analysis. No correlation was present between the counts of herpesvirus-specific T cells and bacterial or fungal infections, possibly, as Th2 rather than Th1 T-cell response helps to curb bacterial infections.The major limitation of our study is the relatively low number of patients. Additionally, we selected patients who did not have GVHD on day 56 (as these patients would not be candidates for withdrawal of immunosuppressive drugs or donor lymphocyte infusion (DLI)), and excluded patients who developed second malignancy (as these patients are typically treated by modalities that alter their immunity). We did evaluate the counts of herpesvirus-specific T cells in the six patients excluded owing to second malignancy (mostly post transplant lymphoproliferative disorder (PTLD)), but did not show the data. All patients had a score of 0. Even if these patients were treated with preemptive treatment for relapse (that is, withdrawal of immunosuppression or DLI), they would likely benefit from the treatment, as withdrawal or decrease of immunosuppression or DLI are treatment options for PTLD.In conclusion, we have shown that herpesvirus-specific T cells can indeed be a surrogate marker of the GVL response in AML patients. The ability to predict patients with increased risk of relapse could lead to preemptive therapy, such as withdrawal of immunosuppression or DLI and reduction in relapse rates. However, preemptive therapy can have adverse effects. The ability to clearly distinguish between patients at high versus low risk of relapse is therefore essential. Our assay has shown a high sensitivity, however, a better specificity would be desirable for guiding preemptive treatment. Perhaps, if combining our assay with an assay of high specificity albeit poor sensitivity (for example, a minimal residual disease assay), the best prediction of relapse could be achieved. Future studies including a higher number of patients and performed prospectively are necessary to confirm the clinical utility of these assays.
Conflict of interestThe authors declare no conflict of interest.