Incidence of valganciclovir‐related leukopenia and neutropenia in solid organ transplant recipients at high risk of cytomegalovirus disease

Valganciclovir (VGCV) prophylaxis is associated with an increased risk of hematologic side effects. We analyzed the impact of VGCV prophylaxis on leukopenia and neutropenia rates and explored risk factors for its occurrence.


BACKGROUND
Cytomegalovirus (CMV) remains one of the most common opportunistic pathogens following solid organ transplantation (SOT) and is a major cause of morbidity and mortality. 1Patients at high risk for CMV disease include those who are CMV-seronegative and receive an allograft from a seropositive donor (D+/R-) and those CMV-seropositive who require anti-lymphocyte antibody treatment. 1High-risk SOT recipients are commonly given universal antiviral prophylaxis with valganciclovir (VGCV) for CMV prevention. 2 VGCV was developed as a prodrug given the poor oral bioavailability of oral ganciclovir (GCV) and a dose of VGCV 900 mg a day gives a drug exposure similar to intravenous GCV at a dose of 5 mg/kg/day. 3,4However, its use is limited by toxicities particularly myelosuppression, which can occur in up to 50% of transplant patients. 5,6][9][10][11] Management of neutropenia post-transplantation can be both costly and negatively impact outcomes, [12][13][14][15] particularly [19][20][21] Importantly, the definitions of leukopenia and/or neutropenia in SOT are variable or missing across the medical literature 6,12,21 which limits comparability between studies.
Multiple drugs used in SOT patients can cause leukopenia and/or neutropenia but the exact contribution of each drug is difficult to quantify given the use of combination therapies that include immunosuppressants and antimicrobials for prophylaxis. 20One common initial intervention when SOT patients develop neutropenia is dose reduction of the antimetabolite agent such as mycophenolate mofetil (MMF) or mycophenolic acid (MPA) followed by withholding of trimethoprimsulfamethoxazole.When neutropenia persists, VGCV may have to be discontinued prematurely which can result in primary CMV infection at the period of most intense immunosuppression.Although VGCV-induced neutropenia is dose-dependent, dose reduction with otherwise stable renal function is not recommended as it may result in subclinical CMV DNAemia and thus promote drug resistance. 6anulocyte-colony stimulating factor (G-CSF) is generally safe in SOT 22,23 but it is associated with significant costs. 6Some studies have suggested no association between leukopenia and neutropenia and/or rates of graft rejection and/or infection 20,21,24 but others have suggested the contrary. 12A systematic review assessing the burden of leukopenia/neutropenia in adult kidney transplant recipients by Raval and colleagues showed that CMV D+/R-status, mycophenolic acid, and tacrolimus were the most common risk factors for leukopenia/neutropenia.Although there was a positive association between leukopenia/neutropenia and acute rejection and opportunistic infections, the results were mixed for graft failure and mortality. 21single-center study in heart transplant recipients found CMV D+/R-status, lower baseline leukocytes, and previous CMV infection to be risk factors for post-transplant neutropenia. 25recent non-inferiority clinical trial by Limaye and colleagues compared VGCV to letermovir for 200 days in CMV D+/R-kidney transplant recipients. 26Leukopenia was defined as a white blood count (WBC) of less than 3500/µL and neutropenia as an absolute neutrophil of less than 1000/µL.The rates of leukopenia and neutropenia through week 28 were significantly higher in the VGCV arm, 64% versus 26% (p < .001),resulting in higher rates of early discontinuation of CMV prophylaxis, 13.5% versus 4.1%, due to adverse events. 26 hypothesized that VGCV prophylaxis in CMV D+/R-SOT recipients results in a high incidence of neutropenia and that neutropenia secondary to VGCV in the first year post-SOT may affect outcomes such as rejection.Our aims were: 1. to analyze the impact of VGCV prophylaxis on neutropenia rates in a multiorgan cohort of SOT recipients; 2. to perform a time-to-event analysis to assess risk factors for neutropenia; 3. to determine the effect of neutropenia on acute rejection rates.CMV-seropositive SOT recipients receiving VGCV prophylaxis were excluded, as the majority of these would have received anti-thymocyte globulin (ATG) induction at our center.CMV D-/R-SOT recipients who received GCV/VGCV prophylaxis for other reasons, such as Epstein-Barr Virus prophylaxis, were also excluded.

