Lung Function Before and After Pediatric Allogeneic Hematopoietic Stem Cell Transplantation: A Predictive Role for DL_COa/VA

Abstract

Background

Pre-allogeneic hematopoietic stem cell transplantation (aHSCT) and post-aHSCT lung function of 41 eligible patients at Riley Hospital for Children were assessed to identify risk factors for post-aHSCT morbidity and mortality.

Observations

One year post-aHSCT pulmonary function tests were significantly lower compared with baseline. These findings recovered at 2 years post-aHSCT. Refractory disease before aHSCT correlated with lower pulmonary function tests after aHSCT. Graft-versus-host disease was significantly associated with higher post-aHSCT residual volume. Importantly, low pre-aHSCT carbon monoxide diffusing capacity adjusted for hemoglobin and alveolar volume was predictive of death.

Conclusions

Among survivors, lung function improves over time after pediatric aHSCT. Measurement of carbon monoxide diffusing capacity adjusted for hemoglobin and alveolar volume before pediatric aHSCT should be further investigated as a predictor of pulmonary dysfunction and mortality.

Allogeneic hematopoietic stem cell transplantation (aHSCT) provides life-saving therapy for a variety of malignant and nonmalignant pediatric diseases. Despite improvements in supportive care, pulmonary complications account for meaningful morbidity and mortality after aHSCT. The incidence of post-aHSCT pulmonary complications in pediatric patients has been reported to be 25% and was associated with higher mortality (65% with pulmonary complications vs. 44% without).¹ Many studies in adults have demonstrated that abnormal pulmonary function tests (PFTs) may be predictive of post-aHSCT mortality and may identify patients at-risk for noninfectious pulmonary complications post-aHSCT.^2–4 Few studies have focused on pediatric aHSCT recipients, and these studies have reported conflicting results on predictors of post-aHSCT morbidity and mortality.

We retrospectively reviewed pre-aHSCT and post-aHSCT PFTs on 41 patients transplanted at our institution over a 5-year period to assess the prevalence of abnormal lung function and to identify PFT variables that may predict post-aHSCT morbidity and mortality. We examined relationships linking lung function abnormalities with age, disease status, earlier pulmonary-toxic chemotherapies, aHSCT conditioning regimen, graft-versus-host disease (GVHD), cytomegalovirus (CMV) immune status, and post-aHSCT death. Importantly, we describe the utility of adjusting carbon monoxide diffusing capacity (DL_CO) for both hemoglobin (DL_COa) and alveolar volume (DL_COa/VA), and its ability to predict post-aHSCT mortality in pediatric patients.

MATERIALS AND METHODS

Patient Characteristics

We retrospectively reviewed PFT and clinical data from patients who underwent aHSCT at Riley Hospital for Children between 2001 and 2006. Institutional review board approval was obtained. A total of 41 of 46 patients from 5 to 19 years of age were eligible for analysis. Five patients were excluded because they were missing baseline PFT data or had nonstandard follow-up times. Thirty-one of 41 recipients had 1-year follow-up PFTs. One year post-aHSCT, 1 patient had transferred care, 4 patients had died (2 of 4 with 1 y post-aHSCT PFTs before death), and 7 patients were alive but without PFTs performed post-aHSCT. Seventeen of 41 recipients had 1 and 2 year follow-up PFTs. At 2 years post-aHSCT, 7 additional patients had died and 13 patients did not follow-up for 2 year post-aHSCT PFTs. Demographic and clinical characteristics are shown in Table 1.

TABLE 1.

