Glycoprotein B subtyping of cytomegalovirus (CMV) in the vitreous of patients with AIDS and CMV retinitis.

A 550-bp region of the cytomegalovirus (CMV) glycoprotein B (gB) gene was amplified by polymerase chain reaction (PCR) from 141 vitreous specimens of 120 patients with AIDS and CMV retinitis from three different metropolitan centers. The distribution of gB subtypes I, II, III, and IV were 19%, 43%, 12%, and 21%, respectively, based on restriction enzyme digestion patterns of PCR-amplified DNA. Two patients had simultaneous infection with two different gB subtypes. The ratio of gB subtypes was similar among the three geographically distinct patient populations. Two of 14 patients with bilateral vitreous specimens had different viral subtypes in each eye. In addition, different gB subtypes were observed in 1 of 6 patients with serial specimens. The ratio of different gB subtypes in the vitreous of patients with CMV retinitis is similar to that previously reported in the peripheral blood of patients with advanced AIDS.


A 550-bp region of the cytomegalovirus (CMV) glycoprotein B (gB) gene was amplified by polymerase chain reaction (PCR) from 141 vitreous specimens of 120 patients with AIDS and CMV
retinitis from three different metropolitan centers. The distribution of gB subtypes I, II, III, and IV were 19%, 43%, 12%, and 21%, respectively, based on restriction enzyme digestion patterns of PCRamplified DNA. Two patients had simultaneous infection with two different gB subtypes. The ratio of gB subtypes was similar among the three geographically distinct patient populations. Two of 14 patients with bilateral vitreous specimens had different viral subtypes in each eye. In addition, different gB subtypes were observed in 1 of 6 patients with serial specimens. The ratio of different gB subtypes in the vitreous of patients with CMV retinitis is similar to that previously reported in the peripheral blood of patients with advanced AIDS.
Cytomegalovirus (CMV) retinitis is one of the most common coded glycoprotein B (gB) in this process. CMV gB has long been known to be the target of virus-neutralizing antisera [1] opportunistic infections of patients with AIDS, but little is and more recently has been shown to mediate virus entry, cellknown about the pathogenesis of this disease. A number of to-cell spread, and syncytium formation in vitro [2,3]. The recent studies have focused on the possible role of virus-enherpes simplex virus type 1 homolog of CMV gB plays a key role in viral neuroinvasiveness in vivo [4]. Genetic studies of the gB coding region of clinical CMV isolates have led to the identification of four CMV gB geno-for CMV retinitis and found that patients infected with CMV of the second specimen was identical to that of the original specimen. In the 1 remaining eye, we initially detected CMV gB subtype II were more likely to develop retinitis than were those infected with other CMV gB subtypes. Furthermore, 70% gB subtype II in the vitreous, but 8 months later we detected CMV gB subtype IV. of the patients who developed CMV retinitis in that study had a prior CMV gB subtype II viremia.
Our study also included 14 pairs of vitreous specimens from patients with bilateral retinitis. The same gB subtype was iden-To further investigate the possible role of CMV gB subtypes in the pathogenesis of CMV retinitis, we have analyzed the gB tified in both eyes in 12 of these patients. In 1 patient, we detected gB subtype I in the right eye and 2 weeks later gB subtypes of CMV DNA in the eyes of patients with AIDSrelated CMV retinitis. This was accomplished by restriction subtype III in the left eye. In a second patient, we detected gB subtype IV in the right eye and 10 weeks later a mixture of digest analysis of polymerase chain reaction (PCR) -amplified CMV DNA from the vitreous of eyes with retinitis.
gB subtypes II and IV in the left eye.
The MaeIII digest patterns of amplified CMV gB DNA from 8 vitreous specimens (7 eyes) did not match any of the four Methods patterns previously described [6]. The digest pattern of 6 of these specimens (5 eyes) was identical and consisted of two sample from 1 eye) revealed ú98% DNA sequence homology using restriction enzymes RsaI and HinfI.
patterns did not match the four patterns described by Shepp et al.
The predicted amino acid sequences from these specimens are the remaining 2 vitreous specimens were consistent with mixed infections. This was confirmed by restriction digest analysis of at least 10 clones of amplified DNA from each specimen. In Results 1 vitreous specimen, we found a mixture of CMV gB subtypes I and III; in the other, we found a mixture of gB subtypes II We successfully amplified the targeted portion of the CMV gB coding region from 141 of 203 vitreous specimens obtained and IV. We next attempted to determine whether the prevalence of from eyes with CMV retinitis. Restriction enzyme digestion of PCR amplification products from 133 specimens (114 patients) infection with the four gB subtypes differed among patients with different patterns of clinical disease. By X 2 analysis, we produced gel electrophoresis patterns typical for gB subtypes I -IV. A more detailed analysis of the gB genotype of the found no statistically significant difference in the ratio of gB subtypes found in the vitreous of patients with newly diagnosed remaining 8 specimens (7 patients) is described below. Of the 62 clinical specimens that failed to yield amplifiable CMV retinitis (37 patients), reactivated disease (76 patients), and clinically resistant retinitis (15 patients) (P Å .24, x 2 test, DNA, 4 were from eyes with newly diagnosed CMV retinitis, 20 were from eyes with inactive CMV retinitis, and 38 were df Å 6).
