We conducted this retrospective study at Bordeaux University Hospital (France). KTRs who received a deceased or living donor kidney between 1 September 2016 and 31 December 2019 were included if they were over 18 years old and if their CMV status was either D+R- or R+.

Induction therapy consisted of thymoglobulin for HLA-sensitized KTRs, and basiliximab for the others. Maintenance treatment included tacrolimus, targeting through level target of 8–10 ng/mL during the first year, followed by 6–10 ng/mL, along with mycophenolic acid (720 mg bid). Steroids were rapidly reduced to 5 mg/day and weaned in non-HLA-sensitized KTRs during the first month post-transplantation. Everolimus was used for a small number of KTRs with a through level target of 5–8 ng/mL.

All KTRs received universal prophylaxis with valganciclovir, aiming for 6 months in D+R- KTRs or 3 months in R+ KTRs. Valganciclovir dosage adjustments were made using the Cockcroft-Gault formula.

KTRs were excluded if they did not take antiviral prophylaxis for at least 6 weeks, if they experienced death, graft loss, or were lost-of follow-up before month 3, and if monitoring of the γδ T lymphocyte subset was not performed at the end of the antiviral prophylaxis. Notably, γδ T lymphocyte measurement at the end of the prophylaxis was part of the routine monitoring of KTRs during this period.

All clinical and biological variables were collected from the R@N database (with final approval from the French Data Protection Authority [CNIL], number 135715). All participants gave written informed consent. The study was performed in accordance with the ethical standards as laid down in the Declaration of Helsinki, and was approved by the Institutional Review Board of the Bordeaux University Hospital.

The endpoints were:

CMV disease was defined as “CMV syndrome” or “probable or proven end-organ CMV disease” using standardized criteria from international guidelines [26].

CMV syndrome was defined by the detection of a positive CMV PCR, with at least 2 additional criteria among the following: fever, malaise or fatigue, leukopenia or neutropenia, thrombopenia or elevation of hepatic aminotransferase.

Proven CMV end-organ disease was defined as the presence of appropriate clinical symptoms together with documentation of CMV in tissue from the relevant organ by immunohistochemistry.

Probable CMV end-organ disease was defined as the presence of appropriate clinical symptoms together with documentation of high viral DNA levels in tissue from the relevant organ by quantitative nucleic acid testing.

The onset of CMV disease was marked by the first detection of CMV DNAemia with CMV symptoms. The duration of CMV disease was the time from the first positive CMV DNAemia until symptom resolution and viral eradication following at least 2 weeks of treatment. The treatment duration was defined as the period during which KTRs received antiviral therapy for CMV disease. Recurrent disease referred to a new episode in KTRs who had previously achieved negative CMV DNAemia following treatment.

Various CMV quantitative nucleic acid testing (QNAT) methods were used throughout the study. Starting in September 2016, QNAT was performed with the LightMix^® Human Cytomegalovirus Kit (TIB MOLBIOL GmbH, Berlin, Germany), with detection and quantification thresholds of 250 and 1000 IU/mL, respectively. From April 2019 onward, the CMV R-GENE^® Kit (Biomerieux, France) was used, with thresholds of 150 and 200 IU/mL. All QNAT assays were conducted in the Department of Virology at Bordeaux University Hospital, adhering strictly to Quality Control for Molecular Diagnostics (QCMD, Glasgow, Scotland) standards since 2004. A CMV QNAT result below the quantification limit was considered negative.

Lymphocyte and Vδ2^neg γδ T lymphocyte counts were analyzed at the end of universal valganciclovir prophylaxis (±1 month). Vδ2^neg γδ T lymphocyte counts were determined by flow cytometry in the Department of Immunology and Immunogenetics at Bordeaux University Hospital, as previously described [19]. To identify the Vδ2^neg γδ T lymphocyte subset and their TEMRA phenotype, we used a panel containing antibodies targeting CD3, γδ TCR, Vδ2 TCR, CD27, and CD45RA (Beckman Coulter, Marseille, France). The Vδ2^neg γδ T lymphocyte subset is rare in CMV-naïve subjects [24]. Results were reported as “not interpretable” (NI) when fewer than 300 events were detected in the Vδ2^neg γδ T lymphocyte gate (Supplementary Figure S1). For clarity, TEMRA Vδ2^neg γδ T cells are referred to as TEMRA γδ T cells throughout this report.

When comparing the incidence of CMV disease across different medication regimens or rejection episodes, only events occurring before the CMV disease onset were included in the “CMV disease” group. All rejection episodes were biopsy-proven. Preformed donor-specific antibodies (DSA) were defined as those present on the day of transplantation or earlier. Post-transplant estimated glomerular filtration rate (eGFR) was defined as the highest eGFR recorded during the prophylaxis period.