Demographic and Clinical Characteristics of 41 aHSCT Recipients

	N (%)
Male	26 (63.4%)
White	37 (90.2%)
Age (y, mean ± SD)	12.5 ± 4.3
Age ≤8 y	10 (24.4%)
Time from diagnosis to aHSCT (d)	828 ± 990
Median time to aHSCT (d)	522
Diagnosis
ALL	17 (41.5%)
AML	12 (29.3%)
Hodgkin lymphoma	1 (2.4%)
Non-Hodgkin lymphoma	3 (7.3%)
Nonmalignant disease*	8 (19.4%)
Pretransplant CMV status
IgG positive	13 (31.7%)
IgM negative†	2 (4.9%)
IgG negative	26 (63.4%)
Earlier pulmonary-toxic chemotherapy	17 (41.4%)
Bleomycin	1 (2.4%)
Methotrexate	14 (34.1%)
Fludarabine	2 (4.9%)
Disease status‡
CR1	10 (24.4%)
CR2	15 (36.6%)
CR3	4 (9.8%)
Refractory	4 (9.8%)
Nonmalignant disease	8 (19.4%)
Conditioning
Total body irradiation	25 (61.0%)
Cyclophosphamide	34 (82.9%)
Busulfan	10 (24.4%)
Fludarabine	4 (9.8%)
aHSCT product
Bone marrow	23 (56.1%)
Umbilical cord blood	14 (34.1%)
PBPC	4 (9.8%)
Donor-recipient HLA match
6/6	27 (65.8%)
5/6	4 (9.8%)
4/6	10 (24.4%)
Related donor	22 (53.7%)
GVHD prophylaxis
Cyclosporine + methylprednisolone	7 (17.2%)
Cyclosporine + methotrexate	31 (75.6%)
Cyclosporine + mycophenolate mofetil	1 (2.4%)
Cyclosporine (single agent)	1 (2.4%)
Tacrolimus + methotrexate	1 (2.4%)
GVHD	22 (53.7%)
Acute	13 (31.7%)
Chronic	4 (9.8%)
Both	5 (12.2%)
Hepatic sinusoidal obstruction syndrome	1 (2.4%)
Post-aHSCT deaths (overall mortality)	11 (26.9%)

^*Nonmalignant disease: severe aplastic anemia (n = 5), Kostmann syndrome (n = 1), Fanconi anemia (n = 1), and hemophagocytic lymphohistiocytosis-HLH (n = 1).

^†Only CMV IgM obtained pre-aHSCT (n = 2).

^‡CR1: first complete remission, CR2: second complete remission, CR3: third complete remission.

aHSCT indicates allogeneic hematopoietic stem cell transplantation; ALL, acute lymphoblastic leukemia; AML, acute myeloid leukemia; CMV, cytomegalovirus; CR, complete remission; GVHD, graft-versus-host disease; HLA, human leukocyte antigen; PBPC, peripheral blood progenitor cells.

Table 1 shows the number of patients who received agents during conditioning that could cause acute or delayed lung injury. Other conditioning agents (not shown) used during conditioning included antithymocyte globulin, etoposide, bis-chloroethylnitrosourea, and melphalan. Details of GVHD prophylaxis are summarized in Table 1. GVHD was documented as “no” or “yes” (biopsy-proven), and was subcategorized as acute, chronic, or having evidence of both.

Pulmonary Function Testing

All PFTs were performed in the Pediatric Pulmonary Function Laboratory at Riley Hospital for Children. All patients had spirometry, lung volume by plethysmography, and DL_CO measured. VA was measured using methane gas. DL_CO was measured by the single breath-hold maneuver⁵ using Vmax Encore (VIASYS Healthcare, Yorba Linda, CA), and was adjusted for hemoglobin (DL_COa) and VA (DL_COa/VA). PFT data were reviewed by a single Pediatric Pulmonologist (Y.J.K.) to ensure reliability. Testing, equipment, and interpretation of PFTs followed the recommendations of the American Thoracic Society and European Respiratory Society.⁵ Lung function variables were expressed as percent-predicted normal values for age, sex, and height-matched controls. Forced expiratory volume in 1 second (FEV₁), forced vital capacity (FVC), FEV₁/FVC, forced expiratory flow at midexpiration (FEF_25%-75%), total lung capacity (TLC), DL_COa, and DL_COa/VA <80% of predicted values were considered abnormal. Residual volume (RV) >120% was considered abnormal and suggestive of air trapping.

Statistical Methods

Statistical analyses were performed using SPSS 16 (SPSS Inc., Chicago, IL). P values <0.05 were considered significant. Logistic regression with generalized estimating equations was used to model the association of abnormal PFTs at baseline, 1 and 2 years post-aHSCT. Repeat measures models and logistic regression were used to test associations between pre-aHSCT and post-aHSCT PFT variables and various clinical characteristics. Logistic regression models were used to model the association of abnormal baseline PFTs with death.

