Analysis of Anti-CMV T-Cell Immunity and CMV Reactivation after Allogeneic Hematopoietic Cell Transplantation

Introduction. Positive anti-CMV serology is indicative of a pre-existing immunologic response against CMV. CMV reactivation and disease are significant complications after alloHCT. We studied anti-CMV T-cell immunity and its effects on CMV reactivation and disease relapse after alloHCT using anti-CMV specific T-cell multimers and CMV antigenemia. Identification of factors which contribute to CMV reactivation and disease relapse could lead to novel interventions to decrease the morbidity and mortality of alloHCT.

Methods. This was a retrospective analysis of prospectively gathered clinical and correlative laboratory data. We quantified anti-CMV T cells in recipients before alloHCT and at days 30 and 100 after alloHCT using CMV specific MHC multimers (Dextramers, Immudex, Denmark). Multimers consist of dextran polymers conjugated to optimized numbers of MHC-CMV peptide complexes and fluorochrome molecules which bind with high avidity to CMV specific T-cells. Dextramers were restricted to HLA-A01, -A02, -A03, -A24, -B07, -B08, and -B35, and were informative in 93% of our predominantly White (97%) population. CMV reactivation was determined with CMV antigenemia based on once weekly testing after WBC recovery until day 100 and as clinically indicated at subsequent follow-up visits. CMV reactivation was defined as the presence of CMV antigen in ≥2/50,000 leukocytes on a single occasion or ≥1/100,000 leukocytes twice consecutively ± anti-CMV therapy, or ≥1/100,000 cells on a single occasion + anti-CMV therapy. Absence of CMV reactivation was defined as CMV antigen in 0-1/100,000 cells without anti-CMV therapy. CMV reactivation was treated preemptively with ganciclovir or foscarnet.

Results. 327 consecutive patients received a 10/10 HLA matched unrelated or 6/6 HLA matched related T-cell replete first alloHCT between 2/1/2008 and 12/31/2016, had informative HLA for multimer testing, and had a pre-alloHCT and at least one post-alloHCT multimer test (Table). Recipients (R) and donors (D) who were both negative for CMV by serology (R-/D-) did not reactivate CMV at any time post-alloHCT. Only 6% of R-/D+ reactivated CMV post-alloHCT at a median day+40 indicating a low rate of CMV transmission from donor to recipient. 52% of R+/D- pairs reactivated CMV at a median 40 days post alloHCT compared to 42% of R+/D+ pairs at a median 42 days post alloHCT. Quantitative anti-CMV T cell immunity was analyzed at 1, 3, 5 and 10 cells/µL. Reconstitution of anti-CMV T cell immunity at any level at day+30 post alloHCT was not associated with CMV reactivation between day +31-100; similar results were seen with anti-CMV T cell immunity at day+100 and CMV reactivation between day+101-365. No differences were seen after stratifying by R+/D+ and R+/D-. Correlation of early (day 0 to+100) CMV reactivation after alloHCT with the cumulative incidence of relapse of the underlying malignancy demonstrated lower relapse rates for patients who had any CMV reactivation (left panel) or >25/50,000 cells positive for CMV (right panel) (Figure).

Conclusions. Our findings suggest that early in the course of T cell replete HLA-matched alloHCT, multimer based measurements of anti-CMV T cells do not predict for protective immunity against CMV. The incidence of disease relapse may be lower after CMV reactivation. Further analysis according to conditioning regimen intensity, GvHD incidence, and lymphocyte chimerism to confirm these findings is underway.

Table

View largeDownload slide

Table

View largeDownload slide

Close modal