Cytomegalovirus reactivation risk after autologous hematopoietic stem cell transplantation: Results of a Tunisian study

Abstract

BACKGROUND

Cytomegalovirus (CMV) reactivation is a potentially severe complication in immunocompromised patients, yet its incidence and impact in recipients of autologous hematopoietic stem cell transplantation (AHSCT) remain insufficiently documented.

AIM

To assess the frequency, timing, and outcomes of CMV reactivation in patients undergoing AHSCT at Aziza Othmana Hospital.

METHODS

We conducted a retrospective descriptive study of all patients who underwent AHSCT between January 2022 and December 2024 and had at least one post-transplant plasma viral load (VL) assessment. CMV VL was quantified by real-time polymerase chain reaction using TaqMan probes (GeneProof^®) with a sensitivity threshold of 150 IU/mL.

RESULTS

Among 277 AHSCT recipients, 17 (6.1%) experienced CMV reactivation. Their median age was 43 years, with a sex ratio of 0.46 (male/female). Underlying diseases included large B-cell lymphoma (n = 5), multiple myeloma (n = 3), and Hodgkin’s lymphoma (n = 4). The median time to reactivation was 26 days post-transplant (11 days after neutrophil recovery). Median peak VL was 1325 IU/mL (range: 150-641000 IU/mL). Six patients required antiviral therapy (median peak VL: 30150 IU/mL), while 11 had spontaneous resolution (median peak VL: 1320 IU/mL). Two patients died in the context of CMV reactivation.

CONCLUSION

CMV reactivation occurs in a noteworthy proportion of AHSCT recipients and may lead to severe outcomes. Routine VL monitoring in the early post-transplant period is crucial, and preemptive therapy should be initiated once clinically relevant VL thresholds are reached to prevent progression to CMV disease and associated mortality.

Key Words: Cytomegalovirus; Reactivation; Autologous hematopoietic stem cell transplantation; Preemptive therapy; Viral load monitoring

Core Tip: This retrospective cohort study analyzed cytomegalovirus (CMV) reactivation after autologous hematopoietic stem cell transplantation in 277 patients. CMV reactivation occurred in 6.1% of cases, often spontaneously resolving but occasionally leading to severe disease and death in patients with high viral loads. Initial and peak viral loads were significantly associated with treatment need, supporting viral quantification as a key decision tool. These findings emphasize the importance of systematic post-transplant CMV monitoring and early preemptive therapy to prevent progression to clinically significant CMV disease.

INTRODUCTION

Cytomegalovirus (CMV) reactivation is a well-known and potentially severe complication in immunocompromised hosts, particularly recipients of allogeneic hematopoietic stem cell transplantation (allo-HSCT), where it is a major cause of morbidity and non-relapse mortality[1]. In contrast, its clinical significance in recipients of autologous hematopoietic stem cell transplantation (AHSCT) has historically been minimized and remains insufficiently documented. This perception stems from the absence of graft-versus-host disease and a typically faster immune reconstitution, which theoretically should lower the risk[2].

However, the myeloablative conditioning regimens used prior to AHSCT induce a period of profound cytopenia and cellular immunodeficiency, creating a credible risk for viral reactivation. Emerging evidence suggests that CMV reactivation in AHSCT recipients is not as rare as once thought and can lead to serious end-organ disease, treatment delays, and even fatal outcomes, especially in patients with certain underlying malignancies like lymphomas[3].

Despite these risks, there are no standardized guidelines for CMV monitoring and preemptive therapy in the autologous setting. Practices vary widely between institutions, and the viral load (VL) thresholds that should trigger intervention are not well-defined[4,5].

Therefore, the objective of this retrospective cohort study was to evaluate the epidemiology and clinical course of CMV reactivation in a large, contemporary cohort of AHSCT recipients. We sought to determine its incidence, identify timing patterns, characterize the VL dynamics, and assess associated outcomes to build an evidence base for optimal management strategies.

MATERIALS AND METHODS

We conducted a retrospective descriptive study of all patients who underwent AHSCT between January 2022 and December 2024 at Aziza Othmana Hospital (Tunis) and had at least one post-transplant plasma VL assessment. As this was an exhaustive retrospective cohort study, all consecutive patients meeting the inclusion criteria during the study period were included, and no sample size calculation was performed a priori. Data were retrospectively collected from patients' medical records. We collected information on patients' demographics, underlying diseases, CMV serostatus, conditioning regimens, CMV reactivation patterns, VL dynamics, treatment modalities, clinical outcomes, and overall survival.