KTRs characteristics are presented as medians and interquartile ranges (IQR) for quantitative variables and as percentages for qualitative variables. Fisher’s exact test or McNemar’s test was used to compare qualitative variables, while Student’s t-test or the Mann–Whitney test was applied to quantitative variables. A p-value <0.05 was considered statistically significant. The relationship between Vδ2^neg γδ T lymphocyte counts was assessed using Spearman’s correlation (rho). Receiver operating characteristic (ROC) curve analysis was conducted to evaluate the performance of lymphocyte counts and TEMRA Vδ2^neg γδ T lymphocyte counts in predicting protection against CMV disease. The probability of CMV disease-free survival, based on lymphocyte levels, was estimated using the Kaplan-Meier method, and the log-rank test was used to compare hazards of CMV disease. Univariate Cox regression analysis was initially applied to identify variables associated with CMV disease. No continuous variable deviated from the assumption of linearity. Covariates with p-values <0.25 in univariate analysis were included in multivariate Cox regression analysis, and variables with p-values <0.05 were retained. Results are presented as hazard ratios (HR) with 95% confidence intervals (95% CI). All analyses were performed using RStudio (version 1.1.423; RStudio Inc., Boston, MA, United States) and Prism (version 10.0.2; GraphPad Software, Boston, MA, United States).

Between September 2016 and December 2019, 606 kidney transplants were performed at Bordeaux University Hospital. Based on the inclusion and exclusion criteria, 344 KTRs were eligible for inclusion in the study (Figure 1).

Table 1 outlines the baseline characteristics of these patients. During the first 2 years post-transplantation, 43 out of 344 KTRs (12.5%) developed CMV disease, with a median onset of 79 days (IQR: 44.0–122.5 days) after discontinuing prophylaxis. Among these, 9 KTRs (20.9%) experienced CMV viral syndrome, and 34 KTRs (79.1%) developed CMV tissue-invasive disease. The median peak CMV DNAemia was 50,320 IU/mL (IQR: 12,432–281,010 IU/mL). The median disease duration was 29.5 days (IQR: 21.5–43 days), and the median treatment duration was 44 days (IQR: 24–55.5 days). CMV recurrence occurred in 6 of the 43 patients (13.6%).

CMV disease occurred in 31.7% (26/82) of D+R- KTRs and 6.4% (17/262) of R+ patients (p < 0.01). Conversely, CMV disease occurred in 4.6% (3/65) of patients treated with mTOR inhibitors and 14.3% (40/279) of patients not treated with mTOR inhibitors (p = 0.03). Interestingly, no significant differences were observed regarding the use of thymoglobulin or the number of treated acute rejection episodes between the groups.

CMV disease characteristics in D+R- and R+ subgroups are detailed in Supplementary Table S1.

We did not observe any episode of CMV disease before the γδ T lymphocyte measurement at the end of the prophylaxis.

Table 2A describes immune profiles of the “CMV disease” and “No CMV disease” KTRs. Vδ2^neg γδ T cells count was higher in the “No CMV disease” group than in the “CMV disease” group (18.4 ± 25.7/mm³ versus 6.0 ± 7.9/mm³; p < 0.01). TEMRA γδ T lymphocytes count was also higher in the “No CMV disease” group (23 ± 26.8/mm³ versus 4.6 ± 6.7/mm³; p < 0.01). Immune profiles in the D+R- and R+ subgroups are depicted in the Tables 2B, C.

It is worth noting that a significant number of immunophenotyping assays did not yield interpretable TEMRA γδ T lymphocyte counts (161/344, 46.8%). These results, labeled as NI (not interpretable), were evenly distributed between the “CMV disease” and “No CMV disease” groups. The proportion of NI patients was similar between those who received thymoglobulin and those who did not (78/183 vs. 82/161, p = 0.13).

Patients with NI results had lower total lymphocytes counts and lower Vδ2^neg γδ T lymphocyte counts than patients with interpretable results, respectively 0.82 G/L vs. 1.01 G/L (p < 0.01) and 4.61/mm³ vs. 27.4/mm³ (p < 0.01) (Supplementary Table S2).

ROC curve analyses for total lymphocyte and Vδ2^neg γδ T lymphocyte counts regarding CMV disease occurrence showed low AUCs of 0.63 and 0.70, respectively (Figures 2A,C). Given the low AUC, we assessed CMV disease-free survival by comparing patients with values above or below the median (lymphocyte count: 761/mm³; Vδ2^neg γδ T lymphocyte count: 7.95/mm³). The probability of CMV disease-free survival was similar between the “high lymphocyte” and “low lymphocyte” groups (p = 0.11) (Figure 2B). However, KTRs with a Vδ2^neg γδ T lymphocyte count >7.95/mm³ had a higher probability of CMV disease-free survival (p < 0.01) (Figure 2D).

