Introduction

Adenovirus (ADV) infections contribute to morbidity and increased mortality after hematopoietic cell transplantation (HCT) [1, 2]. The clinical manifestations of ADV infections in HCT patients most frequently include upper respiratory disease, gastroenteritis, and kerato-conjunctivitis, while encephalitis, myocarditis, and pneumonia occur sporadically. Local infections can disseminate to cause systemic disease and/or progress to lethal disease [2]. Diagnostic and therapeutic management of infections with ADV in immunocompromised host was presented by several working groups [1,2,3,4,5].

Existing data suggest that the incidence of ADV infections in patients after HCT is higher in children than in adults [2, 4]. Recent data showed the incidence of ADV infections in 7.4% of children and 2.9% of adults after allogeneic HCT (allo-HCT) [6], while this infection does not seem to be a major issue after chimeric antigen receptor T-cells (CAR-T) therapy [7, 8]. Further pediatric analysis showed the incidence of ADV infections of 10.5% after allo-HCT and 1.3% after autologous HCT (auto-HCT) [9]. Recent US analysis showed the incidences of ADV infection, any ADV viremia, and ADV viremia ≥1000 copies/mL within 6 months after the allo-HCT in children to be 23%, 16%, and 9%, respectively, while for adults being 5%, 3%, and 2%, respectively [10]. Even higher rates were documented in AdVance study: 32 and 6% of cumulative incidence of ADV infection in children and adults, respectively; and the incidence of ADV viremia ≥1000 copies/mL was 14% and 1.5% in children and adult recipients, respectively [11, 12].

In this European Society for Blood and Marrow Transplantation (EBMT) registry-based study we aimed to analyze the outcome of ADV infection in children and adults after allo-HCT, as well as clinical forms of ADV infection and risk factors associated with adverse outcomes of ADV infection after allo-HCT.

Methods

Study design

All patients with an allo-HCT performed between 2000 and 2022 reported to the EBMT database, who had ADV infection after HCT, were analyzed in this retrospective registry-based study. Basic data of ADV infection including clinical type, diagnostics, therapeutic strategy, and outcome were analyzed. No exclusion criteria were implemented.

Data collection

Transplant data available on age, sex, primary diagnosis, type of donor, type of conditioning (reduced intensity conditioning [RIC] or myeloablative conditioning [MAC], according to the EBMT definitions), remission status of the underlying malignancy at allo-HCT, source of stem cells (i.e., peripheral blood [PB] or bone marrow [BM]), donor and recipient cytomegalovirus (CMV) serostatus, acute/chronic GVHD, concomitant infections, and outcome were analyzed. Acute graft-versus-host disease (aGVHD) was graded according to the criteria of Glucksberg et al. [13], chronic GVHD (cGVHD) was graded as limited or extensive. This study was approved and performed by the Infectious Diseases Working Party (IDWP) of the EBMT in accordance with the principles of the Declaration of Helsinki.

Clinical diagnosis of ADV infection was done locally in the transplant center and reported as: viremia, gastrointestinal infection, cystitis, pneumonia, hepatitis, central nervous system (CNS) infection, septic shock, multiorgan failure (MOF), or other.

Statistical analysis

The primary endpoint was the probability of overall survival (OS) at days +30, +100, at 1 year after allo-HCT. OS was defined as the time from allo-HCT to the last observation or death, regardless of the cause. OS was computed using the Kaplan–Meier estimator, with the log-rank test being used to perform univariate comparisons and the Greenwood formula to compute confidence intervals. Non-relapse mortality (NRM) was defined as death after transplant that was not preceded by recurrent or progressive primary disease. The cumulative incidence of acute and chronic GVHD was estimated considering the corresponding type of GVHD as an event of interest and death without GVHD as a competing risk setting, with the Gray test being used to compare different groups and the delta method to obtain confidence intervals. Factors considered in risk factor analysis were gender, diagnosis, the patient’s age at transplantation, the donor’s age at transplantation, donor/recipient sex matching (female donor to male recipient vs other), type of donor (HLA-matched sibling vs other donor types), type of conditioning (RIC; MAC), disease status at the time of transplantation (i.e., first complete remission vs other), source of stem cells (PB; BM; cord blood, CB), year of transplantation and CMV serostatus of the recipient (seronegative vs seropositive), donor (seronegative vs seropositive), and donor/recipient combinations (seronegative/seronegative, seropositive/seronegative, seronegative/seropositive, seropositive/seropositive).

