Neither time to nadir nor depth of nadir emerged as a valid surrogate of overall survival (OS) across different contemporary first-line regimens for metastatic colorectal cancer (mCRC), according to pooled data from 20 randomized trials.
With tumor nadir defined as the local minimum relative tumor-size change (RTSC) from baseline, a group of European, Australian, and U.S. researchers led by Tomasz Burzykowski, PhD, of the International Drug Development Institute in Louvain-la-Neuve, Belgium, reported the following:
- For 14 chemotherapy-alone trial comparisons involving 4,289 patients, the estimated value of R2 (coefficient of determination, with values closer to 1.00 indicating stronger association) was 0.63 (95% CI 0.30-0.96) for the association between treatment effects on time to nadir and OS; for the association between effects on depth of nadir and OS, the R2 was 0.08 (95% CI 0-0.37)
- For 11 comparisons in 4,854 patients in which at least one anti-angiogenic agent was added to chemotherapy, the corresponding R2 values were 0.25 (95% CI 0-0.72) and 0.06 (95% CI 0-0.35)
- In eight comparisons in 2,684 patients adding anti-epidermal growth factor receptor (EGFR) drugs, the R2 values were 0.24 (95% CI 0-0.83) and 0.21 (95% CI 0-0.78)
These findings suggest that response-based endpoints cannot replace OS in clinical trials, the authors wrote in the study online in JAMA Network Open. “The implications of these results for early drug development and clinical practice are unclear and warrant further studies; the findings of this study reinforce the need to develop more reliable endpoints that reflect tumor biology and patient benefit.”
However, the team said, the different effects of adding anti-angiogenesis and anti-EGFR agents to chemotherapy that emerged in the study merit further investigation.
The study also showed that the addition of an anti-angiogenesis agent to chemotherapy was associated with a later, though not often deeper, nadir. Conversely, the addition of an anti-EGFR agent frequently produced a deeper nadir, with less conclusive results about the timing of occurrence. These differences in response kinetics warrant further investigation, Burzykowski and associates said.
They pointed out that with the current continuum of care with active subsequent lines of therapy for mCRC, it is becoming increasingly difficult to demonstrate gains in OS from first-line treatment trials. This problem has spurred interest in alternative strategies, including surrogate endpoints based on tumor measurements. But that approach would appear of little use, according to the new study.
The studies, all published during the period 2004 to 2012, were drawn from the database of France’s Aide de Recherche en Cancérologie Digestive (ARCAD). Three sets of comparisons were performed:
- Chemotherapy plus anti-angiogenic agents
- Anti-EGFR agents in first-line treatment for mCRC
Surrogacy of time to nadir and depth of nadir was assessed in two ways:
- At the trial level, based on joint modeling of relative tumor-size change vs baseline and OS
- At the patient level, in which the correlation was assessed between modeled RTSC and OS
For chemotherapy, there was an indication that RTSC may provide a strong prediction of survival, the authors said, adding that for anti-angiogenesis, a strong correlation might be inferred after the initial half-year of treatment, but for anti-EGFR agents, the correlation was weak.
“These findings are not surprising, given the weak trial-level association between conventional response rates and OS in mCRC, despite their association with OS at the patient level, both in mCRC and advanced breast cancer,” Burzykowski, et al. wrote. While achieving response may convey prognostic information for patients in clinical practice, it also suggests that response-based endpoints cannot replace OS in clinical trials, the team added.
‘Search Must Continue’
Asked for his perspective, Van Morris, MD, of the University of Texas MD Anderson Cancer Center in Houston, said the ARCAD findings show that the search must continue for informative early surrogates for OS in clinical trials for mCRC.
“The weak associations with overall survival reported here – for example, depth of nadir for anti-angiogenic agents — could reflect intratumoral pathologic changes not captured radiographically by the criteria tested here,” said Morris, who was not involved in the study. “Better surrogates are warranted, and perhaps dynamic changes in circulating tumor DNA levels may provide further insight in future work.”
The use of surrogate endpoints rather than OS in trial design has long been criticized for shortening the duration of studies in order to bring new oncology drugs to market sooner.
Among study limitations, the chief one was the absence of tumor measurements for all patients, Burzykowski and co-authors said, which was a potential source of bias through exclusion of individuals with features that may have differed systematically from those of included patients.
Moreover, extended RAS gene testing was not available at the time of the trials, and that led to a predictably small percentage of patients being falsely considered as having wild-type tumors, the researchers said. In addition, no data were available on tumor sidedness or other potential prognostic or predictive molecular markers, such as the status of microsatellite instability, BRAF, or HER2. Another limitation pertained to the study’s model building, which was restricted by the absence of post-progression measurements.
Data and funding were provided by the Fondation Aide et Recherche en Cancérologie Digestive.
Burzykowski reported having no conflicts of interest; numerous co-authors disclosed various financial ties to industry.
Morris reported having no conflicts of interest in relation to his comments.