Optimal recovery of cytomegalovirus from urine as a function of specimen preparation

Cytomegalovirus (CMV) is a significant pathogen among immunocompromised patients. We compared supernatant and sediment fractions of centrifuged urine for the optimal recovery of CMV by shell vial culture and polymerase chain reaction (PCR). Of 336 urine specimens, 31 (9.23%) were positive by shell vial culture; of these 29 (93.5%) were identified using the sediment fraction and 17 (54.8%) using the supernatant fraction (p = 0.001, chi2). Of the 29 positive sediment fraction specimens, 24 (82.8%) were identified as CMV positive at 24 h and 5 (17.2%) were identified as positive at 48 h. Two (0.064%) of the total 31 positive specimens were lost to microbial contamination in the sediment inoculated cultures. Of the 17 supernatant fraction specimens, 9 (53.9%) were identified as CMV positive at 24 h and 8 (47.1%) were identified as positive at 48 h. Fourteen (45.2%) of the total 31 positive specimens were lost to either toxicity or microbial contamination in the sediment-inoculated cultures. Thirty-four CMV culture-positive specimens were tested by PCR; 5 of these specimens (14.7%) were PCR negative for both sediment and supernatant fractions; 26 (76.5%) were found to be positive using the sediment fraction and negative using the supernatant; 3 (8.8%) were PCR positive for both the sediment and the supernatant. None of the 34 was identified as positive using the supernatant fraction only (p = 0.001, chi2). These findings demonstrate that the method of specimen preparation can significantly affect the outcome of diagnostic testing for CMV from urine specimens.