Cytomegalovirus (CMV) remains a major cause of morbidity after allogeneic hematopoietic cell transplantation (HCT), with recipient serostatus guiding CMV risk stratification and the use of letermovir prophylaxis [1,2,3,4,5]. In routine clinical practice, CMV serostatus is defined by pretransplant CMV‒IgG testing, but interpretation is increasingly challenging in heavily transfused hematology patients, in whom passively transferred antibodies may transiently mimic prior infection [6, 7]. This challenge is particularly evident with highly sensitive CMV‒IgG assays such as Elecsys® (Roche Diagnostics), which frequently detect low-positive results near the analytical cutoff. These findings are commonly interpreted as seropositive, despite uncertainty regarding their clinical relevance. Current CMV management guidelines do not specifically address low-level CMV‒IgG reactivity or indicate a preferred testing platform [8]. Accordingly, this study evaluated whether low-level Elecsys® CMV‒IgG reactivity is reproducible across alternative CMV‒IgG assays and marks an HCT-recipient population at increased risk of post-transplant CMV reactivation.
This single-center retrospective study included all consecutive adult patients (≥18 years) undergoing allogeneic HCT at the University Hospital Basel, Switzerland, between January 2023 and August 2025. Donor and recipient CMV serostatus were assessed as part of routine pretransplant evaluation by Elecsys® (Roche Diagnostics, Rotkreuz, Switzerland). After transplantation, CMV‒DNAemia was monitored weekly in plasma using the cobas®‒CMV assay (Roche Diagnostics) [9], in accordance with current ECIL recommendations [8]. Additional technical details are provided in the Supplementary Methods. The study was approved by the Ethics Committee of Northwestern and Central Switzerland (EKNZ 2025‒00016).
This study included 335 allogeneic HCT-recipients with available pretransplant serum samples (Supplementary Fig. 1). The median age was 63 years (interquartile range, 52‒68), and 136 (40.6%) were female.
To evaluate low-level CMV‒IgG reactivity, 95 Elecsys®-positive sera collected at transplantation were retested using VIDAS® (bioMérieux, Marcy-l’Étoile, France), SERION® (Virion/Serion, Würzburg, Germany), and recomLine (Mikrogen, Neuried, Germany) immunoassays (Supplementary Tables 1 and 2). Among 38 low-positive Elecsys® samples, only 2 (5.3%) were confirmed as positive by any alternative platform, indicating poor reproducibility of low-level Elecsys® reactivity. In the absence of an accepted gold standard for prior CMV exposure, low-positive results were interpreted by orthogonal assay concordance rather than classified as true- or false-positive. Concordance between Elecsys® and other immunoassays increased markedly at CMV‒IgG concentrations >9 U/mL. Receiver operating characteristic analysis using a composite reference standard across the four immunoassays showed a specificity of 64.6% for Elecsys® in the low-positive range, increasing to 100% above 9 U/mL (Supplementary Fig. 2). This threshold was therefore used to distinguish high-positive from low-positive Elecsys® reactivity. Accordingly, 202 allogeneic HCT-recipients (60.3%) were classified as high-positive (>9 U/mL), 71 (21.2%) as low-positive (0.5‒9 U/mL), and 62 (18.5%) as negative (2, 3, 8], and to 37 (52.1%) low-positive HCT-recipients based on clinical judgment (Supplementary Table 3).
Changes in pretransplant CMV‒IgG classification were assessed in 107 allogeneic HCT-recipients with multiple pretransplant serum samples by comparing the first sample obtained within 300 days before transplantation with the sample obtained closest to HCT. At transplantation, 57 recipients were high-positive, 30 low-positive, and 20 negative (Supplementary Table 4). Of 54 initially high-positive recipients, none changed category (0%; 95% CI, 0‒6.6%). Of 19 initially low-positive recipients, 3 (15.8%; 95% CI, 3.4‒39.6%) changed category, with 2 becoming negative and 1 high-positive. Of 34 initially seronegative recipients, 16 (47.1%; 95% CI, 29.8‒64.9%) converted, including 14 to low-positive and 2 to high-positive (Fig. 1a). Compared with initially high-positive recipients, category changes were significantly more frequent in initially low-positive and seronegative recipients (both p
a Sankey diagram showing transitions in Elecsys® CMV‒IgG classification in 107 allogeneic HCT-recipients with multiple pretransplant serum samples available, comparing the first sample obtained within 300 days before transplantation with the sample collected closest to transplantation. Serostatus categories were defined as negative (9 U/mL). Arrow widths are proportional to the number of HCT-recipients transitioning between categories. b, c Cumulative incidence of CMV‒DNAemia through day+100 after HCT in 253 HCT-recipients with serial CMV‒DNAemia results available from transplantation through day +100, stratified by Elecsys® CMV‒IgG serostatus at transplantation and letermovir (LMV) prophylaxis. Estimates were calculated using the Aalen‒Johansen method in a competing-risks framework for any detectable CMV‒DNAemia, including signals below the quantification threshold (b), and CMV‒DNAemia >500 IU/mL (c).