Comparisons for categorical variables were done using Fisher’s exact test or the χ2 test. The proportional hazard assumption was verified using graphical methods: scaled Schoenfeld [14] residuals and graphical checks proposed by Klein–Moeschberger [15] were performed without finding evidence of relevant violations. The univariate and multivariate effect of variables on OS was analyzed using a Cox proportional hazards model in order to estimate hazard risks (HRs). All tests were two-sided, with the type I error rate fixed at 0.05. The Bonferroni correction was applied in the case of multiple subgroup comparisons. Median follow-up was calculated according to the inverted Kaplan–Meier technique [16]. All the analyses were performed using the statistical software SAS (SAS Institute Inc., Cary, NC, USA) version 9.2.

Results

Demographics

There were 2529 patients transplanted between 2000 and 2022 with reported ADV infection. Patient and transplant characteristics are presented in Table 1. With respect to age, 43.9% (n = 1110) were children (<18 years), and 56.1% (n = 1419) were adults. The median age of patients was 23.6 (min–max: 0.1–74.0) years; 62.4% were male. Primary diagnosis: 55.4% (n = 1401) acute leukemia (AL) or MDS, 20.9% (n = 529) other malignant diseases, and 23.7% (n = 599) non-malignant disorders. Stem cell source used for transplantation: PB in 51.7% (n = 1308), BM in 32.3% (n = 818), CB in 15.6% (n = 395), or their combinations in 0.4% (n = 8). According to donor type, the majority of patients had unrelated donors (63.2%), matched-related donors (22.5%), and mismatched-related donors (14.3%). Conditioning: myeloablative in 59.0% (n = 1491), RIC in 39.8% (n = 1006), not determined in 1.3% (32). Total body irradiation (TBI) based conditioning was used in 32.0% (n = 810) patients. Grades 2–4 aGVHD was diagnosed in 44.6%, cGVHD in 33.6% (limited in 17.3%, extensive in 16.3%) of eligible patients.

Table 1 Main characteristics of patients with ADV infection, with comparison of characteristics between children and adults.
Full size table

Clinical manifestations of ADV infection

Viremia was diagnosed in 62.6% (n = 1589), gastrointestinal infection in 17.9% (n = 453), cystitis in 3.7% (n = 94), pneumonia in 4.9% (n = 124), and other not specified in 9.7% (n = 246) patients, while number of reported cases of hepatitis, CNS infection, septic shock, and MOF was below 10 for each complication.

Survival after ADV infection

OS of the entire cohort of patients infected with ADV was 88.7% (95% CI = 87.4–89.9) at day +30, 75.1% (95% CI = 73.3–76.7) at day +100, and 61.1% (59.1–63.0) at 1 year (Fig. 1a, b). Patients infected with ADV within 30 days after HCT had lower OS and higher NRM at day +100 (Fig. 1c, d). The 100-day OS after diagnosis of ADV infections was 79.2% (95% CI = 76.6–81.5) in children and 71.9% (95%CI = 69.4–74.2) in adults (p < 0.0001) (Fig. 1e, Table 2). With respect to the clinical form of ADV disease, the 100-day OS was 82.3% in patients with gastrointestinal infection, 73.8% with viremia, and 66.3% in patients with pneumonia (p = 0.001). Patients with disseminated multiorgan disease had poor outcomes; 3/7 septic shock and 6/6 with MOF died by day 100.

Fig. 1: Survival probability.
figure 1

a OS for all patients; b 100-day NRM for all patients; c OS according to time from HCT to ADV infection up to day 30; d 100-day NRM according to time from HCT to ADV infection up to day 30; e OS in children and adults; f 100-day NRM according to calendar time of transplant; g OS according to CMV reactivation before ADV infection; h 100-day NRM according to CMV reactivation before ADV infection.

Full size image
Table 2 Overall survival in all patients with ADV infection (Kaplan–Meier analysis).
Full size table

With respect to cGVHD the 1-year OS was 54.7% (95% CI = 46.3–62.4) in patients with extensive cGVHD and 65.0% (95% CI = 62.9–67.1) without cGVHD or with limited form of cGVHD (p = 0.01).

In order to show a change in infection management over time (2000–2010 vs 2011–2022) we analyzed NRM in respective periods. The 100-day NRM probability was 27.7% (95% CI = 24.5–31.0) between 2000 and 2010 and 18.7% (95% CI = 16.8–20.5) between 2011 and 2022 (p < 0.0001) (Fig. 1f).

