jnci.oxfordjournals.org JNCI | Correspondence 137 Response

As Dandona et al. note, the recent genome-wide association study from the PanScan collaborators ( 1) produced an important finding validating the association of blood type and incident pancreatic cancer in a dif-ferent population; however, further studies such as the one by Dandona et al. are clearly necessary to verify or refute our findings from the Nurses ’ Health Study and Health Professionals Follow-up Study. In addition, the investigation of ABO blood group antigens in relation to survival is interesting because patients with this disease have a very poor prognosis; novel risk markers and bio-logical insights are needed if we are to improve patient outcomes. Furthermore, previous studies have demonstrated associa-tions of blood group antigens with outcomes in other malignancies ( 2 – 4). We agree with Dandona et al. that further investigations are warranted of ABO blood group and survival in patients with pancreatic cancer