SAN DIEGO – Current thiopurine use is an independent risk factor for nonmelanoma skin cancer in patients with ulcerative colitis, according to a study reported at the annual meeting of the American College of Gastroenterology.
In the retrospective cohort study of more than 14,000 U.S. veterans with ulcerative colitis, current users of thiopurines were more than twice as likely as never users to receive a diagnosis of squamous or basal cell skin cancer. The excess risk, however, disappeared after stopping thiopurines.
"It is crucial to educate physicians and patients about the risk of nonmelanoma skin cancer and possible preventive measures," asserted Dr. Ali Abbas, an internal medicine resident at * University of Florida, Gainesville.
Dr. Ali Abbas
Session comoderator Dr. Stephen B. Hanauer of the University of Chicago noted, "The current quality measures in [inflammatory bowel disease] include a number of measures related to biologic therapy – immunization/vaccination, [tuberculosis] testing, etc. ... Do you believe that yearly skin examination should be added? Is there enough evidence that this should be added to the quality indicators?"
"I think now we have enough evidence to recommend regular skin examination for those on long-term thiopurines," Dr. Abbas replied, while adding that the reversibility of risk after stopping remains controversial.
Dr. Hanauer further noted that number of health care visits also predicted nonmelanoma skin cancer risk. "I would suspect that, while patients are on thiopurines, they are getting more intensive visits, so can you dissociate those factors?"
Multivariate analyses took into account the number of visits, according to Dr. Abbas. "Even after exclusion of this confounding effect, we had a two times increase of the risk," he said. "If you compare our hazard ratio with previously reported hazard ratios, it’s lower – most of them reported were 4 to 5. So I am assuming that we kind of excluded the effect of the detection bias and we present a more independent effect of thiopurines on risk of nonmelanoma skin cancer."
"Do you know what the lag time would be then from exposure to the development of a nonmelanoma? So is it biologically plausible to understand that this would be reversible in patients who stop therapy?" asked Dr. David T. Rubin, codirector of the inflammatory bowel disease center at the University of Chicago, the other session comoderator.
The investigators did not calculate lag time, Dr. Abbas replied. "The biological origin is just the interaction between the UV [ultraviolet] radiation on the skin and the damage that the thiopurine causes to the DNA will prevent the repair. So from a biological point of view, I think it’s understandable to conclude that, once the insult to the DNA or the repair mechanism is gone, the UV radiation effect will also be low."
"Well, I think there is another biologic plausibility," Dr. Hanauer suggested, "which is the potential for viral infections such as [human papillomavirus] contributing to skin cancers, and the known effect of thiopurines on viruses. So certainly, UV is a very strong association, but we are also familiar with individuals who have skin warts."
"You do need to understand lag time between exposure and subsequent neoplasia before you conclude that when you are off therapy, you don’t need screening anymore," Dr. Rubin said.
Upcoming analyses will stratify patients according to the location of skin cancer, for example, looking at anogenital skin cancers separately, Dr. Abbas replied.
In the study, the investigators reviewed pharmacy benefits records for 14,527 veterans with an ulcerative colitis diagnosis who were seen in the Veterans Affairs health care system between 2001 and 2011.
They were 59 years old on average at baseline; 94% were male and 77% were white. Their median follow-up was 8.1 years, according to Dr. Abbas.
Overall, 23% of the patients used thiopurines, for a median of 1.6 years during follow-up, and the median duration of follow-up after stopping these medications was 3.6 years. The median duration of follow-up among thiopurine-unexposed patients was 6.7 years.
Main results showed that the incidence of nonmelanoma skin cancer was 3.7 per 1,000 person-years among unexposed patients, 8.4 per 1,000 person-years during thiopurine use, and 3.0 per 1,000 person-years after stopping thiopurines.
In a multivariate model, patients had a significantly higher risk of nonmelanoma skin cancer while taking thiopurines when compared with never-users (hazard ratio, 2.1). There also was a nonsignificant trend toward reduced risk after stopping, as compared with never users (0.7).
Other factors associated with higher risk included older age, male sex, white race/ethnicity, living in zones with high UV exposure, and more frequent use of the VA health care system.