More frequent follow-up with computed tomography of the thorax and abdomen and serum carcinoembyronic antigen (CEA) testing does not significantly improve mortality rates or improve time to detection of recurrence, results of two studies published in JAMA have suggested.*
In COLOFOL, a randomized clinical trial including more than 2,500 patients with stage II or III colorectal cancer, more frequent follow-up with CT of the thorax and abdomen and serum CEA did not significantly improve 5-year overall mortality or colorectal cancer–specific mortality rates.
In the second study, a retrospective cohort analysis of the National Cancer Database (NCDB) including more than 8,500 patients with stage I-III colorectal cancer, investigators found no significant association between the surveillance testing frequency and time to detection of disease recurrence.
Taken together, these findings suggest a need to revisit clinical practice guidelines, Hanna K. Sanoff, MD, of the University of North Carolina at Chapel Hill said in a related editorial (for details, see “Views on the News”).
The COLOFOL randomized trial, reported by Henrik T. Sørensen, DMSc, head of the department of clinical epidemiology at Aarhus (Denmark) University Hospital, and his colleagues, included 2,509 patients with stage II or III colorectal cancer.
“The question of appropriate follow-up intervals has been controversial, and varying intensity of follow-up has been used within and among countries,” Dr. Sørensen and his coauthors said.
Patients were randomized either to a high-frequency group, in which CT and CEA testing were conducted at 6, 12, 18, 24, and 36 months after surgery, or to a low-frequency group that received testing only at 12 and 36 months after surgery.
Results of COLOFOL showed that the 5-year colorectal cancer–specific mortality rate was similar: 10.6% for the high-frequency follow-up group versus 11.4% for the low frequency group (risk difference, 0.8%; 95%confidence interval, –1.7% to 3.3%; P = .52).
Likewise, 5-year overall mortality was 13.0% for the high-frequency group and 14.1% for the low-frequency follow-up groups (risk difference, 1.1%; 95% CI, –1.6% to 3.8%; P = .43
High-intensity testing did result in recurrences being detected earlier; nevertheless, this did not translate into a reduced mortality rate, investigators said.
The retrospective NCDB analysis, reported by George J. Chang, MD, of University of Texas MD Anderson Cancer Center, Houston, and his coauthors, included 8,529 patients with stage I-III colorectal cancer treated at 1,175 facilities.
Facilities designated as high intensity for imaging performed a mean of 2.87 imaging tests over 3 years, compared with 1.63 for facilities designated as low intensity. Median time to detection of recurrence was similar between arms, at 15.1 months for patients treated at centers with high-intensity surveillance versus 16.0 months for those treated at low-intensity surveillance centers (hazard ratio, 0.99; 95% CI, 0.90-1.09).
High-intensity CEA testing facilities performed a mean of 4.31 tests within 3 years versus 1.63 for low-intensity facilities. Again, investigators found similar median time to detection of recurrence for high- and low-intensity facilities (15.9 months versus 15.3 months, respectively; hazard ratio, 1.00; 95% CI, 0.90-1.11)
Previously, the Follow-up After Colorectal Surgery (FACS) study, a randomized controlled trial, showed no survival benefit to more frequent testing, Dr. Chang and his colleagues noted.
“Based on these data and the recent FACS trial, current National Comprehensive Cancer Network (NCCN) guideline recommendations could be considered overtesting given the absence of improvement in recurrence detection or survival,” they wrote, noting that the NCCN guidelines have suggested CT testing every 6 months for 3 years.
Disclosures for the COLOFOL trial included one investigator who reported potential conflicts of interest with Janssen-Cilag and Merck Serono. For the NCDB study, one coauthor reported a potential conflict of interest related to Johnson & Johnson. No other disclosures were reported.
SOURCES: Sørensen HT et al. JAMA Oncol. 2018 May 22. doi: 10.1001/jama.2018.5623; Chang GJ et al. JAMA Oncol. 2018 May 22. doi; 10.1001/jama.2018.5816.
Correction, 6/8/18: An earlier version of this article misstated the name of the journal in which this study was published.