Prevalence and Risk Predictive Factors of Comorbid Malignancies in Patients with Chronic Myeloid Leukemia

Introduction: As chronic myeloid leukemia (CML) patents are generally diagnosed at old age and live longer by active use of BCR-ABL1 tyrosine kinase inhibitors (TKIs), the occurrence of other malignancy (OM) is becoming a critical issue as a long-term comorbidity. An increased rate of OM has been reported in myeloproliferative disorders and long-term TKI treatment may induce OM in CML. To explore exact prevalence and characteristics of OM, we reviewed medical records of CML patients and compared with those of age-matched Korean population.

Methods: The medical records of 1,469 CML patients who diagnosed between January 2000 and December 2014 were reviewed using Korean data-set of Asia CML Registry (ACR). With a cut-off date of July 2016, age-standardized prevalence rates (A-SPR) of OM (except benign tumors and other leukemias) were analyzed and compared with that of general population in Korea Central Cancer Registry (KCCR). In addition, we analyzed cumulative incidence rate of OM and various risk factors.

Results: The median duration of follow-up was 84 (1-197) months, and 96 CML patients had at least one OM. Forty three patients had a history of OM before a median 69 (1-161) months of CML diagnosis and 53 patients developed OM after a median 53 (range; 0.2-172) months of CML diagnosis. The OM included 32 thyroid cancers, 19 colorectal cancers, 16 stomach cancers, 9 breast cancers, 4 gynecological cancers (3 cervical cancers and 1 uterine endometrial cancers), 3 lymphoma (2 non-Hodgkin's lymphoma and 1 Hodgkin's lymphoma), 3 biliary cancer, 3 skin cancers, 3 prostate cancers, 2 lung cancer, 2 tongue cancer, 2 liver cancer, 2 esophageal cancer, 1 pancreatic cancer, and 1 bladder cancer.A-SPR of OM was 1.7 times higher in CML patients. Hodgkin's lymphoma (8.7 times), thyroid cancer (2.6 times), biliary cancer (2.6 times), colorectal cancer (2 times), non-Hodgkin's lymphoma (1.8 times), cervical cancer (1.8 times), and breast cancer (1.6 times) had a higher A-SPR. On the other hands, skin cancer (3.3 times), lung cancer (2 times), and liver cancer (2 times) were lower than that of general population. With 53 patients who had OM after CML diagnosis, we analyzed the cumulative incidence. The risk of OM was increased over the follow-up period (2.7% at 7 years) Univariate analysis revealed that patients who were more than 37 years old at CML diagnosis (4.3% vs. 0.4%, p<0.001) and who had family history of cancer (8.2% vs. 2.3%, p=0.002) were associated with a higher OM. After adjusting for factors, multivariate analysis showed that older age (HR of 4.19, P<0.001) and family history (HR of 3.17, P=0.001) were independently associated with increased risk. There was no difference in 7-year overall survival (OS) between patients with OM (n=96) and without OM (n=1,373) (84.9% vs. 86.9%, p=0.573). However advanced cancer stages (stage 3 and 4) of OM significantly affected poor OS ( 88.3% vs. 65.6% P=0.0406).

Conclusion: Although comorbid malignancies did not significantly affect CML survival, poor survival in advanced stages and the high risk of other cancers warn the need of systematic screening in long-term CML survivors. In addition, the specific cancer types with a significantly higher A-SPR should be considered for further studies including genetic mechanisms.