RESUMO
BACKGROUND: Cytomegalovirus (CMV) disease occurs in 16% to 20% of low-risk, CMV-positive renal transplant recipients. The cutoffs for quantitative real-time polymerase chain reaction (qPCR) or phosphoprotein (pp65) antigenemia (pp65emia) for starting preemptive therapy have not been well established. METHODS: We measured qPCR and pp65emia weekly from day 7 to day 120 after transplantation, in anti-CMV immunoglobulin Gpositive donor and recipient pairs. Patients and physicians were blinded to the test results. Suspicion of CMV disease led to the order of new tests. In asymptomatic viremic patients, the highest pp65emia and qPCR values were used, whereas we considered the last value before diagnosis in those with CMV disease. RESULTS: We collected a total of 1,481 blood samples from 102 adult patients. Seventeen patients developed CMV disease, 54 presented at least one episode of viremia that cleared spontaneously, and 31 never presented viremia. Five patients developed CMV disease after the end of the study period. The median (95% confidence interval) pp65emia and qPCR values were higher before CMV disease than during asymptomatic viremia (6 [982] vs. 3 [114] cells/10(6) cells; P<0.001 and 3,080 [1,26315,605] vs. 258 [2581,679] copies/mL; P=0.008, respectively). The receiver operating characteristic curve showed that pp65emia 4 cells/10(6) cells or greater showed a sensitivity and specificity to predict CMV disease of 69% and 81%, respectively (area, 0.769; P=0.001), with a positive predictive value of 37% and a negative predictive value of 93%. For qPCR 2,000 copies/mL or higher, the positive predictive value and negative predictive value were 57% and 91%, respectively (receiver operating characteristic area, 0.782; P=0.000). CONCLUSION: With these cutoffs, both methods are appropriate for detecting CMV disease.
