GENEVA – Circulating tumor cell (CTC) count is an independent predictor of both progression-free and overall survival in patients with advanced non–small cell lung cancer (NSCLC), according to data from 550 patients.
Will Pass/MDedge News
Dr. Colin Lindsay
This is the largest CTC study to date and the first to compare test results from multiple centers, reported lead author Colin Lindsay, MD, PhD, of the University of Manchester (England) and colleagues. Among the centers, investigators found minimal variability in results guiding progression-free survival and no significant differences in results predicting overall survival. These findings suggest that CTC testing could be reproducible and reliable on a large scale, Dr. Lindsay said during a presentation at the European Lung Cancer Conference; he added that this conclusion addresses a previous concern about the process.
“A slight problem with the process is still that it is semi-automated,” Dr. Lindsay said at the meeting presented by the European Society for Medical Oncology. “The machine will harvest potential cells and stain potential cells, but the end step of the process is that a trained user in each laboratory will decide which cell is a CTC and which cell isn’t a CTC, and it’s that potential for user variability that was the basis of this study.”
The retrospective study involved 550 patients with NSCLC whose samples were processed at seven centers in multiple European countries, including 209 patients whose data was previously unpublished. The investigators looked for associations between CTC count and survival using Cox regression analysis and evaluated if CTCs could add value to prognostic clinicopathologic models based on c-indices and likelihood ratio statistics. CTC count was assessed as a continuous variable and, based on previous studies, using two categorical thresholds: at least 2 cells per 7.5 mL and at least 5 cells per 7.5 mL. In addition, the investigators looked for associations between NSCLC molecular subtypes and CTC levels.
The results showed that both cutoff levels were predictive of survival, with the higher threshold carrying a poorer prognosis. For progression-free survival, CTC counts of at least 2 cells per 7.5 mL carried a hazard ratio of 1.72, whereas the 5-cell threshold had a hazard ratio of 2.21 (P less than .001 for both). Similarly, overall survival hazard ratios for the lower and higher thresholds were 2.18 and 2.75, respectively (P less than .001 for both). When baseline CTC count was added to the analysis, predictive accuracy increased further, dropping P values tenfold, down to .0001. C-index models had a more modest impact. Although minor heterogeneity was detected among centers for prediction of progression-free survival, overall survival data was broadly reliable. Dr. Lindsay noted that intercenter differences seemed to diminish with greater testing experience. No relationships were detected between molecular subtypes and CTC profiles.
“It’s always good to finish a talk with the white elephant in the room,” Dr. Lindsay said in his concluding remarks. “Is there room for CTCs in non–small cell lung cancer? I believe they have the potential to complement ctDNA work by offering a cellular context, but [CTCs] aren’t there yet for clinical roll-out.”
Will Pass/MDedge News
Dr. Juergen Wolf
Invited discussant Juergen Wolf, MD, of the University Hospital Cologne (Germany) provided a similar conclusion, suggesting that CTCs have a clear place in research, but their clinical value is debatable. He noted that ctDNA, the most similar diagnostic and prognostic tool under development, has a pragmatic edge because ctDNA samples are more amenable to shipping and handling. Dr. Wolf noted that ctDNA also has been shown to have value for treatment planning, specifically for the EGFR T790M resistance mutation. This latter point tied into a larger issue described by Dr. Wolf, who suggested that in the current treatment landscape for NSCLC, predictive testing needs to be actionable.
“We cannot draw a consequence of a prognostic biomarker,” Dr. Wolf said. “In the era of personalized medicine, what we need is predictive markers, predictive of the outcome of specific therapies.”The investigators disclosed financial relationships with AstraZeneca, Novartis, Pfizer, and others.
SOURCE: Lindsay C et al. ELCC 2019, Abstract 21O.