Study design and population
The primary outcome was significant neutropenia, as defined below, and the secondary outcomes were CMV infection, CMV disease, acute rejection, and death.
Participants entered the study at the time of transplant and were followed until death or censoring; the latter was defined as loss to follow-up or administrative censoring on June 19, 2019.
This study was approved by the University of Alberta Research Ethics Board (Pro00058690).

Definitions
Leukopenia was defined as WBC <3500/µL in at least two consecutive measurements, with each measurement at least 24 h apart.
Pre-transplant leukopenia was defined as WBC <3500/µL 0 to 30 days before transplant.Neutropenia was defined as absolute neutrophil count (ANC) <1500/µL; significant neutropenia: ANC <1000/µL; severe neutropenia ANC <500/µL, in at least 2 consecutive measurements, with each measurement at least 24 h apart.Acute kidney injury (AKI) was defined according to the KDIGO 2012 guidelines as an increase in serum creatinine ≥1.5 times baseline, which was known or presumed to have occurred within 7 days, or an increase in serum creatinine of 26.5 micromoles/L (equivalent to ≥ 0.3 mg/dL) or more within 48 h. 27ta on leukopenia, neutropenia, and acute kidney injury were only collected during VGCV prophylaxis.
Febrile neutropenia was defined as a single temperature of >38.

Antiviral prophylaxis
CMV D+/R-SOT recipients were expected to complete 3 months of VGCV prophylaxis, except small bowel and multivisceral transplant recipients who received 6 months, and lung transplant recipients who received 12 months.CMV D+/R-kidney and simultaneous kidney-pancreas transplant recipients received 3 months of antiviral prophylaxis from 2005 to 2012 and 6 months thereafter.
CMV D-/R-recipients of kidney, heart, lung, heart-lung, pancreas, or small bowel transplants who were HSV seropositive, as well as liver transplant recipients with anti-lymphocyte antibodies induction received acyclovir or valacyclovir for the initial 30 days after SOT.

CMV monitoring
In October 2005, an in-house quantitative polymerase chain reaction (PCR), with results expressed in CMV genome copies per mL, was started.In March 2012, RealStar CMV PCR (Altona Diagnostics), which measures the viral load in International Units (IU) per mL, was implemented.On average, the conversion factor is 2-fold, that is, 2 copies/mL by the in-house assay are about 1.0 IU/mL by the Real Star assay.The detection threshold for the in-house assay was below 500 copies/mL (250 IU/mL).All samples below 250 IU/mL were considered negative.
Surveillance after prophylaxis was performed with weekly CMV PCR for an additional 12 weeks.No routine CMV monitoring was performed during universal prophylaxis.

Pneumocystis jirovecii pneumonia prophylaxis
All SOT recipients received trimethoprim-sulfamethoxazole (TMPSMX) for Pneumocystis jirovecii pneumonia (PJP) prophylaxis for a minimum of 6 months, with lung transplant recipients receiving lifelong PJP prophylaxis.In our center, patients with anaphylaxis or delayed hypersensitivity to sulfa drugs, undergo TMPSMX desensitization at the time of transplant.

Statistical analysis
Categorical variables were expressed as proportions and compared by Chi-square test.Continuous variables were expressed as means and standard deviation if normally distributed and medians and interquartile range if not normally distributed and compared by a t-test or Wilcoxon rank-sum testing.Associations between covariables and neutropenia were assessed using competing-risks survival regression expressed as crude or adjusted sub-distribution hazard ratios (SHRs) with the respective 95% confidence intervals (CIs).The significance level was set at a two-sided p-value of less than .05.Cumulative incidence was visualized using competing-risk regression plots and survival analysis was visualized using Kaplan-Meier plots.The null hypothesis was tested via the log-rank test.
To help with interpretation, the number of days on VGCV prophylaxis was treated as a categorical variable based on the following five ordinal categories: 0 days, >0 and ≤90 days, >90 and ≤180 days, >180 and ≤270 days, and >270 days.
Multivariable competing-risks survival regression models assessing the relationship between VGCV prophylaxis as a time-varying variable (i.e., by adding interaction term with time), and significant neutropenia (<1000/ µL), accounting for the competing risk of death,  Patients in the VGCV-exposed cohort were less likely to be on MMF/MPA (p < .001),tacrolimus (p = .035),and prednisone (p =

