Role of Respiratory Viral Infections in Airflow Obstruction After Hematopoietic Cell Transplantation

ABSTRACT

Background

Respiratory viral infections (RVIs) in allogeneic hematopoietic cell transplant (allo‐HCT) recipients can cause airflow obstruction (AFO). Despite this risk, the impact of community‐acquired RVI on PFTs in pediatric patients after allo‐HCT remains unclear.

Objective

The study determines the role of RVI in the development of AFO post‐transplant among survivors of allo‐HCT.

Study Design

This retrospective study analyzed St. Jude patients undergoing allo‐HCT between 2003–2020. AFO was defined by z‐scores and forced expiratory volume in 1 s (FEV1)/forced vital capacity (FVC) ratio expressed as lower limit of normal (LLN). Baseline parameters were obtained before allo‐HCT, and Year 1 parameters within 425 days post allo‐HCT. Spirometry measures were calculated using the global lung function Initiative (GLI) calculator. Bivariate and multivariate logistic regression identified risk factors for AFO following allo‐HCT.

Results

A total of 397 patients were included, with a median age of 12.7 years. RVI occurred in 13% of patients within the first 100 days after transplantation. AFO developed in 12% of patients. Among those with AFO, 68% had a baseline FEV1/FVC ratio above the LLN, and 11% had a history of lower respiratory tract infection (LRTI) (p < 0.05). The median time from transplant to RVI was shorter in patients with AFO (12 vs. 34 days). A baseline LLN value was a significant predictor of AFO. However, after adjusting for known covariates, RVI was not independently associated with AFO risk.

Conclusion

RVIs within 100 days of allo‐HCT did not significantly affect AFO following transplant in pediatric patients.

Respiratory viral infection severity was not independently associated with post‐transplant airflow obstruction. Earlier infection onset after allo HCT may be more important than viral virulence or host factors in long term pulmonary outcomes.

Abbreviations

AFO

airflow obstruction

AI

artificial intelligence

Allo‐HCT

allogeneic hematopoietic cell transplant

ALSAC

American Lebanese Syrian Associated Charities

BAL

bronchoalveolar lavage

BMSC

bone marrow stem cells

CI

confidence interval

CMV

cytomegalovirus

CT

computed tomography

EMR

Electronic medical records

ERS

European respiratory society

FEV

forced expiratory volume

FVC

forced vital capacity

GLI

global lung function initiative

GVHD

graft versus host disease

HRV

human rhinovirus/enterovirus

ICU

intensive care unit

IQR

inter‐quartile range

LLN

lower limit of normal

LRTI

lower respiratory tract infection

NA

not applicable

OR

odds ratio

PBSC

peripheral blood stem cells

PCR

polymerized chain reaction

PFT

pulmonary function test

PIV

parainfluenza virus infection

RSV

respiratory syncytial virus

RVI

respiratory viral infections

SAS

statistical analysis system

SJCRH

St. Jude Children's Research Hospital

TRM

treatment related mortality

URTI

upper respiratory tract infection

1. Introduction

Allogenic hematopoietic cellular transplant (allo‐HCT) is an efficacious treatment for patients with a range of malignant and nonmalignant diseases. Despite significant advances in allo‐HCT, treatment‐related mortality (TRM) remains a concern. Improved therapies have resulted in a growing population of long‐term survivors who may face late complications from allo‐HCT [1]. Among these complications, abnormal pulmonary function is prevalent and a main cause of death following disease relapse in allo‐HCT survivors. Moreover, recent studies have shed some light on the significant role of community‐acquired respiratory viral infections (RVIs) such as respiratory syncytial virus (RSV) infections as a vital cause of long‐term morbidity [2] and mortality among this cohort [3].

The impact of community RVIs on lung functions in children undergoing allo‐HCT has not been well studied, and to date, no research has thoroughly examined long‐term pulmonary outcomes in this context [2]. Erard and colleagues studied the first 100 days post allo‐HCT and identified community RVIs as a significant risk factor for the development of airflow obstruction (AFO) in adults [4].

In this study, we aimed to retrospectively evaluate the role of community RVIs in the development of AFO in children undergoing allo‐HCT using a nested cohort at St Jude Children's Research Hospital (SJCRH, Memphis, TN, USA). Furthermore, we hypothesized that RVIs in the first 100 days post allo‐HCT have a direct and lasting impact on long term lung function in this population.

