Risk of second cancers in chronic myeloproliferative neoplasms

To the editor:

We read with interest the article by Frederiksen et al about the increased risk of developing hematologic and nonhematologic cancers in a large cohort of Danish patients with myeloproliferative neoplasms (MPNs).¹  In a previous study²  we reported that MPN patients have a 3.44-fold higher risk of lymphoid neoplasms compared with the general population, in line with current¹  and other results.³  Considering nonhematologic cancers, Frederiksen et al reported a standardized incidence ratio (SIR) value of 1.2 (95% CI: 1.0-1.4) for essential thrombocythemia (ET) and 1.4 (95% CI: 1.3-1.5) for polycythemia vera (PV).¹  These estimates were based on 1578 ET and 4625 PV patients after a median follow-up of 4.0 and 5.0 years, respectively.

We have followed a series of 733 MPN patients, 302 PV, 375 ET, and 56 primary myelofibrosis, consecutively diagnosed at the Hematology Section, University of Florence, from 1980 to 2006, all under an active clinical follow-up protocol and residing in a defined area of Tuscany, central Italy. The identification of cancer cases was obtained through individual chart records and linkages with the hospital discharge system, pathology department registries, Regional Cancer Registry, local town offices, and Regional Mortality Registry. For each subject, the period at risk started from the date of MPN diagnosis to the date of diagnosis of subsequent primary cancer, death, or December 31, 2006 (end of follow-up), whichever came first. The mean follow-up period was 6.45 years with a total of 4724.72 person-years. The specific cancer incidence rates in the general population were provided by the Tuscany Cancer Registry, which is active in the same area since 1984 and involves 1 161 000 inhabitants; SIRs were calculated as the ratio of observed to expected cases. The study was conducted in accordance with institutional guidelines after approval by local ethics committee.

Results showed the absence of a specific pattern of risk of cancer of all sites (SIR = 0.87, 95% CI: 0.64, 1.14) except for melanoma cases for which we found a significantly elevated risk compared with the general population (SIR = 3.69, 95% CI: 1.39, 9.64; Table 1); however, because of the low number of cases (2 PV, 1 ET, and 1 myelofibrosis) such an association remains to be validated. The analyses stratified by sex (361 males/372 females), JAK2V617F mutational status (340 JAK2V617F mutated/137 wild-type), and cytotoxic therapy with hydroxyurea (59.6% of 659 evaluated) did not show any specific risk pattern (not shown in detail).

Our study is based on a smaller population compared with Frederiksen et al¹  and has the advantages of a slightly longer follow-up and of being conducted in a narrower area with an active local registry. Our carefully controlled patient series reduces the risk of possible misclassification at baseline, potentially present in a very large series obtained in a country-wide study, as mentioned by the authors.¹  In any case, we believe that the low SIR value (1.2 to 1.4) found in the Danish study,¹  together with our negative results, should call for much caution before accepting d'emblée the idea that the incidence of nonhematologic cancers is specifically increased in MPNs and, even more, before discussing such topic with the patients.