Human Cytomegalovirus (HCMV), a widespread Herpes virus, usually produces asymptomatic infections in immunocompetent hosts. Serious disease can occur in immunocompromised individuals and in congenitally infected newborns. Symptoms in these patients range from a mild disease to life threatening multiorgan system disease. Conventional methods for laboratory diagnosis of CMV infection include serology, virus culture by conventional tube method or rapid shell vial assay and antigen detection. Culture is the "gold standard" but is a relatively insensitive laboratory method and serology results are difficult to interpret especially in immunocompromised patients 123. pp65 antigenemia assay is used as a test for monitoring those at higher risk of developing CMV disease and to initiate pre-emptive therapy 45.

Detection of HCMV DNA in clinical specimens by nucleic acid based amplification methods such as Polymerase chain reaction (PCR) contributes to a rapid and early diagnosis 678. Primer pairs for the detection of the genes coding for the Immediate Early (IE) antigen and Late antigen (LA) were initially used for the detection of HCMV genome in urine and peripheral blood leucocyte specimens 910. Since then, a variety of primer pairs are being used for routine diagnosis of HCMV infection in various patient populations. Sequence variations in the viral genome have been shown to affect the ability of the PCR using different primer sets to detect HCMV DNA 11121314. Little is known about the sequence variations in the regions of HCMV genome under the present study viz: morphological transforming region II (mtr II), UL 83 and glycoprotein O (gO) gene. The PCRs for the aforementioned regions were already standardized in our laboratory using cultures of CMV AD-169 strain (ATCC VR-538). pp65 antigenemia assay is a rapid, reliable and superior to both rapid shell vial assay and conventional test tube culture in the detection of HCMV in the clinical specimens from immunocompromised patients indicating active HCMV disease. 1516. Therefore, the antigenemia assay was considered as "gold standard" in the present study to evaluate the efficacy of the three Polymerase chain reaction tests to detect HCMV genome in the clinical specimens of clinically suspected HCMV disease in immunocompromised patients in Chennai, India.

Of the total 92 specimens from 74 patients tested, the pp 65 antigenemia was present in 48 clinical specimens from 38 patients and these were considered as positive for the "gold standard", definition for active CMV disease. The patients in whom pp65 antigenemia was positive presented a mean of 44.7 positive cells in the antigenemia assay (range 13 – 162 cells). HCMV was isolated from three clinical specimens (one peripheral blood leucocyte and two urine) from three patients, positive for pp65 antigenemia. Of the total 48 clinical specimens positive by the "gold standard", when tested by PCR methods all were positive for mtr II region, 27 for UL -83 gene and 21 for gO gene. This increase in clinical sensitivity by the PCR for mtr II over the UL 83 and gO PCRs were 44% and 56% respectively. The increase in the clinical sensitivity of the PCR for mtr II was statistically significant (P < 0.0001 by Fisher's exact test for two proportions). However, the difference in the clinical specificity between the three PCR tests were not statistically significant (P >0.08, by Fisher's exact test for two proportions). A summary of the results for the clinical specimens is presented in Table 2. All the 45 peripheral blood leucocytes obtained from the controls (seropositive healthy donors) remained negative by pp65 antigenemia assay and all the three PCR tests for HCMV.

Human Cytomegalovirus has long been recognized as a major cause of life- threatening complications in immunosuppressed individuals. There is perceived need for the use of a reliable technique that allows an early detection of the viral activation to help decide on early use of pre-emptive therapy in those at greater risk of the disease. Technique such as virus isolation though most specific cannot be practiced on a regular basis due to lack of its sensitivity and non-availability of human diploid fibroblasts in this part of the world. Quantitative pp 65 antigenemia, used to monitor and detect CMV disease, is well established to have a higher positive predictive value for the disease 3. Since antigenemia is cell based and a low frequency event, a sufficient number of granulocytes are necessary for a reliable result 3. This becomes difficult in Bone marrow transplant or other patients with severe leucopenia. Other difficulties include the necessity for immediate processing of the specimen (within 6 hrs, stored specimens may give erroneous results), difficulty in processing a large number of specimens at a time and subjective component in slide reading, which requires expertise 3. The drawbacks of pp65 antigenemia assay or virus isolation may be overcome by the use of rapid, sensitive and normalized method such as PCR for the detection of HCMV genome in the clinical specimens, which can be applied on a large scale of clinical specimens without any difficulty in a standard laboratory for routine diagnosis of HCMV infection.

Amplification of HCMV genome by PCR is a rapid and sensitive method for detection of HCMV in clinical specimens. The choice of PCR primers for HCMV genome detection in a clinical specimen is crucial since the genome of HCMV is reported to be highly variable 11121314. The primers targeting regions such as the Major immediate early gene exon 4, regions of gene coding late antigen, glycoprotein B (gB) and glycoprotein H (gH) which are widely used have failed to detect the HCMV genome in certain clinical specimens due to primer target mismatch owing to the large sequence variations in HCMV genome. Nucleotide substitutions and even deletions of certain ORFs have been found along the genome of HCMV 17. Studies by Distefano et al., suggested that PCR for gB gene was more reliable than the Major immediate early gene exon 4 or Late antigen gene in detecting HCMV genome in the clinical specimens from congenital and perinatal infections in Argentina 13. Studies by Wirgart et al. suggested that the DNA polymerase gene and gB gene were more conserved and can be used for diagnosis of HCMV infections in different patient populations 14. PCRs for gB and gH genes of HCMV were compared in a Brazilian study on the renal transplant recipients by Aquino and Figueiredo where in a multiplex format of both the genes was suggested for reliable detection of HCMV genome 18. PCRs for MIE gene (IE1), glycoprotein B (gp 58) and structural phosphoprotein (pp 150) were compared with pp 65 antigenemia in heart and lung transplant recipients by Barber et al. where in all the PCR showed a high sensitivity of 100% though gp 58 was associated more with a positive PCR signal than the other two PCRs 19.

There are only a few reports available on the primers targeting the regions of HCMV genome under the present study 202122. Therefore, the efficacy of these three primers were evaluated on different clinical specimens obtained from different high risk immunocompromised patient populations against pp65 antigenemia assay as gold standard. Blood from the healthy seropositive controls did not yield any detectable CMV DNA by any of the three PCRs following amplification. Thus, though the PCRs were considered as highly sensitive methods, they did not detect low – level latent HCMV infection present in the healthy immunocompetent individuals in this study.

pp65 antigenemia assay detected active CMV disease in 38 of 74 patients. The virus isolation had a low sensitivity, positive only in three out of 12 patients with CMV disease as evidenced by a positive antigenemia and these were positive by PCR for all the three regions. We hypothesise that PCRs for all the three regions may become positive only with a high viral load.

The failure of the PCR targeting UL 83 may be due to its lower analytical sensitivity, as it is a uniplex PCR. The PCR for gO gene though a nested PCR, shows a lower clinical sensitivity than the uniplex PCR for UL 83 and this may be attributed to the strain variations in gO gene leading to the primer target mismatch and hence loss of an amplification signal.

Results of our study showed that the PCR for mtr II had 100% sensitivity, 100 % negative predictive value, 87% positive predictive value and 84% specificity. Therefore, it is the most suitable for routine use in Chennai region in India.

The author(s) declare that they have no competing interests.

The corresponding author, H.N. Madhavan is the principal investigator of the study; is involved in the design, supervision, data analyses and writing of the report. P. Sowmya performed all the virological investigations, nucleotide sequencing and analyses of data. All authors were involved in the preparation of this "Research Article".