Functional control of CMV reactivation is profoundly influenced by CMV serostatus after nonmyeloablative hematopoietic cell transplantation following a TLI-ATG preparative regimen

Joanna M. Schaenman1,2, Marcy L. Vana2, Chanu Rhee2, Jonathan Wong1, Shelly Navato2, Laura Johnston2, Ruby M. Wong2, Dora Y. Ho1, and Janice M. Brown1,2 Divisions of Infectious Diseases1 and Blood and Marrow Transplantation2, Stanford University Medical Center, Stanford, California 94305-5623, USA.

A substantial body of data supports the importance of recapitulation of effective immune function for control of cytomegalovirus (CMV) following hematopoietic cell transplantation (HCT). However, many questions remain regarding immune reconstitution of functional viral control after nonmyeloablative (NMA) HCT, especially with preparative regimens containing total lymphoid irradiation and antithymocyte globulin (TLI-ATG). We analyzed 197 patients who underwent NMA HCT at Stanford University Hospital between 2001 and 2007; 126 patients had either donor or recipient seropositivity for CMV. TLI was administered 11 days prior to transplant, and ATG on D-11 to -7. Nine patients were eliminated from analysis for either participation in a CMV prophylaxis trial or for insufficient data. HCT recipients were screened weekly for CMV reactivation using the Amplicor CMV test (Roche Molecular Diagnostics) on peripheral blood for the first 100 days after HCT. Any positive result led to preemptive treatment with intravenous ganciclovir or valganciclovir. Statistical analysis was performed using SAS Enterprise Guide. There were no significant differences between groups with respect to sex, age, race, underlying hematologic disease, or donor type. 35% had acute leukemia as their underlying diagnosis, 18% chronic leukemia, 32% lymphoma, and 15% myelodysplastic syndrome. 63 patients received HCT from siblings including one partially matched related donor (53%), 6 patients received haplo-identical transplants (5%), and 48 patients had unrelated donors (41%). Data regarding reactivation by serogroup

SerogroupNo. patientsPatients with CMV reactivation (%)Median days to first reactivationNo. patients with multiple reactivations (%)*Average no. reactivations per patientMax. no. reactivations
* Patients included if there was less than one month of missing data: D+/R− 15 patients, D+/R+ 55 patients, D−/R+ 35 patients evaluated. 
D+/R− 15 2 (13.3) 56 0 (0) 0.1 
D+/R+ 57 27 (47.4) 17 5 (9.1) 0.6 
D−/R+ 45 29 (64.4) 12 5 (14.3) 0.9 
SerogroupNo. patientsPatients with CMV reactivation (%)Median days to first reactivationNo. patients with multiple reactivations (%)*Average no. reactivations per patientMax. no. reactivations
* Patients included if there was less than one month of missing data: D+/R− 15 patients, D+/R+ 55 patients, D−/R+ 35 patients evaluated. 
D+/R− 15 2 (13.3) 56 0 (0) 0.1 
D+/R+ 57 27 (47.4) 17 5 (9.1) 0.6 
D−/R+ 45 29 (64.4) 12 5 (14.3) 0.9 

Analysis by logistic regression with correction for age, diagnosis, donor type, and steroid use was statistically significant (p=0.005) for the comparison of all three serogroups. Other covariates reaching statistical significance were diagnosis (increased reactivation with acute leukemia and lymphoma) and donor type (OR 2.7 for unrelated versus related donors (95% confidence interval 1.2–6.1, p=.016). There was no difference in the degree of chimerism in whole blood on day 28 with a mean of 80% for the D+/R+ and 79% for the D−/R+ patients (t test, pooled p-value 0.90). The incidence and severity of acute graft versus host disease (GVHD) (Grade > II) were not statistically different between groups, with 0 D+/R− patients, 2 D+/R+ patients, and 4 D−/R+ patients with acute GVHD. Sustained CMV viral load was significantly greater in the D−/R+ group based on analysis by a mixed effect model with correction for the covariates listed above (p=.0006 for serogoup*time). None of the other covariates analyzed had a statistically significant effect in this model. CMV serostatus of both donor and recipient had a significant effect on CMV reactivation over the first 100 days after NMA HCT following a TLI-ATG preparative regimen. The relative paucity of reactivation and the longer time to first reactivation in the D+/R− group is predicted by previously published observations. Donor CMV serology did not influence the timing of CMV reactivation in the D+/R+ compared with the D−/R+ group, however, donor seropositivity did play an important role in the reconstitution of effective antiviral immune function as manifested by an increased burden of detectable CMV in the D−/R+ as compared with the D+/R+ group. These observations have led us to an ongoing exploration of the kinetics of the antiviral immune response in an animal model in an attempt to refine our understanding of the interaction between the magnitude and timing of CMV exposure and the reconstitution of functional immunologic control after HCT.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author

Sign in via your Institution