Characteristics of the cohort
.041) at 6 months.Although the proportion of kidney transplants was higher in VGCV-unexposed SOT recipients and the proportion of liver transplants higher in VGCV-exposed SOT recipients, characteristics were otherwise relatively similar between the groups (Table 1).The cumulative incidence of significant neutropenia (ANC < 1000/µL) was significantly higher in VGCV-exposed compared to VGCVunexposed (log-rank p < .001, Figure 2).Although the cumulative incidence of neutropenia did not significantly differ by transplant type, heart transplant recipients had the lowest incidence of neutropenia (Figure 1).
Overall, VGCV prophylaxis in CMV D+/R-SOT recipients increased the incidence rate of significant neutropenia by 13.5-fold (

VGCV prophylaxis duration and early discontinuation rates
The median duration of VGCV prophylaxis in CMV D+/R-was 90 days   In CMV D+/R-, early discontinuation or temporary withholding of VGCV compared to no interruptions to VGCV prophylaxis did not result in higher rates of CMV disease [26 (40%) versus 47 (34.1%), respectively (p = 0.410)].

Febrile neutropenia and acute rejection
Episodes of febrile neutropenia up to 30 days after stopping VGCV VGCV-unexposed SOT recipients (p = .305).In multivariable competing risk analysis, CMV D+ status was associated with an increased risk of acute rejection (Figure S2), but significant neutropenia was not (Table S1).

TA B L E 3
Univariable and multivariable competing risk and cause-specific Cox regression analyses of factors associated with significant neutropenia.

Adjusted HR (95% CI) Model B p-Value
Recipient age (continuous)    Notably, AKI during VGCV prophylaxis also increased the risk of significant neutropenia by 78%.This finding is intriguing and may reflect the challenges of adjusting VGCV dosing when renal function is not at a steady state.

Mortality
After adjusting for multiple factors, heart and liver compared to kidney transplant status were associated with an 87% and 71% decrease in risk in neutropenia, respectively, which was consistent across competing risk and Cox regression analyses.This is likely driven by the shorter duration of VGCV prophylaxis in such groups, but other factors may have contributed.It is also important to note that antimetabolite use at 6 months did not constitute an independent risk factor, but most SOT recipients received it during the period of VGCV prophylaxis.Although ATG and sirolimus were protective of neutropenia in the multivariable Cox model, this was not the case in the competing risk model.To our knowledge, this represents one of the largest studies in multiorgan SOT recipients assessing incidence rates of and risk factors for neutropenia related to VGCV exposure in CMV-seronegative SOT recipients.
Notably, significant neutropenia resulted in dose-reduction and discontinuation of antimetabolite agents and those with significant neutropenia had significantly higher rates of acute rejection.However, significant neutropenia did not increase the risk of acute rejection in multivariable analysis.
Letermovir is a relatively novel antiviral that was initially approved for CMV prevention in allogeneic stem cell transplant recipients 29  12 months albeit the rates of neutropenia were similar in the 2 arms. 30V D+/R-lung transplant recipients, who experience high rates of CMV infection, CMV disease, and death, 31,32 merit special attention.Pre-emptive therapy for CMV prevention is not recommended in this population 2 with robust evidence supporting the extension of VGCV prophylaxis to 12 months in those at high risk for CMV in order to decrease the rates of CMV disease.33 For example, in our cohort, 55% (11/20) of CMV D+/R-lung transplant recipients developed significant neutropenia and VGCV had to be discontinued or temporarily withheld in 45% (9/20).Although there is considerable interest in using letermovir as prophylaxis in this high-risk population, data on lung transplantation is limited to observational studies concerning reports of breakthrough CMV infection.34,35 Therefore, clinical trials evaluating the efficacy of letermovir in lung and other non-kidney transplant recipients are still needed.
Our study is not without limitations.First, this study is retrospective, with the potential for unmeasured confounders and data reporting errors.With regard to unmeasured confounders, we did not collect data on donor demographic factors.However, we do not believe those factors would be expected confounders in the relationship between VGCV exposure and the primary outcome of neutropenia.
Second, we did not collect data on non-CMV opportunistic infections, such as fungal infections, or bacterial bloodstream infections.
Third, our study was not adequately powered to perform subgroup analyses to assess risk factors for neutropenia specific to transplant groups.
Fourth, maintenance immunosuppression was assessed as a timefixed variable at 6 months, which does not reflect the entire exposure to immunosuppressants, particularly antimetabolites.
Last, this is a single center and our findings do not necessarily apply to centers with differing demographic factors, indications for transplant, the proportion of CMV D+/R-and CMV prevention strategies.
Additionally, these findings may not be applicable to CMV-seropositive SOT recipients.
In conclusion, VGCV prophylaxis in CMV D+/R-SOT recipients is limited by high rates of neutropenia.Future prospective studies should aim to identify those at high risk of VGCV-related neutropenia and assess alternative CMV prophylactic strategies in SOT recipients.

