CA 19-9 and Lewis antigens in pancreatic cancer.

rituximab, which supports a similar incidence of HBV reactivation in our reported study. Nonetheless, we agree with Koo et al that in addition to the less informative anecdotal experience and isolated case reports, more information on experience of oncologists is needed. HBV reactivation is a phenomenon that includes two well-known major categories (ie, HBsAg positive v HBsAg negative/anti-HBc positive). These have both been well described and examined in reviews by experts. These reviews acknowledge that the definition of HBV reactivation has been arbitrary. Therefore, we disagree with the application by Koo et al of another term to describe the same condition; this may merely create more confusion for readers. We would like to confirm that our reported series involved consecutive patients treated at one institute in an endemic area where uniform protocol was available. Finally, we would like to clarify again that to prevent HBV reactivation and its associated morbidity and mortality, we suggested two possible approaches. The first was to closely monitor HBV DNA and serum biochemistry during and for at least 6 months after completion of rituximab therapy in anti-HBc–positive patients, particularly those who are also antiHBs negative, and administer antiviral prophylaxis promptly on detection of reactivation. However, it has been commonly observed that this approach has not been universally successful, possibly because of delays in antiviral administration; it may not be cost effective, and it can be difficult to conduct in clinics that lack adequate laboratory support. We thus suggested the alternative of prophylactic antiviral therapy. Koo et al may consider the number of patients studied to have been too few. However, we maintain our study shows that physicians should take into account the cost of intense monitoring and assess whether to consider an alternative approach to a potentially ineffective therapy.