Definitions

CMV reactivation was defined as a single plasma CMV DNA level ≥ 150 IU/mL detected by real-time polymerase chain reaction (PCR). Patients with detectable but sub-threshold VLs (< 150 IU/mL) were not considered to have reactivation.

Neutrophil engraftment was defined as achieving an absolute neutrophil count > 0.5 × 10⁹/L for 3 consecutive days.

Probable CMV disease: Compatible clinical signs and symptoms along with CMV detection (such as viremia) but without histopathological confirmation from tissue biopsy.

Proven CMV end-organ disease was defined according to established criteria: (1) CMV pneumoniae required the presence of pulmonary infiltrates on imaging with histopathological or bronchoalveolar lavage confirmation of CMV; and (2) CMV colitis was defined by gastrointestinal symptoms with endoscopic and histopathological evidence of CMV infection.

CMV monitoring protocol

CMV VL was quantified using real-time PCR with TaqMan probes (GeneProof^®, Czech Republic) with a lower limit of quantification of 150 IU/mL.

Antiviral prophylaxis and treatment strategy

All patients received antiviral prophylaxis with acyclovir 400 mg orally twice daily from day-7 until neutrophil engraftment, followed by 400 mg twice daily until day +30 post-transplant in the absence of reactivation. Preemptive therapy was implemented at the clinician discretion.

First-line preemptive therapy consisted of (1) Ganciclovir: 5 mg/kg intravenously every 12 hours for induction (7-14 days) followed by 5 mg/kg daily for maintenance, provided adequate blood counts (absolute neutrophil count > 1.0 × 10⁹/L, platelets > 50 × 10⁹/L); (2) Valganciclovir: 900 mg orally twice daily for induction followed by 900 mg daily for maintenance (alternative to intravenous ganciclovir when oral route feasible); and (3) Foscarnet: 90 mg/kg intravenously every 12 hours for induction followed by 90 mg/kg daily for maintenance (reserved for patients with cytopenias or ganciclovir resistance).

Treatment duration was individualized based on VL response, with induction therapy continued until two consecutive negative PCR results, followed by maintenance therapy for 2-4 weeks.

All transfused blood products used in the transplant unit are systematically leukoreduced to minimize the risk of transfusion-transmitted CMV infection.

Statistical analysis

Descriptive statistics were performed on demographic and clinical data, presenting counts and percentages for categorical variables and median (interquartile ranges) for continuous variables. χ² test or Fisher's exact test was used to compare categorical variables, while Mann-Whitney U test was used for continuous variables, depending on data distribution and sample sizes. Time to reactivation was calculated from the date of stem cell infusion to the first positive CMV PCR result. P < 0.05 was considered statistically significant. All analyses were performed using Statistical Package for the Social Sciences version 25.0 (IBM SPSS Statistics 25).

RESULTS

Cohort characteristics

A total of 277 autologous stem cell transplant (ASCT) recipients were included in the study. Among them, 17 patients (6.1%) experienced CMV reactivation, defined as a VL ≥ 150 IU/mL. The median age of affected patients was 43 years (interquartile ranges: 32-58), with a notable female predominance (sex ratio male/female: 0.46). The distribution of underlying diseases in this group was as follows: (1) Large B-cell lymphoma (n = 5, 26.3%); (2) Multiple myeloma (n = 3, 21.1%); and (3) Hodgkin's lymphoma (n = 4, 21.1%). The remaining six patients had other lymphoma subtypes.

Patients without antiviral treatment

Of the reactivated patients, 11 (65%) did not require specific antiviral therapy. In this group, the median age was 39 years. The median initial VL was 690 IU/mL and the median peak VL was 1320 IU/mL. The median time to reactivation was 25 days post-transplant, with neutrophil engraftment occurring at a median of 12 days. Ten patients had a favorable course with spontaneous clearance of CMV viremia. One patient developed probable CMV gastrointestinal disease and died.

Patients with antiviral treatment

The remaining six patients (35%) required antiviral therapy. Their median age was 45 years. The median initial VL was 1631 IU/mL, and the median peak VL reached 30150 IU/mL. The median time to reactivation was 24 days post-transplant, with a median time to neutrophil engraftment of 11 days. Antiviral regimens included foscarnet (n = 3), ganciclovir (n = 2), and valganciclovir (n = 1). Five treated patients achieved viral clearance, with a median time to negativity of 46 days and a median duration from treatment initiation to viral clearance of 33 days. Among them, three developed probable CMV disease, all with favorable outcomes.