A
retrospective cohort study of adult (age ≥18 years) CMVseronegative SOT recipients from either CMV seropositive (CMV D+/R-) or seronegative (CMV D-/R-) donors, transplanted between July 2005 and March 2019, at the University of Alberta Hospital, Edmonton, Alberta.Patients with early death, re-transplantation, or loss of follow-up within the first 3 months post-transplant were excluded.Patients who underwent re-transplantation more than 3 months from the most recent transplant, were included if they were CMV seronegative immediately before re-transplantation.
were adjusted to factors presumed to be associated with the outcome of neutropenia, such as age, transplanted organ, ATG or anti-IL-2 (basiliximab or daclizumab) induction, maintenance immunosuppressants at 6 months (i.e., antimetabolite, calcineurin inhibitor, prednisone, or sirolimus), pre-transplant leukopenia and AKI during VGCV prophylaxis (model A).For comparison, we calculated cause-specific hazard ratios (HRs) via multivariable Cox regression using the same variables (model B).Additional information on competing risk models for acute rejection can be found in the Supporting Information Appendix.Proportional hazard assumptions were tested with Schoenfeld residuals for Cox regression models and with visual inspection of log-log plots for competing risk regression models.All analyses were performed with Stata/SE (StataCorp LLC, version 17.0).

4 DISCUSSION
In this single-center retrospective cohort study of 430 multiorgan SOT recipients, VGCV prophylaxis for up to 3 months versus no prophylaxis increased the risk of significant neutropenia by nearly 40-fold on competing risk multivariable analysis.Importantly, significant neutropenia F I G U R E 3 Kaplan-Meier survival plots for cytomegalovirus-donor seropositive/recipient seronegative (CMV D+/R-) (valganciclovir [VGCV]-exposed) versus CMV D-/R-(VGCV-unexposed).developed <3 months in 43.8% and between 3 and 6 months in 42.7% of VGCV-exposed recipients.Although VGCV duration >6 months versus no VGCV was not associated with increased neutropenia risk, this is likely due to the fact that this study is underpowered to detect differences in that category since only 22 SOT recipients received VGCV >6 months.Nonetheless, this finding supports the notion that patients who develop VGCV-related neutropenia likely do so earlier into the VGCV course.
and was most recently granted approval by the United States Food and Drug Administration for CMV prevention in CMV D+/R-kidney trans-plant recipients based on the findings of a clinical trial led by Limaye and colleagues. 26In this trial, letermovir was deemed non-inferior to VGCV, and neutropenia (defined as ANC < 1000/µL) through day 28 occurred in 4.1% in the letermovir arm compared to 19.5% in the VGCV arm [rate difference −15.4 (−20.7 to −10.5)].In another clinical trial in CMV D+/R-liver transplant recipients, pre-emptive CMV therapy versus VGCV prophylaxis resulted in a lower incidence of CMV disease at Belga: Conception and design; data analysis and interpretation; drafting.Cristina Hernandez: Data collection and critical revisions; Dima Kabbani: Data interpretation and critical revisions; Carlos Cervera: Conception and design; data analysis and interpretation; drafting and critical revisions.All authors approved the final version of the manuscript.