2. Materials and Methods

2.1. Study Design, Setting, and Patients

We conducted a single‐center, retrospective, longitudinal study using spirometry and clinical data. SJCRH. Ethical approval was granted by SJCRH Ethics Committee. The cohort consisted of 438 cases treated with allo‐HCT from 1st January 2003 until 31st December 2020, had a pulmonary function test (PFTs) available before (baseline PFT) and had a routine RVI test in the first 100 days post allo‐HCT. Patients who did not undergo PFT measurements within the timeframe (36 patients) were excluded from the study. Patients diagnosed with nonrespiratory viruses were excluded (3 patients). Two duplicate cases were removed. The final cohort consisted of 397 patients (Figure S1).

2.2. Clinical Data

Data were obtained from a centralized electronic medical record (EMR) database housed at SJCRH. Donor type was determined according to donor‐recipient ABO compatibility and stem cells source. Donor type includes matched sibling donor, haplo, unrelated, or others. Stem cell source was classified as peripheral blood stem cells (PBSC), bone marrow stem cells (BMSC), or cord blood. Graft‐versus‐host disease (GVHD) type was classified as acute, chronic, both acute and chronic, or none. The preparative regimen was classified as myeloablative, nonmyeloablative, or reduced intensity. The types of RVIs, the occurrence of viral co‐infection, and their association with infection site were evaluated. Requirement for oxygen supplementation, mechanical ventilation, and the use of anti‐viral therapy were also assessed.

2.3. Study Definitions

RVIs were defined by microbiologically detected respiratory viruses on nasopharyngeal wash, mouth, tracheal aspirate, or bronchoalveolar‐lavage specimens. The type of RVI testing varied over time, from culture (2003–2009), direct fluorescent antibodies (2010–2013), to PCR (2014–2020). RVIs were classified based on their potential for disease and complications in this population: Class I: viruses that are found most likely to cause severe disease (respiratory syncytial virus, human metapneumovirus, parainfluenza virus, influenza A and B, and adenovirus); Class II: viruses that are found less likely to cause severe disease (human rhinoviruses, human seasonal coronaviruses, and human bocavirus) [3].

URTI was defined as the presence of respiratory signs or symptoms without radiological evidence of LRTI infection. LRTI was defined as a virus‐ positive lower respiratory tract sample (i.e., bronchoalveolar lavage [BAL], a lung biopsy specimen, or an autopsy specimen) with or without radiographic abnormalities or a positive upper respiratory tract sample with radiographic abnormalities or a new radiologic abnormality, positive BAL, or diagnosis by a physician within 28 days of the initial RVI diagnosis.

Criteria for assessing the degree of LRTI severity are not consistent in the literature and are often not applicable to this population, who may be over 24 months old and have baseline oxygen requirements. For these reasons, we defined severity based on oxygen use and the need for mechanical ventilation: mild (no need for supplemental oxygen), moderate (supplemental oxygen for greater than 4 h), or severe (need for mechanical ventilation). Baseline RVI locations were classified as URTI, LRTI, or no RVIs. ICU admission was determined by overall clinical status rather than by the presence of a positive RVI test alone.

Co‐infections were defined as testing positive in a microbiological test for other organisms during an RVI infection episode.

2.4. Outcome

PFT was performed using a spirometry system (Platinum Elite from MGC Diagnostics Corporation). Data acquisition was conducted through the Breeze software Suite (version 7.1). All spirometry indices were obtained by a registered respiratory therapist at SJCRH.

According to Current European respiratory society (ERS) technical standards, fixed percent predicted forced expiratory volume (FEV1) and percent predicted forced vital capacity (FVC) have been replaced by z‐scores and lower limit of normal (LLN) to more accurately detect abnormal lung function. Therefore, we utilized the global lung function initiative (GLI) online calculator to derive z‐scores and LLN [5].

Z‐scores represent the number of standard deviations measured value deviates from the population means. A z‐score of −1.645, corresponding to the LLN, has traditionally been defined as AFO (i.e., where the FEV1/FVC ratio is less than the LLN) [5].

Percent predicted values were converted to measure parameters, which were then used to calculate z‐scores and LLN using GLI calculator [6]. An excel data set containing the required and optional input variables (age, height, gender, ethnicity, baseline FEV1, year 1 FEV1, baseline FVC and year 1 FVC) was uploaded for analysis. Race‐neutral z‐score and LLN were obtained automatically. Any z‐score below −1.645 was classified as AFO [5].