To assess the impact of low-level Elecsys® reactivity on post-transplant CMV reactivation, clinical outcome analyses were performed in 253 HCT-recipients with serial CMV‒DNAemia results available from transplantation through day+100. The cumulative incidence of any CMV‒DNAemia, including detectable signals below the quantification threshold, was highest in high-positive HCT-recipients receiving letermovir prophylaxis at 40.5% (95% CI, 32.6‒48.5). In contrast, incidence was 16.7% (95% CI, 1.4‒31.9) in low-positive HCT-recipients without prophylaxis, 3.7% (95% CI, 0‒11.0) in those receiving prophylaxis, and 5.6% (95% CI, 0‒11.7) in seronegative recipients. All groups with low-positive or negative CMV‒IgG at transplantation showed significantly lower incidence of CMV‒DNAemia than those with high-positive serology (p 1b).
In Fine‒Gray regression analyses, the lowest subdistribution hazard for CMV‒DNAemia was observed in recipients with low-positive Elecsys® results receiving letermovir prophylaxis (sHR, 0.09; 95% CI, 0.01‒0.71; p p p = 0.09), compared with high-positive recipients on letermovir prophylaxis as the reference group. Thus, recipients with low-positive Elecsys® results, irrespective of letermovir prophylaxis, as well as seronegative recipients, had an approximately 60‒90% reduction in the subdistribution hazard of CMV‒DNAemia compared with high-positive recipients. Furthermore, higher Elecsys® CMV‒IgG concentrations at transplantation showed a trend toward increased risk of CMV reactivation (p = 0.054). Myeloablative conditioning was independently associated with higher risk (sHR, 1.86; 95% CI, 1.10‒3.15; p p
When analysis was restricted to CMV‒DNAemia above 500 IU/mL, 13 events occurred among high-positive HCT-recipients receiving letermovir, corresponding to a day+100 cumulative incidence of 8.8% (95% CI, 4.2‒13.4). Only a single case occurred in a low-positive HCT-recipient (D + /R + ) without letermovir prophylaxis (4.3%; 95% CI, 0‒12.9) (Fig. 1c), in whom CMV‒DNAemia was first detected on day+26, peaked at 29 000 IU/mL by day+44, and declined rapidly after initiation of valganciclovir therapy. This temporal pattern and viral kinetics are more consistent with primary CMV infection than with reactivation from latent virus [10].
This study shows that low-positive Elecsys® CMV‒IgG results are often not reproducible on alternative platforms and are not associated with the CMV reactivation risk observed in truly seropositive HCT-recipients. These findings have important implications for CMV risk stratification and clinical management in the bone marrow transplantation setting. Current ECIL-10 guidelines do not recommend a specific CMV‒IgG testing platform, and interpretation of indeterminate pre-HCT serology remains assay- and center-specific [8]. Our findings therefore, suggest that low-positive CMV‒IgG values near the assay cutoff may lead to different CMV serostatus assignments depending on the platform used, consistent with prior reports in which initially Elecsys®-seropositive patients were reclassified as seronegative after confirmatory testing and showed CMV reactivation rates similar to truly seronegative individuals [11, 12]. Although transfusion histories of HCT-recipients were not fully available and no accepted gold standard for prior CMV exposure exists, the clinical outcome data presented here support that some low-positive Elecsys® reactivity may reflect non-specific or passively transferred antibodies rather than true prior CMV infection. In this context, misclassification of seronegative HCT-recipients as seropositive may affect CMV risk stratification, HCT-donor selection, and letermovir allocation. Such misclassification likely arises, at least in part, from assay-dependent differences in assay design and analytical sensitivity near the lower limit of detection, where low-positive Elecsys® reactivity frequently lacked orthogonal confirmation. Accordingly, confirmatory testing with an alternative CMV‒IgG assay or immunoblot may aid clinical interpretation, and non-reproducible low-positive results may reasonably be considered functionally seronegative under routine post-transplant CMV‒DNAemia monitoring. If validated in larger multicenter cohorts, these findings may inform refinement of CMV serostatus assignment strategies and consideration of assay-specific factors in future CMV management guidelines.
