Use of antibiotics linked to increased risk for proximal colon cancer
Use of antibiotics appeared consistently associated with higher subsequent risk for proximal colon cancer, according to results of a nationwide, population-based study published in Journal of the National Cancer Institute.
Researchers also reported an inverse association between antibiotics use and rectal cancer among women.
“Previous studies had indicated that there was an association between antibiotics use and colorectal cancer risk. In this study, we were able to take advantage of comprehensive registry data to further break down and validate this relationship,” Sophia Harlid, PhD, research fellow in the department of radiation sciences in oncology at Umeå University in Sweden, told Healio. “Although our study provides an additional reason to restrict the use of antibiotics, when possible, the main reason for this remains the risk [for] introducing bacterial resistance.”
The analysis included 40,545 patients with colorectal cancer (mean age at diagnosis, 71.9 years; 52.9% men) from the Swedish Colorectal Cancer Register matched 1 to 5 by age, sex and county with 202,720 cancer-free controls.
Researchers used data from the Swedish Prescribed Drug Register to categorize antibiotics exposure within both groups as none (18.7% in the colorectal cancer group vs. 22.4% in the control group), low (1-10 days; 14.4% vs. 13.9%), moderate (11-60 days; 46.1% vs. 44.4%), high (61-180 days; 16.5% vs. 15.4%) and very high (> 180 days; 4.3% vs. 3.9%).
Median follow-up was 8 years.
Results showed positive associations between colorectal cancer and moderate (OR = 1.15, 95% CI, 1.12-1.18) or very high (OR = 1.17, 95% CI = 1.10-1.24) antibiotics use compared with no use (P for trend < .001).
The association was attenuated for very high vs. no use (OR = 1.02, 95% CI, 0.95 to 1.09) and no longer statistically significant in an analysis excluding antibiotics prescribed within 2 years of diagnosis for cases or index date for controls. Researchers subsequently applied the 2-year exclusion to all analyses.
Results of site-specific analyses — excluding the 2-year washout — showed the association was limited mostly to proximal colon cancer (adjusted OR for very high vs. no antibiotics use = 1.17, 95% CI, 1.05-1.31).
“Our study provides important support for the hypothesis that antibiotics can affect colon cancer risk (limited to the proximal colon),” Harlid said. “However, the risk increase for the individual is small, and not a cause for personal concern. The results bring us a little bit closer to understanding the relationship between the bacteria in our intestines and their impact on our health.”
In analysis stratified by sex, researchers reported an inverse association for rectal cancer and antibiotics use among women only with moderate use (OR = 0.86; 95% CI, 0.8-0.92 and high use (OR = 0.84; 95% CI = 0.76 to 0.94) compared with no use (P for trend < .001). Additionally, they observed a statistically significant interaction (P = .002) between antibiotics use and sex for rectal cancer, but not for proximal or distal colon cancer.
“The inverse association between antibiotics and rectal cancer in women, but not men, was not expected,” Harlid told Healio. “This very interesting finding will need to be further confirmed before it can be established as a true association. It was also a bit surprising that the association between antibiotic use and proximal colon cancer risk was so strong despite the fact that our follow-up time was only 5 to 10 years.”
Among classes of antibiotics, researchers observed associations of quinolones and sulfonamides and/or trimethoprimes with increased risk for proximal colon cancer among both men and women. This may reflect the impact of these antibiotics on bacterial diversity, they wrote.
Study limitations included possible unmeasured confounding, such as medical comorbidities and diet, and a lack of data on antibiotics administered during inpatient care and on antibiotics use before 2005 — the start of the pharmaceutical register — which resulted in less than 10 years of follow-up time, Harlid and colleagues wrote.
Harlid said further studies are needed to confirm the findings in different populations and with longer follow-up.
“There is also a need to further investigate the possibility of a causal link between antibiotics use and colon cancer risk, but this will require studies of a different design, as well as a better understanding of the interplay between microbes and host,” Harlid said.
For more information:
Sophia Harlid, PhD, can be reached at Department of Radiation Sciences, Oncology, Umeå University, SE-90187, Umeå, Sweden; email: firstname.lastname@example.org.