While this study followed the methodology described by Erard et al. [4], we made specific modifications to enhance the statistical power of post‐transplant PFT data. The timeframe for PFT collection was expanded to include any assessment performed within year +60 days posttransplant to account for missing FEV1 and FVC data (n = 98).

2.5. Statistical Analysis

Statistical analysis was performed using Statistical Analysis System (SAS) version 9.4 (SAS Institute, Cary, NC, USA). Patient characteristics, treatment variables, and outcomes were summarized and compared between participants and nonparticipants. The outcome AFO was categorized as a binary variable. Pearson's chi‐square or Fisher's exact test was used to compare categorical variables. Continuous data were compared using the Wilcoxon rank‐sum test at a significance level of 0.05. The median and range were calculated. Variables with insufficient data to calculate statistics were mentioned with their values. Univariate logistic regression analysis was used to examine the potential predictors of AFO. Multivariate logistic regression analysis was performed to assess the impact of GVHD, FEV1/FVC ratio before transplant, and the different community respiratory viruses among those with AFO. In addition, we used a multivariate model to assess the effect of RVIs in different locations among those with AFO An additional analysis was performed to determine the association and impact of cases with AFO with the URTI and LRTI after adjusting for age at allo‐HCT, baseline FEV1/FVC ratio, and GVHD category. These covariates were found to be significant risk factors for AFO in the literature [4]. Panels were generated using GraphPad Prism 10.lnk.

3. Results

3.1. Cohort Characteristics

Among 397 patients, 54 (14%) developed RVI within the first 100 days posttransplant. Of these, 11% exhibit AFO by one‐year post‐transplant. Oxygen supplementation was required in 78% of patients with RVI than in those without RVI (43%, p < 0.05). In addition, 14% of those with RVI required ICU admission (Table 1).

TABLE 1.

Epidemiologic characteristics of the study cohort —Descriptive statistics.

	All patients (n = 397)	RVI (n = 54)	No RVI (n = 343)	p a
Demographic variables
Age at allo‐HCT, in years
Median [Range]	13 [5.1–25.2]	13 [5.3–20.3]	12.9 [5.1–25.2]	0.61
Gender (n, %)
Female	181 (46)	25 (47)	156 (45)	0.79
Male	216 (54)	29 (53)	187 (55)
Race (n, %)
Black	74 (19)	11 (19)	63 (18)	0.98
White	286 (72)	39 (74)	247 (72)
Other	37 (9)	4 (7)	33 (10)
Transplant variables
Donor type (n, %)
Matched Sibling Donor	92 (23)	11 (20)	81 (24)	0.75
Unrelated	141 (36)	22 (41)	119 (35)
Haplo	163 (41)	21 (39)	142 (41)
Others	1 (0.3)	0 (0)	1 (0.2)
Cell source (n, %)
PBSC	181 (46)	26 (50)	155 (46)	0.48
BMSC	210 (53)	28 (50)	182 (52)
Cord	6 (2)	0 (0)	6 (2)
GvHD (n, %)
Acute and chronic GVHD	7 (2)	0 (0)	7 (2)	0.09
Only Acute	22 (6)	0 (0)	22 (7)
Only Chronic	6 (2)	0 (0)	6 (2)
None	362 (91)	54 (100)	308 (90)
Pulmonary test variables
Baseline FEV1/FVC ratio b
Less than LLN	35(9)	2(4)	33(10)	0.15
More than LLN	362(91)	52(96)	310(90)
Median [Range]	0.88 [0.5–1.09]	0.88 [0.76–1.01]	0.88 [0.5–1.09]	0.27
Year 1 FEV1/FVC ratio c
Less than LLN	46 (12)	6 (11)	40 (12)	0.90
More than LLN	351 (88)	48 (89)	303 (88)
Median [Range]	0.88 [0.33–1.08]	0.89 [0.67–0.99]	0.88 [0.33–1.08]	0.40
AFO
Yes	47 (12)	6 (11)	41 (12)	0.86
No	350 (88)	48 (89)	302 (88)
Infection variables
Oxygen required, n (%)
Yes	188 (47)	43 (78)	145 (43)	< 0.0001
No	209 (53)	11 (22)	198 (57)
Oxygen required in days
Median [Range]	6 [1–69]	7 [1–18]	5 [1–69]	0.24
Mechanical ventilation required, n (%)
	(n = 188)	(n = 43)	(n = 145)
Yes	54 (29)	6 (14)	15 (10)	0.03
No	134 (71)	37 (86)	130 (90)
Mechanical ventilation required in days
Median [Range]	9 [1–47]	18.5 [2–24]	8 [1–47]	0.71