The ROC curve for TEMRA γδ T lymphocyte count (excluding NI KTRs) yielded an AUC of 0.79 and defined an optimal threshold of 4.65/mm³ (sensitivity 79.4%, specificity 78.9%) (Figure 3A). Of the 344 KTRs, 135 (39.2%) had a TEMRA γδ T lymphocyte count >4.65/mm³. KTRs with a count >4.65/mm³ had a higher probability of CMV disease-free survival compared to those with a count classified as NI or ≤4.65/mm³ (p < 0.01) (Figure 3B).

We further analyzed the 161 NI KTRs and found that their Vδ2^neg γδ T lymphocyte counts were very low, similar to those of KTRs with TEMRA γδ T lymphocyte counts ≤4.65/mm³, and much lower than those with counts >4.65/mm³ [median: 3.2/mm³ (IQR: 0.1–7.25), 2.9/mm³ (IQR: 0.2–8.15), 24.4/mm³ (IQR: 16.15–40.58), respectively] (Figure 3C). Thus, we grouped KTRs with TEMRA γδ T lymphocyte counts classified as NI and ≤4.65/mm³. KTRs with TEMRA γδ T lymphocyte counts >4.65/mm³ had significantly higher CMV disease-free survival rates than those in the combined NI and ≤4.65/mm³ group (p < 0.01) (Figure 3D).

We conducted a univariate analysis to identify factors associated with CMV disease (Table 3). The following factors were included in the multivariate analysis: R+ serostatus [HR 0.16 (95% CI 0.09–0.29); p < 0.01], total lymphocyte count >761/mm³ [HR 0.56 (95% CI 0.30–1.05); p = 0.07], Vδ2^neg γδ T lymphocyte count >7.95/mm³ [HR 0.34 (95% CI 0.17–0.68); p < 0.01], TEMRA γδ T lymphocyte count >4.65/mm³ [HR 0.18 (95% CI 0.07–0.46); p < 0.01], and the use of mTOR inhibitors [HR 0.27 (95% CI 0.06–1.11); p = 0.07]. In the multivariate analysis, only R+ serostatus [HR 0.23 (IQR 0.11–0.45); p < 0.01] remained independently associated with CMV disease. The TEMRA γδ T lymphocyte count was no longer significantly associated with CMV disease [HR 0.39 (95% CI 0.14–1.09); p = 0.07] (Table 4).

Total lymphocyte counts and Vδ2^neg γδ T lymphocyte counts were significantly lower in D+R- patients compared to R+ patients (0.74 ± 0.46 G/L vs. 0.95 ± 0.59 G/L, p < 0.01; Vδ2^neg : 9.9 ± 6.15 vs. 21.72 ± 26.1; p < 0.01 (Table 2D).

TEMRA : 13.84 ± 19.72 vs. 23.94 ± 26.54; p < 0.01 (Figure 4A). Among the 135 patients with TEMRA γδ T lymphocyte counts >4.65/mm³, only 4 were D+R-, the majority being R+ (n = 131) (p < 0.01). Finally, the number of interpretable results was lower in D+R- patients compared to R+ patients (22 versus 161) (Figure 4B).

Based on these findings, we evaluated the predictive value of a TEMRA γδ T lymphocyte count >4.65/mm³ for protection against CMV disease in the subgroup of R+ KTR, as detailed in Table 5.

Of the 262 R+ KTRs (including NI KTRs), 131 (50%) had a TEMRA γδ T lymphocyte count >4.65/mm³. The sensitivity of this test was low (52.2%), but specificity was high (82.3%). The positive predictive value (i.e., protection against CMV disease in KTRs with a TEMRA γδ T lymphocyte count >4.65/mm³) was 97.7%, while the negative predictive value was only 10.7%. R+ KTRs with a TEMRA γδ T lymphocyte count >4.65/mm³ had a significantly higher probability of CMV disease-free survival than those with counts classified as NI or ≤4.65/mm³ (p < 0.01) (Figure 4C). The probability of CMV disease-free survival was also higher in R+ KTRs with TEMRA γδ T lymphocytes count > 4.65/mm³ than in the group gathering R+ KTRs with TEMRA γδ T lymphocytes count “NI” and ≤4.65/mm³ (p = 0.02) (Figure 4D).

Univariate analysis identified total lymphocyte counts >761/mm³ [HR 0.35 (95% CI: 0.12–0.99); p = 0.05], Vδ2^neg γδ T lymphocyte counts >7.95/mm³ [HR 0.31 (95% CI: 0.11–0.84); p = 0.02] and TEMRA γδ T lymphocyte counts >4.65/mm³ [HR 0.28 (95% CI: 0.09–0.87); p = 0.03] as factors associated with CMV disease (Table 6). In the multivariate analysis of R+ KTRs, only a TEMRA γδ T lymphocyte count >4.65/mm³ remained independently associated with protection against CMV disease [HR 0.27 (95% CI: 0.09–0.85); p = 0.03] (Table 7).