Risk factor analysis for overall survival

In multivariate analysis, factors contributing to worse OS at day +100 in all analyzed patients with ADV infection were: age > 18 years, type of ADV infection (pneumonia and viremia), diagnosis of AL or MDS, second or higher HCT, PB or CB stem cell source, time from HCT to ADV infection <100 days and Karnofsky/Lansky score <90 (data not shown). When age was analyzed as continuous variable, a 10-year increase in age was an adverse risk factor for survival with HR = 1.11 (95% CI = 1.08–1.14), p < 0.0001.

Since the risk of mortality was higher in adults at day +100 (HR = 1.48, 95% CI = 1.23–1.77; p < 0.0001), and also at 1 year (HR = 1.48, 95% CI = 1.21–1.81; p < 0.0001), the risk factors analysis was performed separately for children and adults (Table 3).

Table 3 Univariate and multivariate analysis of 100-day overall survival in children and adults (in Cox model).
Full size table

Factors contributing to increased risk of death by day +100 in children infected with ADV in univariate analysis were: type of ADV infection (viremia vs gastrointestinal infection), CMV seropositivity of donor or recipient, second or higher HCT, stem cell source (PB vs BM), ex vivo T-cell depletion, and Lansky/Karnofsky score <90. In multivariate analysis, two factors retained significance: CMV seropositivity of donor and/or recipient (p = 0.02), and Lansky/Karnofsky score <90 (p < 0.0001).

Factors contributing to increased risk of death by day +100 in adults infected with ADV in univariate analysis were: type of ADV infection (viremia or pneumonia vs gastrointestinal infection), second or higher HCT, shorter time from HCT to ADV infection (<30 days), and Karnofsky score <90. In multivariate analysis, three factors were significant: type of ADV infection (viremia or pneumonia vs gastrointestinal infection) (p = 0.0004), second or higher HCT (p = 0.0003), and shorter time from HCT to ADV infection (p = 0.003).

CMV coinfections

In this cohort of patients infected with ADV, the proportion of CMV-seropositive (46.3% children vs 60.2% adults) and CMV-seronegative (46.6% children vs 36.7% adults) recipients reflected general population trend with higher rate of CMV seropositivity in adults. Recipient and/or donor CMV seropositivity in patients infected with ADV contributed to worse 100-day OS in children, but not in adults (Table 3). Overall, 868/2511 (34.6%) patients had CMV reactivation before diagnosis of ADV infection. OS at 100 days and 1 year after ADV infection, as well as cumulative incidence of NRM did not differ between CMV-infected and CMV non-infected patients (Fig. 1g, h).

Causes of deaths

Out of 2529 analyzed patients infected with ADV, 938 (37.1%) patients died within 1 year after ADV infection, including 339 children (30.5% of total number of children) and 599 adults (42.2% of total number of adults) (OR = 1.7, 95% CI = 1.4–2.0; p < 0.0001). Death rate from relapse, progression or secondary malignancy at 1 year was 6.3% in children and 8.5% in adults (OR = 1.4, 95% CI = 1.02–1.9; p = 0.0359). NRM rate was 24.2% in children and 33.7% in adults (OR = 1.6, 95% CI = 1.3–1.9; p < 0.0001). Infection-related deaths were reported in 12.0% children and 13.3% adults (ns). Deaths from ADV infection were reported in 16 children (1.4% of all patients; 4.7% of all deaths in pediatric population) and 12 adults (0.8% of all patients; 2.0% of all deaths in adult population) (ns).

Discussion

ADV infection results in a wide array of clinical presentations [2, 3]. In immunocompetent individuals ADV usually causes mild disease, while severe and life-threatening infections can occur in the immunocompromised populations, such as allo-HCT patients [17, 18].

This study aimed to analyze clinical manifestations and outcome of ADV infections after allo-HCT in children and adults. The most frequent clinical forms of ADV infection were viremia (63.1%), and gastrointestinal infection (16.9%), while pneumonia and cystitis contributed to 4.9% and 3.7%, respectively. Patients with ADV viremia or pneumonia had significantly lower survival in comparison to patients with other clinical manifestations. Since gastrointestinal infection can progress to systemic viremia, the role of fecal screening in children should be underlined to identify the need of pre-emptive therapy in order to prevent progression to viremia.

In spite of only about 20% of allo-HCT being performed in children [19], 43.9% of reported ADV infections occurred in children. This shows that ADV infection is a much more frequent problem in children than in adults. OS was significantly higher in children than in adults at each time point: OS at day +100 after ADV diagnosis was higher in children than adults (79.2% vs 71.9%) and the risk of 1 year mortality was also higher in adults. These figures might reflect not only better survival from ADV infections in children, but also overall better post-transplant survival in children than adults. Undoubtedly, the impact of age itself of patients with ADV infection was a continuous risk factor for worse survival, reaching risk of 11% by each decade of life. Apart from age of patients, we found several major adverse risk factors for OS of patients with ADV infection. Interestingly, several differences between children and adults were observed.