Abbreviations: AFO, airflow obstruction; allo‐HCT, allogenic hemopoietic cellular transplant; BMSC, bone marrow stem cells; FEV1/FVC ratio, forced expiratory volume in 1 s (FEV1)/forced vital capacity (FVC) ratio; GVHD, graft versus host disease; LLN, lower limit of normal; LRTI, lower respiratory tract infection; Oxygen, oxygen requirement started within 30 days of infection and extension for any current oxygen; PBSC, peripheral blood stem cells; RVI, respiratory viral infection; URTI, upper respiratory tract infection.

^a Statistical analyses were performed using Pearson's chi‐square or Fisher's exact test or Wilcoxon rank‐sum test at a significance level of 0.05. Bold type indicates statistical significance.

^b Baseline tests were considered between Day −1 to Day −101 of allo‐HCT.

^c Year 1 test were considered within 425 days post allo‐HCT.

A total of 60 episodes of RVIs were identified among 54 patients (Table S1). The majority (53, 88%) of episodes were URTI, while more than one quarter (16, 27%) progressed to LRTI. Information on site of infection was unavailable for one case. Human rhinovirus/enterovirus (HRV) was the most frequently detected pathogen, followed by adenovirus and RSV. Co‐infections were observed in 14 (23%) episodes. More than half of the viral infection episodes involve Class I viruses [1]. Forty‐seven (12%) cases were identified as having AFO, defined as per current pulmonary function guidelines [5].

Table S3 summarizes the demographic and transplant characteristics of the allo‐HCT cohort with AFO. Among the 47 patients with AFO post allo‐HCT, the median age was 13.7 years (IQR 5–21). The majority of these patients were white (70%). Forty percent of these patients had an unrelated donor, and 53% underwent a myeloablative conditioning regimen. The most frequent stem cell source was bone marrow, accounting for 53% (25 cases). Eight cases (2%) died within 1 year of transplant, and none of the cases experienced RVI. Four patients (9%) experienced chronic GVHD, two classified as severe (Table S3). Notably, both patients had lung involvement as the site of GVHD, with chest computed tomography (CT) findings confirming chronic GVHD (Table S2). The median baseline FEV1/FVC ratio in this cohort was 0.88 (IQR 0.50–1.09), with the majority (91%) of them “More than LLN”, reflecting normal pulmonary function at baseline. Six patients (12%) with AFO had RVI. Among these, five had a Class I virus and one had a Class II virus. About half of these patients (23, 49%) required supplemental oxygen for an average of 7 days [IQR 1–33]. Additionally, three patients required mechanical ventilation for 2 days. URTI was noted in six cases (13%), and the time from transplant to URTI was significantly (p = 0.04) shorter in those with AFO (median = 11.9 days) when compared to those without AFO (median = 33.7 days) (Table 2, Figure S2).

TABLE 2.

Comparison of clinical and infection characteristics between AFO and no AFO cases.

	All patients: N = 397, n (%)	AFO: N = 47, n (%)	No AFO: N = 350, n (%)	p a
Baseline FEV1/FVC ratio b (in liters)
Less than LLN	35 (9)	15 (32)	20 (6)	< 0.00001
More than LLN	362 (91)	32 (68)	330 (94)
Median [Range]	0.88 [0.50–1.09]	0.81 [0.69–1.02]	0.89 [0.50–1.09]	< 0.0001
Year 1 FEV1/FVC ratio c (in liters)
Less than LLN	46 (12)	46 (98)	0 (0)	< 0.00001
More than LLN	351 (88)	1 (2)	350 (100)
Median [Range]	0.90 [0.33–1.08]	0.70 [0.33–0.80]	0.89 [0.75–1.08]	< 0.0001
Type of viruses based on virulence
No RVI	343 (87)	41 (87)	302 (86)	0.24
Class I virus	33 (8)	5 (11)	28 (8)
Class II virus	21 (5)	1 (2)	20 (6)
RVI location, n (%)
Episode 1
No RVI d	345 (87)	41 (87)	304 (86)	0.16
Baseline LRTI	14 (1)	4 (2)	10 (2)
Baseline URTI	47 (12)	5 (11)	42 (13)
Episode 2
	(n = 54)	(n = 2)	(n = 52)
No subsequent Episode	46 (87)	0 (0)	46 (87)	NA
Baseline LRTI	2 (2)	1 (0)	1 (2)
Baseline URTI	6 (11)	1 (100)	5 (11)
Time to URTI from transplant (in days)
	(n = 53)	(n = 6)	(n = 47)	0.04
Median [Range]	28.46 [2–97]	11.9 [10–13]	33.68 [2–97]
Oxygen
Yes	188 (47)	23 (49)	165 (47)	0.81
No	209 (53)	24 (51)	185 (53)
Oxygen days
Median [Range]	5 [1–87]	7 [1–33]	9.1 [1–87]	0.39
Mechanical ventilation
Yes	21 (11)	3 (6)	18 (5)	0.71
No	167 (89)	44 (94)	332 (95)
Mechanical ventilation days
Median [Range]	11.3 [1–47]	2 [2–23]	2.5 [1–47]	0.60
Death within one year of HCT
Yes	8 (2)	2 (4)	6 (2)	0.24
No	389 (98)	45 (96)	344 (98)

