancer incidence was strongest in ovarian cancer trials (Pearson ρ, 0.99) and lung cancer trials (Pearson ρ, 0.92).
- A correlation between the two endpoints was moderate for breast cancer (Pearson ρ, 0.70) and weak for colorectal cancer (Pearson ρ, 0.39; 95% CI, −0.27 to 0.80) and prostate cancer (Pearson ρ, −0.69; 95% CI, −0.99 to 0.81).
- Given the varied results by cancer type, the authors extrapolated that, in screening trials evaluating multicancer early detection tests, using late-stage cancer incidence as an endpoint instead of cancer mortality "is likely to be invalid."
IN PRACTICE:
"For clinical trials that screen for lung and ovarian cancers, incidence of stage III-IV cancer may be a suitable alternative endpoint to cancer-specific mortality," the authors said. However, "stage III-IV cancer incidence is unlikely to be a suitable alternative endpoint to cancer-specific mortality in screening trials for breast, colorectal, and prostate cancers."
Based on the findings of the current study, "cancer-related mortality remains the most appropriate endpoint for clinical evaluation of the new blood-based tests that aim to detect many cancers for which there is no evidence that screening is beneficial,” wrote Peter Bach, MD, with DELFI Diagnostics Inc., Baltimore, Maryland, in an accompanying editorial. These "studies might take a little longer, but will yield a more reliable answer."
SOURCE:
The study, led by first author Xiaoshuang Feng, PhD, with the Genomic Epidemiology Branch, International Agency for Research on Cancer, Lyon, France, and an accompanying editorial from Peter Bach, MD, were published online on April 7 in JAMA.
LIMITATIONS:
The sample size for each cancer type was small, and the study did not account for sampling variability in estimated effects on late-stage cancer and cancer mortality. The study did not account for differences across trials in the number of participants, number of outcomes, or length of follow-up. Observations from randomized trials of screening for individual cancers may not apply to multicancer screening.
DISCLOSURES:
The study was funded by the International Agency for Research on Cancer. The authors reported no conflicts of interest. Bach reported stock ownership in and being company executive of DELFI Diagnostics, a for-profit company that is developing blood-based cancer screening tests, including a test for lung cancer screening.