Two fatalities occurred in this cohort (11.8% mortality rate among reactivated patients)

Patient 1 was a 55-year-old female with diffuse large B-cell lymphoma, transplanted on D + 0. Neutrophil engraftment occurred on D + 12. CMV was first detected at D + 28 with a VL of 2450 copies/mL, peaking at 196000 copies/mL on D + 36. Foscarnet therapy was started on D + 40, but the patient died on D + 41 with gastrointestinal symptoms consistent with probable CMV gastrointestinal disease.

Patient 2 was a 59-year-old female with diffuse large B-cell lymphoma, transplanted on D + 0. Neutrophil engraftment occurred on D + 10. CMV reactivation was first detected on D + 30 with a VL of 915 copies/mL, peaking at 4360 copies/mL. The patient developed probable CMV pneumoniae associated with candidemia on D + 33 and died on D + 35.

In both cases, the diagnosis of CMV disease was based on compatible clinical presentation and significant viremia in the absence of histopathological confirmation.

Comparative analysis

When comparing patients requiring antiviral therapy with those who spontaneously cleared CMV, initial VL (P = 0.015) and peak VL (P = 0.007) were significantly higher in the treated group. No significant differences were observed for age (P = 0.309), sex (P = 0.205), time to reactivation (P = 0.66), or time to neutrophil engraftment (P = 0.884). Detailed comparisons are summarized in Table 1.

Table 1 Comparison between patients requiring antiviral therapy and those with spontaneous clearance, median (interquartile ranges).

Variable	Treated (n = 6)	Not treated (n = 11)	P value (Mann-Whitney U/Fisher exact)
Age (years)	45 (33, 61)	39 (23, 60)	0.309
Sex (male/female)	4/2	6/5	0.205
Initial VL (IU/mL)	1631 (612, 50200)	690 (150, 1990)	0.015
Peak VL (IU/mL)	30150 (812, 641000)	1320 (150, 18000)	0.007
Time to reactivation (days)	24 (9, 29)	25 (5, 38)	0.66
Time to neutrophil engraftment (days)	11 (10, 12)	12 (10, 19)	0.884

VL: Viral load.

DISCUSSION

This retrospective cohort study of 277 ASCT recipients reported a 6.1% incidence of CMV reactivation, defined by a VL ≥ 150 IU/mL. This is consistent with the clinical understanding that CMV reactivation is less frequent after autologous transplantation than after allogeneic transplantation, where rates commonly range between 20%-50%, depending on factors like conditioning regimens and patient risk profiles. Variability in CMV reactivation rates across studies may be attributed to differences in conditioning intensity, immune recovery, monitoring frequency, and transfusion policies. In our setting, systematic leukoreduction of transfused blood products and the absence of graft-versus-host disease likely contribute to the relatively low reactivation rate observed. The 6.1% numbers fit well within the 4%-20% range reported in previous literature for ASCT patients, highlighting the relative rarity of CMV reactivation in this setting compared to allo-HSCT cohorts, which show much higher rates (up to approximately 50%)[6,7].

The predominance of female patients and the underlying diseases (mainly B-cell lymphomas and multiple myeloma) reflect typical populations undergoing ASCT, supporting the external validity of this cohort[8].

Among the 17 reactivated patients, 6 (35.3%) required antiviral therapy, while the remaining 11 (64.7%) experienced spontaneous viral clearance. Our analysis demonstrated that both initial VL and peak VL were significantly higher in patients who required treatment (P = 0.015 and P = 0.007, respectively), whereas age, sex, time to reactivation, and time to neutrophil engraftment showed no significant association with therapeutic need (Table 1). This key finding aligns with current preemptive therapy protocols that rely heavily on VL as a trigger for intervention rather than patient demographics or other parameters, as demonstrated by Tan et al[9]; who showed that preemptive antiviral therapy relies primarily on VL measurements for initiation in hematopoietic cell transplant recipients. Their study found that starting treatment at lower VLs shortened the duration of viremia and reduced clinical risk, underscoring VL as the main determinant for therapy rather than patient demographics or timing variables. Similarly, other studies have confirmed that VL monitoring represents the most reliable predictor of clinically significant CMV infection in transplant recipients[10]. Furthermore, a recent systematic review by Sadowska-Klasa et al[11]; concluded that antiviral preemptive therapy initiated at CMV VL thresholds of approximately 2-3 Log10 IU/mL effectively prevents CMV disease progression, providing robust evidence supporting VL as the key trigger for treatment initiation across transplant populations[9,11].