Abbreviations: AFO, airflow obstruction; FEV1/FVC ratio, forced expiratory volume in 1 s (FEV1)/forced vital capacity (FVC) ratio; LLN, lower limit of normal; RVI, respiratory viral infection; Oxygen, oxygen requirement started within 30 days of infection and extension for any current oxygen; URTI, upper respiratory tract infection; LRTI, lower respiratory tract infection; NA, not applicable.

^a Statistical analyses were performed using Pearson's chi‐square or Fisher's exact test or Wilcoxon rank‐sum test at a significance level of 0.05 to compare AFO cases with those who did not experience AFO post‐transplant. Bold type indicates statistical significance.

^b Baseline tests were considered between Day −1 to Day −101 of allo‐HCT.

^c Year 1 test were considered within 425 days post allo‐HCT.

^d One patient was classified as having no RTI because physician notes were unavailable.

Risk factors were analyzed for their potential association with increased risk of AFO using univariate logistic regression: age, gender at birth, transplant donor, cell source, GVHD, baseline FEV₁/FVC ratio, baseline RVI location, oxygen requirement, mechanical ventilation requirement, antiviral treatment, Class I and Class II virus. These factors were selected based on previous studies [4, 7]. Among those with AFO, only baseline FEV1/FVC ratio was found to be a significant predictor (OR = 0.13, 95% CI [0.06–0.28]). The remaining variables were not significantly associated with AFO (Table S4).

Moreover, similar results were found when multivariate logistic regression was used to determine significant predictors responsible for AFO. Baseline FEV1/FVC ratio was found to be a significant predictor (OR = 0.30, 95% CI [0.12–0.79]) after adjusting for age, GVHD, and RVI (Table S5).

When compared the impact of AFO between those exposed to Class II virus and Class I virus [3], there was no difference in their impact (Table S6). Furthermore, our model did not find any significant association between risk of AFO and LRTI after adjusting for potential confounders [4], age at allo‐HCT, baseline FEV1/FVC ratio, and GVHD categories that have been found to be risk factors for AFO in previous literature (Table S7, Figures 1 and 2).

FIGURE 1.

Risk factors associated with lung function among patients with RVI Lung function z‐scores are compared between patients with and without AFO across transplant‐ and infection‐related variables transplant and RVI related risk factors. (A) Preparative regimen: At α = 0.05, AFO was associated with significantly lower mean z‐scores independent of regimen type. (B) Donor type: A similar association was observed across donor sources, with no evidence of effect modification. (C) Timing of viral infection post‐transplant and (D) virus classification were not associated with differences in z‐scores by AFO status. (E) Respiratory viral infection (RVI): Among patients without RVI, z‐scores did not differ by AFO status. In contrast, among those with RVI, both URTI and LRTI in the presence of AFO were associated with significantly lower z‐scores compared with RVI cases without AFO. “+” denotes the group mean. * stands for p < 0.05.

FIGURE 2.