To control for contamination of vitreous samples from non-from eyes with reactivated CMV retinitis.
On the basis of restriction enzyme digest analysis of ampli-ocular sources, we attempted to amplify CMV gB DNA from the vitreous of 52 patients with and without AIDS and ocular fied viral DNA, 19% of the vitreous samples from patients with AIDS and CMV retinitis were infected with CMV gB inflammation not caused by CMV. We were unable to amplify CMV DNA from any of these 52 vitreous samples. subtype I, 43% with gB subtype II, 12% with gB subtype III, and 21% with gB subtype IV. A similar ratio of gB subtypes was found in each of the three geographically distinct study Discussion populations (table 1). By x 2 analysis, there was no statistically significant difference in these ratios (P Å .79, x 2 test, df Å 6).
In this study, we have analyzed the gB coding region of CMV in the vitreous of patients with AIDS to determine the Our study included serial vitreous specimens from 6 eyes obtained 5 -9 months apart. In 5 of these eyes, gB subtyping viral subtypes responsible for CMV retinitis. CMV gB subtype / 9d52$$oc31 09-01-98 19:36:50 jinfa UC: J Infect Overall, gB subtype testing of vitreous specimens as we have done in this study was unable to provide additional prog-The distribution of CMV gB subtypes in the eyes of our study patients, while different from that reported in the blood nostic or predictive information about which patients may have resistant or recurrent CMV retinitis. Furthermore, using historic of patients with solid organ transplants [5,7,8], closely resembles the distribution of gB subtypes in the blood of AIDS controls, we were unable to find statistically significant differences between the distribution of gB subtypes in the vitreous patients at risk for CMV retinitis [6,9]. Thus, it is unlikely that infection with any one specific CMV gB subtype, such as and blood of AIDS patients. Analysis of other regions of the CMV genome, including other regions of the gB coding region, gB subtype II, confers increased risk of development of retinitis. This conclusion appears to conflict with the data of Shepp may prove valuable in this regard. et al. [6], who, in a smaller study, found that viremia with CMV gB subtype II was associated with a higher risk of CMV This study sought to determine whether high-dose acyclovir improves posttransplant survival in cytomegalovirus (CMV)-seropositive patients when ganciclovir is given for prophylaxis or as early therapy. Three groups were studied: Group 1 (n Å 112) received ganciclovir from engraftment without prior acyclovir treatment, group 2 (n Å 114) was given ganciclovir for CMV pp65 antigenemia without prior acyclovir, and group 3 (n Å 133) received ganciclovir at engraftment with prior intravenous acyclovir (500 mg/m 2 every 8 h) from day 5 before transplant until engraftment. In a multivariable Cox model, there was no significant difference in the adjusted risk of transplant survival between the groups during the first 2 years after transplant (relative risk for mortality: group 1, 1.0; group 2, 0.75 (95% confidence interval [CI], 0.52 -1.1); group 3, 1.04 (95% CI, 0.74 -1.47). The incidence of CMV disease and CMV-related mortality was not significantly different between the groups. Thus, high-dose acyclovir does not appear to improve survival when ganciclovir is given either at engraftment or for CMV pp65 antigenemia.
Two studies have shown improved survival in patients given sible for the observed survival benefit [2,6]. Other reasons for the controversy include the availability of more potent anti-CMV high-dose acyclovir (500 mg/m 2 every 8 h) intravenously (iv) from day 5 before transplant (05) until day 30 after transplant compounds, such as ganciclovir and foscarnet, the high cost of the proposed acyclovir regimens, especially when added to alone or with subsequent oral acyclovir (800 mg 4 times/day until day 210) [1-3]. Despite these positive results, the issue conflicting results on the efficacy of high-dose acyclovir in human immunodeficiency virus-infected patients [7-10], results of acyclovir for prevention of cytomegalovirus (CMV) disease remains controversial among stem cell transplant centers [4,5].
in solid organ transplant recipients that show superior efficacy of ganciclovir compared to acyclovir [11], and the lack of effi-The major reason for this controversy is a poor understanding of the mechanisms of the observed survival benefit. The differ-cacy of high-dose acyclovir for prevention of CMV in autologous marrow transplant recipients [12]. ences in survival in these studies could not clearly be attributed to a reduction of CMV-associated mortality [1][2][3]. It has been It is unclear whether high-dose acyclovir has additional benefit if ganciclovir is given at engraftment or for early treatment hypothesized that early suppression of CMV infection or other herpesviruses, such as human herpesvirus 6, during the first based on CMV pp65 antigenemia or polymerase chain reaction (PCR) detection of CMV DNA, the two most common strate-month after transplant by iv high-dose acyclovir may be respongies at this time. Because there are no published reports or ongoing controlled studies addressing the issue and because the cost of the proposed acyclovir regimen is substantial, we