Limited data suggest previously unrecognized risk
Multiple sclerosis patients taking the drug mitoxantrone had a higher incidence of colorectal cancer and acute myeloid leukemia (AML) than the general population, researchers in Germany found. Patients also had “mildly increased” incidence of other malignancies.
In a German cohort of 676 mitoxantrone-treated patients with MS who had complete follow-up, rates of colorectal cancer were three times higher (standardized incidence ratio 2.98, 95% CI 1.20-6.14) than in the general population, Mathias Buttmann, MD, of the University of Wuerzburg, and colleagues reported online in the journal Neurology.
Buttmann and his team used electronic records and manual chart review to identify 677 patients taking the drug who had been treated at the University of Wuerzburg from 1994 to 2007. Of those, 676 had good follow-up data and were included in the study. Median follow-up time was 8.7 years, and median cumulative mitoxantrone dose was 79 mg/m2. About one-third had purely relapsing MS when starting mitoxantrone.
Rates of AML were 10 times higher (SIR 10.44, 95% CI 3.39-24.36) than in the general population, the researchers also found. The link between mitoxantrone and AML in MS patients had been known “for years,” Buttmann told MedPage Today, but he and his colleagues wanted to study it further and search for links to other cancers.
“Our main challenge was to locate as many as possible of our patients and to collect follow-up information,” Buttmann said. “It took us years of detective work but we finally were successful in 676 out of 677 identified patients, which we initially hadn’t dared to hope for in this retrospective study.”
The researchers learned that 37 of the patients, 5.5%, had developed malignancies of some kind. Nine had breast cancer, seven had colorectal cancer, four had AML, two had each glioblastoma multiforme, lung, pancreatic, or prostate cancer, and one each had nine other kinds of cancer.
Paul Wright, MD, a neurologist at Northwell Health in New York, who was not involved in the study, said the colorectal cancer revelation from the study will be a surprise to doctors who treat MS patients. Although a larger follow-up study would provide more information, he said Buttmann and his colleagues’ work is enough to raise awareness about the possibility of other cancers in patients receiving this drug.
“Because colon cancer is very easy to screen for, and at certain ages it should be done anyway, I think physicians should use their clinical judgment to discuss it with patients and offer [additional screenings] to them,” Wright said.
AML patients’ median age was 38. The median length of time between starting mitoxantrone and developing AML was 35 months, but one of the four patients had a latency of more than 20 years. That patient also received another immunosuppressive drug prior to mitoxantrone.
All the patients with AML were alive at the end of the follow-up period, having experienced “full remission,” according to the study.
The colorectal cancer patients’ median age was 58. The median length of time between starting the drug and being diagnosed with cancer was 74 months. They were diagnosed at advanced stages and three died of their cancer.
In all, 55 of the 676 patients died over the course of the study, 12 of them from cancer.
For any malignancy, the SIR with mitoxantrone exposure for MS patients relative to the general German population was “mildly” increased. The only additional risk factor was older age at start of mitoxantrone, said the researchers; other potential confounders for which they had data did not appear to influence the results.
The study’s main limitation is the lack of good pre-existing data on the incidence of cancers in patients with MS; the researchers didn’t have similar cancer-incidence data for German MS patients not taking mitoxantrone.
“Because of inconsistent data on the incidence of malignancies in patients with MS as compared to the general population, it remains unclear whether and how the chosen control group influenced our results,” the authors wrote.
Other limits of the study included a small cohort size, but the authors wrote that they avoided underreporting and bias because they were searching for significant diagnoses that would have likely been reported, and they sought redundant information from multiple sources. The authors concede that they may still have underestimated mitoxantrone’s link to cancer development because they could have missed undiagnosed malignancies and that additional cancers may have developed after the follow-up period.
Buttmann told MedPage Today that the study results should be taken with “more than one grain of salt” because the absolute number of colorectal cancer cases was so low.
“Despite statistical significance when comparing to the general population, we may have observed an increased incidence of colorectal cancer just by chance and our results definitely need independent confirmation,” he said.
Buttmann said it’s too soon for clinicians to draw practical conclusions from the study and he said further work will confirm or refute his findings. He said he’ll let his patients know about the potential increased risk of colorectal cancers, but remind them that these results were only found his small study.
“Right now I would stick to the general guidelines for colorectal cancer screening but not recommend additional invasive safety measures to mitoxantrone-treated patient,” he said.
University research funds sponored the study, for which the authors report no pharmaceutical funding.
Buttmann was a consulant for Teva Pharma and Roche and has recieved clinical research grants from Merck Serono and Novartis. He’s also recieved speaker fees from Biogen, and travel grants from Bayer HealthCare Pharmaceuticals, Genzyme, and Novartis.
By Syndney Lupkin
