In patients with metastatic colorectal cancer (mCRC), the correlation between tumor location and improved survival varies based on stage of disease, according to a study.
Deborah Schrag, MD
Deborah Schrag, MD
In patients with metastatic colorectal cancer (mCRC), the correlation between tumor location and improved survival varies based on stage of disease, according to a results of a recent population-based study.
The study was designed to corroborate the findings from the CALGB/SWOG 80405 phase III trial, which demonstrated that survival outcomes in patients withKRASwild-type metastatic mCRC were significantly longer among those with tumors originating on the left versus the right side of the colon.
In the population-based analysis, researchers determined that although there was an association between left-sided tumors and improved progression, it was not consistent across all stages of disease.
“The relationship between tumor location and prognosis varies by stage, with right-side stage IV colon cancer having worse survival than left," said Deborah Schrag, MD, chief, Division of Population Sciences, Department of Medical Oncology, Dana-Farber Cancer Institute, who presented the data at the 2016 ASCO Annual Meeting. “This association is weaker and attenuated for stage III and is not present for stage II CRC. Lead time bias and older age certainly contribute to this effect.”
Data included in the study came from patients with stage II, III, and IV incident primary CRC from 18 tumor registries in the National Cancer Institute Surveillance, Epidemiology, and End Results Program (SEER) program from 2002 to 2012. A subgroup of patients linked to Medicare claims between 2004 and 2011 and treated with either cetuximab (Erbitux) or panitumumab (Vectibix) was also evaluated.
“The strength of this data is that it is a population-based registry with large representative cohorts with detailed information including AJCC (American Joint Committee on Cancer) stage and prospectively ascertained information about primary tumor location and stage, said Schrag. “Unfortunately, there are no molecular data, includingKRASstatus, and chemotherapy treatments are available only for the subgroup enrolled in Medicare.”
The study included 138,069 patients with primary right-sided CRC with a median age of 70.2 years, and 55% of these patients were female. Among the 97,311 patients with primary left-sided colon cancer, the median age was 65.4 and 46% were female.
Previous studies have shown that right-sided colon cancers tend to appear at an older age. In this study, the categorization of primary CRC sites was done slightly differently than other studies, said Schrag. Right-sided colon cancer was defined as cancer that is located in the cecum, ascending colon, or transverse colon, and left-sided colon cancer was defined as cancer of the splenic flexure, descending colon, or sigmoid colon.
The study found that tumor location for stage III/IV CRC has not shifted in distribution since 2000. Between 2000 and 2003, 39% of patients had right-sided tumors and 29% of patients had left-sided tumors. From 2004 to 2008, 38% of patients had right-sided tumors and 27% had left-sided ones. Finally, from 2009 to 2012, 37% of patients had right-sided tumors and 27% had left.
In patients with stage IV CRC (n = 64,770), the median survival was 9.5 months in those with right-sided tumors and 15.5 months in those with left-side tumors. The unadjusted hazard ratio (HR) was 1.32 (95% CI, 1.30-1.35) and the adjusted HR was 1.25 (95% CI, 1.22-1.27) compared with left-sided tumors in this population.
For those with stage III CRC (n = 91,009) the unadjusted HR was 1.35 (95% CI, 1.32-1.38) and the adjusted HR was 1.12 (95% CI, 1.09-1.14) in those with right-sided tumors compared with left-sided tumors. Median survival was 62.5 months in those with right-sided tumors and 93.5 months in those with left-sided tumors in this patient population.
Tumor location had the smallest impact on patients with stage II CRC, results showed. There was an unadjusted HR for patients with right-sided versus left-sided CRC of 1.06 (95% CI, 1.03-1.08) and an adjusted HR of 0.89 (95% CI, 0.87-0.92).
“In stage II CRC there was much less of an impact, in fact, the hazard ratio flips the other way and there is no meaningful difference in survival,” said Schrag.
Based on these findings, researchers then investigated whether lead time may account for the survival differences between the right- and left-sided tumors, said Schrag.
“Left-sided tumors bleed and symptoms start earlier than right-sided tumors,” said Schrag. “Right-sided tumors cause occult bleeding and symptoms often start later. That difference is lead time may artificially make survival appear longer for left-sided tumors. We took a look at this by diving deeper and examining substage.”
Substage was broken into IIIa (n = 7636), IIIb (n = 33,957), IIIc (20,628), and III NOS (28,788). The most significant difference was seen between IIIa and IIIc, said Schrag, as 32% of patients had right-sided tumors in substage IIIA compared with 45% of patients in substage IIIc.
“This provides some evidence of stage migration and potential lead-time bias,” said Schrag. “When looking at the survival curves split out by substage IIIa, IIIb, and IIIc, it can be seen that most of the survival difference is contributed by the stage IIIc patients, with an adjusted hazard ratio of 1.32.”
Schrag D, Weng S, Brooks G, et al. The relationship between primary tumor sidedness and prognosis in colorectal cancer.J Clin Oncol34, 2016 (suppl; abstr 3505). Abstract 3505.