Longitudinal changes in lung function from baseline to year 1. Changes in lung function z‐scores from baseline to year 1 are shown across transplant‐ and infection‐related variables, transplant and RVI related risk factors. (A) Preparative regimen: Mean z‐scores were significantly lower at year 1 compared with baseline, independent of regimen type. (B) Donor type: Declines in z‐scores were observed across all donor sources. (C) Timing of RVI: RVI occurring > 30 days posttransplant was associated with lower year 1 z‐scores, whereas RVI ≤ 30 days posttransplant was not, suggesting a greater contribution of late RVI to lung function decline in this cohort. (D) Virus classification: Class I viruses were associated with lower year 1 z‐scores. (E) Site of viral infection: Declines from baseline were observed regardless of RVI presence or infection site. “+” denotes the group mean. * stands for p < 0.05.

While Figure 2 presents descriptive trends, we extend this by conducting an unadjusted linear regression analysis (Table S8). None of the potential predictors showed statistical significance, suggesting the observed effect may be due to chance.

4. Discussion

This study evaluated the impact of RVI in pulmonary outcomes among pediatric allo‐HCT recipients with particular focus on AFO [8].

The overall prevalence of AFO in our cohort was lower than that reported in a previous study (12% vs. 26%) [9]. Though it remains higher than the 2%–10% trend observed in some earlier studies [10]. Unlike previous studies involving pediatric/adult populations, we did not observe a significant association between community acquired respiratory viruses and allo‐HCT‐related AFO [4, 9]. Consistent with earlier findings [11, 12, 13], HRV was the most frequently identified pathogen in our cohort (Table S10). However, in contrast to a prior study [11], linking RSV and human metapneumovirus infections to progression towards LRTI, we observed HRV to be the predominant virus associated with LRTI. Although our multivariate models did not identify any statistically significant predictor, Ogimi C et al. reported that allo‐HCT recipients with multiple viruses detected had a comparable risk of LRTI progression to those infected with HRV, PIV, influenza, or adenovirus [11].

A noteworthy observation in our study was the association between baseline PFT and timing of infection onset. Patients with preexisting AFO experienced URTI earlier after transplant than those without AFO (Figure S2 and Graph S1), suggesting a potential link between baseline lung function and susceptibility to RVIs. This finding supports the hypothesis that the compromised baseline PFT may weaken airway defense mechanism, thereby increasing the vulnerability to infection. Consistent with this, a prior study found LRTI to be a strong predictor of pulmonary impairment among adults allo‐HCT patients age 19–77 years [14]. Together, these results underscore the importance of vigilant pulmonary monitoring and early intervention in patients with preexisting AFO to mitigate secondary infection risks.

Analysis of potential predictors for AFO (Table S5) revealed no significant difference in infection severity or outcomes among Class I and Class II. Similarly, the multivariate model (Table S7) did not find any significant predictor of AFO, suggesting the presence of RVI alone does not account for posttransplant AFO in this cohort.

This lack of statistical significance may be attributed to sample size limitations, which could have affected the power of the analysis and the ability to detect true associations. This also explains the reason behind not finding many risk predictors despite finding similarity in prevalence with other studies [11].

A key strength of this study is the application of “race‐neutral z‐score classification” for diagnoses of AFO which is a significant advance [5], a method recently endorsed in pulmonary function interpretation, particularly in a pediatric population. Nonetheless, several limitations must be acknowledged. First, the relatively small cohort size limited our ability to detect modest association or stratify AFO by severity (mild, moderate, and severe) [6]. Second, preexisting respiratory or systemic conditions that may contribute to airflow decline were not systematically evaluated, potentially leading to AFO misclassification. This study did not assess LRTI episodes caused by common RVI prior to myeloablative allo‐HCT, which have been linked to increased mortality and hospitalization risk [15].

As detailed in the method section, modifications were made to account for missing PFT data which may have introduced misclassification bias. Selection bias is also possible as cases lacking PFT data were excluded. Lack of consistency in the dates of PFT is a possible limitation; however, it does not affect the overall conclusion. Respiratory viral testing was performed based on clinical indication. Patients underwent multiplex PCR panel testing when they developed symptoms suggestive of RVI, including upper respiratory tract symptoms, fever with or without localizing findings, or lower respiratory tract symptoms. This strategy may have led to underestimation of subclinical infections and carries the inherent risk of false‐negative results associated with standard diagnostic assays. Another important consideration is the inclusion of patients who had baseline AFO at the start of the study (Table 2). Including these patients may introduce confounding variables when identifying risk factors for developing AFO. However, after excluding these patients from the analysis, no significant predictors for subsequent development of AFO were identified (Table S11). Additionally, because this cohort included patients from the pre‐PCR diagnostic era, some RVIs may have been underdiagnosed or misclassified. However, when we compared the descriptive statistics between the full cohort and PCR era subsets (2014 onwards), no significant difference was observed (Table S9). Another limitation is the absence of temporal adjustment of HSC era. Given the substantial improvement in transplant practice over the past two decades, unaccounted era effects could have influenced outcomes. Finally, we excluded non‐RVI viral infections such as cytomegalovirus (CMV), which is known to be associated with chronic GvHD. Importantly, none of the excluded CMV cases demonstrated chronic GvHD in our data set.

