Introduction: Prolonged ganciclovir (GCV) or valganciclovir (VGCV) prophylaxis (12 months) is recommended after lung transplantation (LTx). Intermittent GCV prophylaxis has been adopted in many developing countries to overcome the high cost of VCGV, and the difficulties of daily intravenous GCV. Due to the lack of a robust biomarker of CMV-immunity by the end of prophylaxis, viral surveillance followed by pre-emptive therapy is necessary to early detection and treatment of CMV infection, and to avoid the development of late CMV disease. Although several commercial assays have been tested, the cut-off of CMV viral load to introduction of CMV treatment remains elusive.
Methods:Prospective study in 38 consecutive LTx recipients to establish the cut-off CMV DNAemia as detected by the COBASÒAmpliprep/ COBASÒTaqmanÒ CMV test. GCV prophylaxis was given 3 times a week for 3 months. Prophylaxis was extended up to 6 months in D+/R- patients and those developing acute rejection in the first 3 months. CMV surveillance was done with antigenemia (AG) at recommended intervals and GCV 5mg bid was introduced if AG ≥10 cells. Either positive test occurring during prophylaxis was considered as prophylaxis failure. Kappa index was used to evaluate the agreement between tests. The cut-off of CMV DNAemia was determined by ROC curve. Results. AG was positive in 20 patients (52.6%) and the COBAS in 34 (89.5%). DNAemia detected by COBAS anteceded AG in 13 of the 20 episodes (65%). In the comparison between tests, 465 of the 643 samples (72.3%) showed concordant. Kappa index was 23.6%. COBAS tested positive in a significantly larger number of samples than AG (p<0.0001). The cut-off of COBAS corresponding to AG≥10 cells was 745.5 IU/mL (Sensitivity 100%; specificity 95.9%). Prophylaxis failure occurred in 18 patients (46.2%), being 16 identified by COBAS (88.8%) and 11 by AG (61.1%). Seven pts (18.4%) developed CMV disease: 3 CMV-GI, 3 CMV-IP, and 1 disseminated disease (lung and colon), at a median of 180 (85-252) days. Five patients died during follow-up, three due to acute rejection (2 associated with pulmonary aspergillosis and 1 with bacterial pneumonia), one due to necrotizing pancreatitis and 1 with chronic lung allograph dysfunction. There was no difference on survival among CMV-pacientes and non-CMV LTx recipients (p=0.61).
Conclusion: CMV surveillance by COBAS was more sensitive than the AG test to detect prophylaxis failures in intermittent regimens. A viral load of 745.5 IU/mL in the COBAS platform should prompt pre-emptive GCV therapy. Despite prophylaxis failures, the incidence of CMV disease was low. The presence of CMV immunity may explain these findings.