Determining the Appropriate PCR Trigger for Pre-Emptive Anti-CMV Therapy in Allogeneic Stem Cell Transplant Recipients.

Background: CMV-related complications are a major cause of morbidity and mortality post allogeneic stem cell transplantation (ASCT). Where a pre-emptive strategy is used, early detection is crucial as this will minimise the risk of CMV infection progressing to CMV disease. As a result, PCR techniques are increasingly replacing pp65 antigen detection as a means of monitoring CMV viral load in at-risk patients. Initiating treatment at too low a cut-off may result in over-treatment since very low level viraemia may not necessarily represent live or replicating virus. However, using too high a cut-off might lead to unacceptable treatment delay. As a result of these uncertainties the optimal pre-emptive strategy remains unknown. We therefore sought to determine whether, using quantitative real time PCR, the manufacturer’s detection cut-off or the validated lower quantitation limit (LQL) was the more appropriate treatment trigger.

Methods: CMV surveillance was done by weekly (twice weekly for positive results) quantitative real time PCR (Artus RealArt™ CMV LC PCR Kit, Hamburg, Germany) until immunosuppression was stopped. The manufacturer’s detection cut-off was 6.5 × 10² copies/ml and the validated LQL was 1 × 10⁴ copies/ml. Treatment was initiated following two consecutive results above the LQL. Below the LQL, treatment was only started if symptoms of CMV infection (e.g. fever, cytopenia) were also present.

Results: Between 01/04 and 12/05 fourty four patients, median age 43y (range 18–69), underwent ASCT from sibling (n=27) or unrelated (n=17) donors for AML/MDS (n=22), ALL (n=8), CML (n=5), NHL (n=5), myeloma (n=3) and osteogenesis imperfecta (n=1). The conditioning regimen was myeloablative (Cy-TBI, Bu-Cy) in 21/44 and reduced intensity (Flu-Mel, Flu-Cy) in 23/44. Campath-1H was used in 16/44. The recipient/donor CMV serostatus was R+/D+ (n=11), R+/D− (n=9), R−/D+ (n=4) and R−/D− (n=20). With a median follow up of 10 months (range <1–26), 12/24 (50%) patients at high risk of CMV reactivation (6 of 11 R+/D+, 6 of 9 R+/D−, 0 of 4 R−/D+) had CMV detected compared with 1/20 (5%) low risk R−/D− patients. CMV detection occurred at a median of 21d (range 3–128) post ASCT. 12/13 patients with detectable CMV had initial levels below the LQL. All required treatment, 5 because of co-existent symptoms of CMV infection despite consecutive readings below the LQL and 7 because the second reading was above the LQL. Conditioning therapy, donor type and use of CAMPATH-1H did not appear to influence risk of CMV detection. There was no CMV related death in this cohort.

Conclusion: We conclude that, using the above kit, the detection cut-off of 6.5 × 10² copies/ml rather than the manufacturer’s validated LQL of 1 × 10⁴/ml is the more appropriate trigger for initiating pre-emptive anti-CMV therapy in patients undergoing ASCT. Since manufacturers do not consider quantification below the LQL to be reliable, we recommend that where the detection cut-off is significantly lower than the LQL, transplant centres seek to determine their own appropriate treatment trigger.