Clinical outcomes were generally favorable across both treatment groups, reflecting the overall good prognosis associated with CMV reactivation in the autologous setting. Among patients with spontaneous viral clearance, the majority experienced uncomplicated courses, although one patient developed probable CMV gastrointestinal disease and subsequently died. This observation underscores that even patients with lower VLs can occasionally progress to severe disease, emphasizing that CMV can remain a serious threat even in the absence of antiviral therapy, consistent with findings by Fesler et al[12], who noted that while ASCT patients have a relatively low risk of CMV reactivation, outcomes can be serious when CMV disease develops, supporting the need for vigilant monitoring.

Among the six treated patients, four achieved successful viral clearance with favorable outcomes, while two fatalities occurred. Both deaths were associated with exceptionally high VLs and severe end-organ involvement, including gastrointestinal disease in one case and pneumopathy with concurrent candidemia in the other. These outcomes reinforce that despite preemptive antiviral treatment, CMV can remain a serious threat at high viral burdens, aligning with the findings of Lee et al[3], who analyzed CMV reactivation after autologous stem cell transplantation and reported variable outcomes, including some mortality associated with severe CMV disease despite antiviral treatment, highlighting the significant clinical impact of CMV in this setting. Three of the treated patients developed probable CMV disease but all responded favorably to antiviral therapy, demonstrating the efficacy of prompt therapeutic intervention in appropriately selected cases.

The overall CMV-related mortality rate of 17.6% (3/17 reactivated patients) in our cohort resonates with reports from other autologous transplant settings, underscoring the potential severity of CMV complications despite their relative infrequency in the ASCT context. These outcomes emphasize that while immune recovery post-AHSCT often suffices to control viral reactivation spontaneously, close monitoring remains essential to identify patients at risk for severe complications.

The study highlights ongoing clinical challenges: Although CMV reactivation post-ASCT is less common, it can lead to severe outcomes, demanding vigilant monitoring. There remain no standardized guidelines specific to CMV monitoring and treatment thresholds in ASCT recipients, resulting in varying clinical practices and potentially affecting patient outcomes. Future prospective studies with larger cohorts and standardized monitoring protocols are necessary to optimize management strategies[13].

Study limitations

Our findings are limited by the lack of detailed demographic and clinical data for the full cohort, preventing the assessment of other potential risk factors such as comorbidities, disease status at transplant, or prior therapy. Additionally, our study has several limitations inherent to its retrospective and single-center design. The frequency of VL monitoring was not standardized and was performed at the discretion of the treating clinician, potentially leading to an underestimation of the true incidence of reactivation, particularly short-lived, low-level viremias. Furthermore, the small number of events limits the power of statistical analysis to identify independent risk factors for reactivation or for progression to high-level viremia.

CONCLUSION

This comprehensive analysis of CMV reactivation dynamics and clinical outcomes in 277 AHSCT recipients provides valuable real-world evidence that enriches our understanding of viral complications in this population. Our results demonstrate that while CMV reactivation occurs in only 6.1% of patients, it represents a clinically significant event that can lead to severe complications and mortality, particularly in patients with high VLs. The strong correlation between VL parameters and treatment necessity, independent of demographic factors or clinical timing, reinforces VL quantification as both a predictive and therapeutic decision-making tool, consistent with current international guidelines.

Based on our findings and their alignment with existing literature, we advocate for implementing a structured, risk-adapted management approach for CMV in AHSCT recipients that includes: (1) Regular monitoring of CMV DNAemia via PCR during the first 2-3 months post-transplant for all at-risk patients, particularly CMV-seropositive recipients; (2) Preemptive therapy initiated upon detection of VLs exceeding predefined thresholds (approximately 1000-5000 IU/mL or 2-3 Log10 IU/mL) to prevent progression to end-organ disease; and (3) Sustained vigilance for late reactivation in patients with prolonged immunosuppression or additional risk factors. These findings also highlight that, although CMV reactivation is less frequent in AHSCT than in allo-HSCT, its clinical impact can be significant in cases of high VL or delayed diagnosis.

Despite the inherent limitations of our retrospective, single-center design, this study contributes important insights into the clinical course and outcomes of CMV reactivation in a real-world autograft population. The variability in current monitoring practices and the demonstrated potential for severe outcomes underscore the urgent need for standardized, evidence-based protocols. Future prospective, multicenter studies with larger cohorts and comprehensive risk factor assessment are essential to definitively establish optimal monitoring schedules, refine VL thresholds for treatment initiation, and develop risk stratification algorithms that will optimize CMV management and improve outcomes for AHSCT recipients.

ACKNOWLEDGEMENTS

The authors thank the medical and nursing staff of the Hematology Unit at Aziza Othmana Hospital for their assistance in patient management and data retrieval. We also acknowledge the contributions of the Microbiology Laboratory team for their support in cytomegalovirus polymerase chain reaction testing and quality control.