We read with great interest the recently published article entitled "Late cytomegalovirus (CMV) infections after kidney transplantation under the preemptive strategy: Risk factors and clinical aspects" by of G. Ono et al 1 The authors found that these late-onset CMV infections were associated with a higher rate of invasive disease and graft loss than earlier infections occurring in the first 6 months following the transplantation. In a previous article, we found that lateonset CMV diseases were associated with a lower DNAemia peak, a shorter CMV therapy and fewer recurrences than early-onset diseases because of a faster anti-CMV immune response. 2We thank the editors of Transplant Infectious Disease for giving us the opportunity to discuss the differences between the two studies:1. The term "Late-onset CMV disease" is still commonly used indiscriminately to describe two different clinical contexts. A previous editorial suggested-in a very relevant way-using the term "post-prophylaxis CMV diseases" for those occurring after universal prophylaxis discontinuation and "late-onset CMV diseases" for those related to new primary infection and reactivation occurring >4 months after transplantation among patients with preemptive strategy, and those occurring >6 months after universal prophylaxis discontinuation. 3 We strongly believe that these two late conditions are unique entities and partly explain why the outcomes are different in the two studies. "Late-onset" CMV diseases were actually "post-prophylaxis" in our study; and were "late-onset" in Ono et al one. The other consequence is that preemptive strategy may not be related to the occurrence of late CMV episodes in Ono et al study because CMV diseases occurred long after the discontinuation of CMV DNAemia monitoring (median 473 days [317-864]). 2. Secondly, Ono et al have compared early-onset and late-onset CMV infections including CMV DNAemias. Asymptomatic DNAemias have probably been missed in the first group which was no longer monitored for CMV and mechanically led to a relative increase of CMV diseases percentage among patients with late CMV infections (79.1%) vs those with early episodes (41.9%).To overcome this bias in our study, we excluded all asymptomatic events and compared only the first episodes of CMV diseases. 2 3. These two studies did not focus on the same population. Our population was homogeneously composed of D + R-patients.Consequently, CMV diseases were primo-infections and the favorable outcome of late-onset events was a direct result of a more effective anti-CMV immune response, occurring after the immunosuppressive burden decrease. 2 In that scenario, we observed that late-onset diseases had a faster viral clearance, a shorter antiviral treatment, and less relapses than early-onset diseases. In the article of Ono et al, their population included mostly R+ patients (71.4%), with CMV reactivation favored by an enhancement of immunosuppression or concomitant systemic infections. Moreover, the quickness of CMV clearance, the antiviral treatment...