Certain findings in our study translated to poorer outcomes. We focused on short-term outcome on 100 days, as this parameter is more closely related to clinical course of ADV infection than outcome at 1 year. Two factors contributed to poorer survival in children: CMV seropositivity of donor and/or recipient, and Lansky/Karnofsky score <90. Three factors contributed to poorer survival in adults: ADV viremia or pneumonia, second or higher HCT, and shorter time from allo-HCT to ADV infection (both <30 days and <100 days). It shows that factors directly related to ADV infection, such as early infection or viremia or pneumonia contribute to mortality in adults but not in children. On the other hand, pre-transplant factors such as CMV seropositivity or overall general performance score, but not factors related to ADV infection, contribute to mortality in children. Possible explanations include that children might face ADV infections more often than adults, but the survival is better in children than adults.

CMV seropositivity of donor and/or recipient (p = 0.02) was a significant adverse risk factor in children with ADV infection, but not in adults, which provides further evidence that CMV seropositivity negatively influences transplant outcomes, a well-known factor in transplant setting [20]. Recently, it has been shown to have adverse impact in patients infected with SARS-CoV-2 [21].

Lansky/Karnofsky performance score <90 was a significant adverse factor for children with ADV infection (p < 0.0001) in multivariate analysis, while for adults this was significant in univariate analysis only. Performance status is an important factor for survival after transplant and usually reflects presence of comorbidities [21].

Clinical type of ADV infection had a significant effect on survival in multivariate analysis in adults, with an adverse impact of ADV viremia or pneumonia in comparison to gastrointestinal infection (p = 0.0004). In children, viremia had an adverse effect, when compared to gastrointestinal infection, but in univariate analysis only. This finding underlines the low risk of severe complications in ADV-driven gastrointestinal infection. Based on recent reports, gut infection seems to be the most frequent clinical manifestation of ADV infection in transplant patients [5, 22].

Second or higher HCT had adverse impact in adults in multivariate analysis (p = 0.0003), while in children was significant in univariate analysis only. This observation is in line with knowledge on the worse transplant outcome in advances stages of disease, usually related to relapse of primary disease [23].

Finally, shorter time from transplant to ADV infection had significant adverse impact on survival in adults (p = 0.003). This was particularly evident if ADV infection appeared within the first 30 days after HCT, as shown on Fig. 1e. Surprisingly, this effect was not observed in children, even in univariate analysis. These data support the consideration of screening for ADV viremia in the first 100 days regardless of age, especially as ADV infection seems not to be a major problem in adult patients after day +100.

With respect to changes over calendar time, the 100-day NRM decreased from 27.7% between 2000 and 2010 to 18.7% between 2011 and 2022, underlining the progress in management of infection. However, shifts or progression in clinical course of infection, diagnostic and therapeutic strategies were not analyzed in this study.

This is one of the largest ever study of patients with ADV infection after allo-HCT. It underlines several important and practical findings. Two prognostic factors were related to ADV infection: clinical form of ADV manifestation, and time from HCT to ADV infection. Other factors identified are well-known associations of OS after allo-HCT, including BM as a more frequent source of stem cells in children than in adults, which may have influenced better survival in children [24].

It was shown recently that additional ADV-related factors can contribute to OS in patients with ADV infection after HCT. Peak ADV viral load and ADV time-averaged area under the curve correlated with NRM in T-cell depleted HCT and in pediatric patients, which supports the potential utility of ADV viral kinetics as endpoints in clinical analyses of ADV therapies [25, 26]. With increased ADV disease burden, there was an increased risk of mortality after both auto- and allo-HCT [27].

Our study has some limitations. Retrospective data, collected from the registry are imperfect, with quantitative and repeated data on blood ADV-DNA being not available. Moreover, the study covered relatively long period of inclusion, thus different diagnostic and therapeutic strategies were implemented in participating centers. There is also a possibility of a bias toward higher number of children being reported, reflecting maybe that screening in children is more often routinely used than in adults.

In conclusion, we have shown that in patients infected with ADV, short-term 100-day OS and non-relapse survival are better in children than adults. Factors directly related to ADV infection, such as early infection or viremia or pneumonia contribute to mortality in adults but not in children, while pre-transplant factors such as CMV seropositivity or overall general performance score, but not factors directly related to ADV infection, contribute to mortality in children. Results of this study might justify the need of regular screening both in children and adults.