In conclusion, RVIs occurring within the 100 days after allo‐HCT were not significantly associated with long term pulmonary decline in this pediatric cohort. Our finding suggests that the timing of infection—particularly early post‐transplant events—may play a more critical role than viral type or host factors. Larger, prospective studies are warranted to validate these findings and to better delineate long term pulmonary effects of RVIs in pediatric allo‐HCT recipients.

Author Contributions

D.R.H. and A.Y.S. conceptualized and designed this study. D.R.H., A.Y.S., and S.A.P. drafted the initial manuscript and reviewed and revised the manuscript. J.A.F. and K.J.A. coordinated and supervised data extraction. M.P., P.M., A.C., A.D., A.G., and K.J.A. collected the data. J.A.F. and S.A.P. carried out the initial data analysis, followed by S.A.P. conducting the statistical analysis. S.A.P., A.Y.S., J.A.F., R.H.D., M.P., P.M., A.C., A.D., A.G., K.J.A., R.T.H., Y.L., D.K., S.S., G.M.M., B.M.T., and D.R.H. reviewed and revised the manuscript for important intellectual content and approved the final manuscript as submitted.

Funding

This study was funded through American Lebanese Syrian Associated Charities (ALSAC). No additional external funding was received for this study.

Disclosure

Gabriela Maron is an advisor/consultant and provides grant/research support to SymBio Pharamaceuticals. Randall Hayden serves as a member of the advisory board for Roche Diagnostics, T2 Diagnostic, and Cepheid. No generative artificial intelligence (AI) tools were used in the creation of this manuscript.

Supporting information

Figure S1: Cohort selection. This flow chart illustrates potential eligible cases extracted (n = 438), following which 98 cases were excluded based on inclusion and exclusion criteria. 397 total cases were identified as meeting all criteria and were included in this study. Created in Lucidchart.com.

Table S1: Epidemiological characteristics—cases with RVI episodes, n(%). *Number of LRTI and URTI are counted distinctly for each episode and each case; Class I virus: respiratory syncytial virus, influenzas A and B human metapneumovirus, parainfluenzas virus, adenovirus; Class II: human rhinovirus, coronavirus, human bocavirus; mild: no need for supplementary oxygen; moderate: need for supplementary oxygen; severe: need for mechanical ventilation.

Table S2: GVHD organ involved per GVHD type.

Table S3: Demographic and transplant characteristics of allo‐HCT cases with AFO—Descriptive statistics.

Figure S2: Comparison of Post‐transplant URTI among AFO and non‐AFO cases (in days).

Table S4: Risk factors for Airflow obstruction—Univariate analysis; numbers in brackets display 95% CI. *Statistical analysis was performed using logistic regression. Bold type indicates statistical significance.

Table S5: Risk factors for Airflow obstruction—Univariate and multivariate analysis; numbers in brackets display 95% CI. *Statistical analysis was performed using logistic regression. Bold type indicates statistical significance.

Table S6: Impact on AFO post‐transplant due to different viruses and their location; numbers in brackets display 95% CI. *Statistical analysis was performed using logistic regression.

Table S7: Determining risk of AFO associated with infection location—after adjusting for age at allo‐HCT, baseline FEV1/FVC ratio and GVHD, a multivariate analysis. Numbers in brackets display 95% CI. *Statistical analysis was performed using logistic regression. Bold type indicates statistical significance.

Table S8: Exploring predictors of AFO using unadjusted linear regression models.

Table S9: Characteristics of PCR diagnosed RVI among HCT patients. *RVI detected within 100 days of transplant.

Table S10: Distribution of RVIs and anti‐viral treatment.

Table S11: Comparison of clinical and infection characteristics between AFO and no AFO (Excluding cases with FEV1/FVC ratio less than LLN).

Graph S1: Modeling days post‐transplant for the RVIs versus Year 1 AFO.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.