Short- and long-term risks of cardiovascular disease following radiotherapy in rectal cancer in four randomized controlled trials and a population-based register.

AIM: A population-based cohort and four randomized trials enriched with long-term register data were used to clarify if radiotherapy in combination with rectal cancer surgery is associated with increased risks of cardiovascular disease (CVD).

METHODS: We identified 14,901 rectal cancer patients diagnosed 1995-2009 in Swedish nationwide registers, of whom 9227 were treated with preoperative radiotherapy. Also, we investigated 2675 patients with rectal cancer previously randomized to preoperative radiotherapy or not followed by surgery in trials conducted 1980-1999. Risks of CVD overall and subtypes were estimated based on prospectively recorded hospital visits during relapse-free follow-up using multivariable Cox regression. Maximum follow-up was 18 and 33 years in the register and trials, respectively.

RESULTS: We found no association between preoperative radiotherapy and overall CVD risk in the register (Incidence Rate Ratio, IRR = 0.99, 95% confidence interval (CI) 0.92-1.06) or in the pooled trials (IRR = 1.07, 95% CI 0.93-1.24). We noted an increased risk of venous thromboembolism among irradiated patients in both cohorts (IRRregister  = 1.41, 95% CI 1.15-2.72; IRRtrials  = 1.41, 95% CI 0.97-2.04), that remained during the first 6 months following surgery among patients treated 2006-2009, after the introduction of antithrombotic treatment (IRR6 months  = 2.30, 95% CI 1.01-5.21). However, the absolute rate difference of venous thromboembolism attributed to RT was low (10 cases per 1000 patients and year).

DISCUSSION: Preoperative radiotherapy did not affect rectal cancer patients' risk of CVD overall. Although an excess risk of short-term venous thromboembolism was noted, the small increase in absolute numbers does not call for general changes in routine prophylactic treatment, but might do so for patients already at high risk of venous thromboembolism.