Colorectal Adenomas and Cancers After Childhood Cancer Treatment: A DCOG-LATER Record Linkage Study.

Abstract

RESULTS

Among 5843 five-year survivors (median follow-up = 24.9 years), 78 individuals developed an adenoma. Cumulative incidence by age 45 years was 3.6% (95% confidence interval [CI] = 2.2% to 5.6%) after abdominopelvic radiotherapy (AP-RT; 49 cases) vs 2.0% (95% CI = 1.3% to 2.8%) among survivors without AP-RT (28 cases; Pdifference = .07) and vs 1.0% (95% CI = 0.3% to 2.6%) among siblings (6 cases) (Pdifference = .03). Factors associated with adenoma risk were AP-RT (hazard ratio [HR] = 2.12, 95% CI = 1.24 to 3.60), total body irradiation (TBI; HR = 10.55, 95% CI = 5.20 to 21.42), cisplatin (HR = 2.13; 95% CI = 0.74 to 6.07 for <480 mg/m²; HR = 3.85, 95% CI = 1.45 to 10.26 for ≥480 mg/m²; Ptrend = .62), a hepatoblastoma diagnosis (HR = 27.12, 95% CI = 8.80 to 83.58), and family history of early-onset CRC (HR = 20.46, 95% CI = 8.10 to 51.70). Procarbazine was statistically significantly associated among survivors without AP-RT/TBI (HR = 2.71, 95% CI = 1.28 to 5.74). Thirteen CRCs occurred.

CONCLUSION

We provide evidence for excess risk of colorectal adenomas and CRCs among childhood cancer survivors. Adenoma risk factors include AP-RT, TBI, cisplatin, and procarbazine. Hepatoblastoma (familial adenomatous polyposis-associated) and family history of early-onset CRC were confirmed as strong risk factors. A full benefit-vs-harm evaluation of CRC screening among high-risk childhood cancer survivors warrants consideration.

BACKGROUND

Although colorectal adenomas serve as prime target for colorectal cancer (CRC) surveillance in other high-risk groups, data on adenoma risk after childhood cancer are lacking. We evaluated the risk of histologically confirmed colorectal adenomas among childhood cancer survivors. A secondary aim was to assess CRC risk.

METHODS

The DCOG-LATER cohort study includes five-year Dutch childhood cancer survivors and a sibling comparison group (n = 883). Colorectal tumors were identified from the population-based Dutch Pathology Registry (PALGA). We calculated cumulative incidences of adenomas/CRCs for survivors and siblings. For adenomas, multivariable Cox regression models were used to evaluate potential risk factors. All statistical tests were two-sided.