Circulating Tumor Cell Counts Predict mCRC Survival

July 1, 2007

"In patients with metastatic colorectal cancer (mCRC), the number of circulating tumor cells (CTCs) before and after initiation of treatment is a significant independent predictor of progression-free survival and overall survival," Neal J. Meropol, MD, reported

ASCO—"In patients with metastatic colorectal cancer (mCRC), the number of circulating tumor cells (CTCs) before and after initiation of treatment is a significant independent predictor of progression-free survival and overall survival," Neal J. Meropol, MD, reported at the American Society of Clinical Oncology 43rd Annual Meeting (abstract 4010). In this study, patients with 3 or more CTCs per 7.5 mL of blood after 3 to 5 weeks of treatment had significantly worse progression-free survival (PFS) and overall survival (OS) than those with 2 or fewer cells per sample, said Dr. Meropol, director of the Gastrointestinal Cancer Program at Fox Chase Cancer Center.

He noted that the management of metastatic colorectal cancer is "complicated and now more nuanced," due to the multiple therapeutic options. "A noninvasive prognostic and predictive marker could guide therapeutic decisions," he said.

Dr. Meropol led a prospective multicenter international study involving 430 patients with measurable metastatic colorectal cancer scheduled to receive first-, second-, or third-line therapy. The investigators drew 7.5 mL of peripheral blood from each patient before starting therapy and at subsequent timepoints. The tumor cells were immunolabeled and counted using a fully automated immunomagnetic cell separation technique developed by Immunicon Corporation (CellSearch).

To set a threshold CTC count, the researchers analyzed data from the first 100 patients: 3 or more CTCs/7.5 mL was set as the cut-off for "unfavorable" prognosis, and 2 or fewer as "favorable."

A total of 320 patients who had undergone follow-up imaging or died were included in assessment of the primary endpoint: agreement between CTC count and response. In these patients, the number of CTCs at 3 to 5 weeks predicted radiographic response (see Table 1). "Elevated CTC number at 3 to 5 weeks has high specificity (93%) but low sensitivity (27%) for early disease progression," Dr. Meropol said. Overall agreement, or accuracy, was 78%.

Further analysis showed that the number of CTCs at time of first follow-up imaging added prognostic information, Dr. Meropol said. Patients who had not progressed and had fewer than 3 CTCs at first follow-up had overall survival of 18.8 months, compared with 7.1 months for nonprogressing patients with 3 or more CTCs. Likewise, for patients who progressed or died, overall survival was 8.3 months and 3.1 months for those with fewer than 3 CTCs and 3 or more CTCs, respectively.

Baseline levels of CTCs were significant predictors of outcome. Progression-free survival was 7.9 months vs 4.5 months, and overall survival was 18.5 months vs 9.4 months for those with favorable vs unfavorable CTC counts, respectively, at baseline. "In multivariate analyses, which included age, ECOG performance status, and line of therapy, CTC at baseline and at 3 to 5 weeks remained the most significant independent predictor of outcome," he said. CTC counts were also predictive of outcome at other timepoints (Table 2).

The study also showed that a change in CTC over time correlated with improved outcome. Median PFS was 6.2 months for those whose CTC count fell from unfavorable at baseline to favorable at 3 to 5 weeks, compared with 1.6 months for those whose CTC count was unfavorable at both baseline and 3 to 5 weeks (P = .02).

Dr. Meropol concluded that "CTC enumeration is complementary to imaging, and may provide early evidence of treatment success or failure." He noted that the study is being confirmed in a large randomized phase III trial in metastatic colorectal cancer patients receiving first-line therapy.