The utility of cytomegalovirus (CMV) quantitative polymerase chain reaction in diagnosis of CMV disease in patients with AIDS

Abstract number: 1134_01_280

Brantsæter A.B., Goplen A.K., Holberg-Petersen M., Jeansson S., Bruun J.N.


HIV infected individuals on no or failing antiretroviral therapy are at risk of CMV disease. Reliable diagnostic tests are important to identify patients at risk, to diagnose manifest disease and to monitor therapy. Studies have shown various capacities of diagnostic tests and have often been limited by lack of autopsy, important because CMV disease is often first diagnosed after death. The objective of this study was to explore the diagnostic utility of CMV quantitative PCR in AIDS patients who died and had a subsequent autopsy.


Patients included in this retrospective study were HIV and CMV seropositive and died during the period 1991–2002 during follow-up at Ullevaal University Hospital, Oslo. A full autopsy including neuropathological examination was required for inclusion. Paraffin-embedded sections were routinely stained with haematoxylin–eosin and immunohistochemistry for CMV was performed in a few cases of doubt to confirm CMV disease. In the case of CMV retinitis, ante-mortem diagnosis was based on typical ophthalmoscopic findings. Otherwise, characteristic histopathological features of cytomegalocytes with inclusions were required for diagnosis of CMV disease. The last one or two stored plasma samples (according to availability) either before diagnosis of CMV alive or before death were analysed by COBAS AMPLICOR CMV Monitor (Roche).


A total of 125 patients were included. 55 patients were diagnosed with CMV disease, of which 39 were first diagnosed at autopsy. 70 patients did not have CMV disease. A total of 85 plasma samples were tested before diagnosis in patients with cytomegalovirus disease and 117 samples were analysed from patients without CMV disease. Median time between the last sample and diagnosis for patients with and without CMV disease were 69 and 74 days, respectively. Geometric mean virus titre in the last sample in patients with viraemia with and without CMV disease were 5894 (median 3140) and 1850 (median 1860) copies/ml, respectively. Results of sensitivity, specificity, positive and negative predictive values of the test are presented in table 1.

Table 1.  Sensitivity, specificity, Positive and negative predictive value for CMV disease by changing viraemia cut-off, time of sampling and number of positive tests.

Quantitative PCR result Sensitivity (95% CI) Specificity (95% CI)Positive predictive value (95% CI)Negative predictive value (95% CI)
Any viraemia0.51 (0,38-0,64)0.87 (0,77-0,93)0.76 (0,60-0,87)0.69 (0,59-0,78)
Viraemia> 2 000   copies/ml0,29 (0,19-0,42)0,96 (0,88-0,99)0,84 (0,62-0,95)0,63 (0,54-0,78)
Vireamia> 10 000   copies/ml0,16 (0,09-0,28)1,00 (0,95-1,00)1,00(0,70-1,00)0,60 (0,51-0,68)
Sample < 30 daysab0,50 (0,29-0,71)0,87(0,53-0,98)0,90 (0,60-0,98)0,44 (0.23-0,67)
One of two tests   positivea0,58 (0,38-0,73)0,84 (0,72-0,91)0,65 (0,46-0,80)0,78 (0,66-0,87)
Two of two tests   positivea0,23 (0,12-0,41)1,00 (0,94-1,00)1,00(0,65-1,00)0,71 (0,60-0,80)
CI = Confidence interval a Any viraemia. b Patients with samples taken within 30 days of either CMV diagnosis Calive or post-mortem or before death.


CMV disease continues to occur in HIV infected individuals even at a time of widespread use of HAART. In this retrospective autopsy based study quantitative PCR had high specificity and positive predictive value in patients with high viral load or recurring viraemia but was limited by poor sensitivity and negative predictive value.

Session Details

Date: 01/08/2007
Time: 00:00-00:00
Session name: XXIst ISTH Congress
Location: Oxford, UK
Presentation type:
Back to top