Appropriate Threshold of Quantitative Cytomegalovirus DNA Polymerase Chain Reaction for Preemptive Treatment in Stem Cell Transplant Recipients

Introduction

Cytomegalovirus (CMV) remains a crucial virus in complications after allogeneic stem cell transplant (SCT). A preemptive treatment has been used to prevent the progression to CMV disease; however, no consensus has been reached on the CMV-DNA viral load threshold that should trigger preemptive treatment. We monitored both CMV-DNA and antigenemia and started preemptive treatment based on antigenemia, that enabled us to evaluate the clinical course of patients with positive CMV-DNA without antiviral treatment.

Methods

Adult patients who received allogeneic SCT at our institutes from 2022 to 2023 were retrospectively reviewed. CMV antigenemia and quantitative CMV-DNA polymerase chain reaction (PCR) assays were simultaneously evaluated from 2022. The study included 109 patients with available PCR and antigenemia data. High risk of CMV infection was defined as in-vivo T cell depletion, cord blood transplantation, or prednisolone use. All patients received CMV antigenemia-guided preemptive treatment for CMV infection defined as 20 positive cells/2 slides for low-risk and 3 positive cells/2 slides for high-risk patients. Prophylactic letermovir was administered until 100 days after SCT in patients scored 4 points or more in our scoring system (+2 for D+R+, +3 for D-R+, +1 for lymphoma, and +1 for unrelated or haploidentical donor).

Results

This study simultaneously evaluated 2,508 samples. The median CMV-DNA load was 0 (range: 0–260,000) IU/mL and median CMV antigenemia level was 0 (range: 0–959) cells per 2 slides. CMV-DNA and antigenemia level revealed a moderate linear correlation (r=0.63). The antigenemia levels of 0, 1–2, 3–4, 5–9, 10–19, and ≥20 positive cells per 2 slides corresponded to median CMV-DNA loads of 0, 210, 340, 495, 1,400, and 8,000 IU/mL, respectively. The cumulative incidence for CMV infection based on CMV antigenemia and CMV disease at day 200 after transplant was 36.7% and 3.9%, respectively. CMV-PCR became positive at any level before CMV antigenemia reached the threshold in 59 patients. Of the 57 evaluated patients, CMV antigenemia subsequently reached the threshold value in 34 (60%) patients, whereas the other 23 (40%) patients experienced spontaneous clearance of CMV viremia without positive CMV antigenemia. Spontaneous CMV DNA clearance was significantly associated with the absence of steroid use for GVHD. Next, we simulated the clinical course of these patients according to the various CMV-PCR thresholds for preemptive treatment between 50 and 1,000 IU/ml. In high-risk patients without letermovir (n=34), increasing the threshold from 50 IU/ml to 150 IU/ml increased the proportion of patients who did not require antiviral agents only by 3% but delayed the start of preemptive treatment in 23%. In high-risk patients with letermovir (n=28), increasing the threshold from 50 IU/ml to 150 IU/ml increased the proportion of patients who did not require antiviral agents by 7% but further increase of the threshold resulted in more patients with negative PCR but positive antigenemia. In low-risk patients without letermovir (n=44), increasing the threshold from 50 IU/ml to 500 or 750 IU/ml increased the proportion of patients who did not require antiviral agents by 18% but further increasement of the threshold resulted in the delay of preemptive treatment in 5% of the patients. There were only 9 low-risk patients with letermovir, but a threshold similar to that for low-risk patients without letermovir seemed appropriate.

Conclusion

For high-risk patients with or without letermovir, a threshold of 50 IU/ml and 150 IU/ml, respectively, may be recommended, while a threshold of 500-750 IU/ml may be proposed for low-risk patients, to decrease the overuse of antiviral agents with minimal delay in preemptive treatment.