RESULTS

Patient Characteristics and Pulmonary Function Testing

The clinical characteristics of the 41 aHSCT patients are summarized in Table 1. Seventeen of 41 patients (41.4%) received pulmonary-toxic chemotherapy before aHSCT. No patients had lung irradiation before aHSCT. Baseline PFT data are summarized in Tables 2 and 3. Absolute numbers and percentages of patients with abnormal PFT parameters at baseline are summarized in Table 3. Twenty-nine of the 41 patients (70.7%) studied had 1 or more abnormal PFT parameters at baseline. Nine of 41 patients (21.9%) had preaHSCT obstructive lung patterns evidenced by an abnormally low FEV₁, FEV₁/FVC, FEF_25%-75%, or RV. Fourteen patients (34.1%) presented to aHSCT with abnormally low DL_COa and DL_COa/VA. We found no significant association between individual before pulmonary-toxic chemotherapy and baseline PFTs.

TABLE 2.

PFT Values at Pre-aHSCT, 1 Year and 2 Year Post-aHSCT

PFT Parameter	Years From aHSCT*	Mean (%)	SD (± %)	From Baseline to 1 y Post-aHSCT(P)	From Baseline to 2 y Post-aHSCT (P)	Change Between 1 and 2 y Post-aHSCT (P)
FEV1	2	88.1	18.0	0.007	0.405	0.157
FEV1	1	83.3	17.5
FEV1	0	91.7	14.7
FVC	2	88.5	20.8	0.007	0.443	0.146
FVC	1	83.4	19.7
FVC	0	91.9	14.1
FEV1/FVC	2	87.9	7.59	0.393	0.895	0.370
FEV1/FVC	1	88.8	6.52
FEV1/FVC	0	88.0	6.91
FEF25%-75%	2	103	26.9	0.761	0.086	0.166
FEF25%-75%	1	97.3	25.6
FEF25%-75%	0	96.3	29.1
RV	2	95.6	22.3	0.142	0.041	0.575
RV	1	91.9	30.2
RV	0	82.4	31.7
TLC	2	92.4	16.3	0.013	0.523	0.186
TLC	1	88.7	17.6
TLC	0	94.1	14.3
DLCOa	2	76.9	13.7	0.002	0.212	0.258
DLCOa	1	72.0	18.9
DLCOa	0	82.2	14.9
DLCOa/VA	2	85.5	13.9	0.686	0.342	0.074
DLCOa/VA	1	81.8	16.0
DLCOa/VA	0	82.9	13.3

^*Time 0 = Pre-aHSCT (n = 41), Time 1 = Estimated 1 y post-aHSCT (n = 31), Time 2 = Estimated 2 y post-aHSCT (n = 17).

aHSCT indicates allogeneic hematopoietic stem cell transplantation; DL_COa, carbon monoxide diffusing capacity adjusted for hemoglobin; FEF_25%-75%, forced expiratory flow at midexpiration; FEV₁, Forced expiratory volume in 1 second; FVC, forced vital capacity; PFT, pulmonary function test; RV, residual volume; TLC, total lung capacity; VA, alveolar volume.

Statistically significant values are in bold.

TABLE 3.

Number and Percent of Patients With Abnormal PFTs at Each Time Point

PFT Parameter	Baseline* n (%)	1 y Post-aHSCT† n (%)	2 y Post-aHSCT‡ n (%)	Baseline to 1 y Post-aHSCT P	Baseline to 2 y Post-aHSCT P
FEV1	8 (19.5%)	11 (35.5%)	3 (17.6%)	0.013	0.774
FVC	8 (19.5%)	11 (35.5%)	5 (29.4%)	0.009	0.190
FEV1/FVC	3 (7.3%)	3 (9.7%)	2 (11.8%)	0.362	0.256
FEF25%-75%	13 (31.7%)	7 (22.6%)	4 (23.5%)	0.310	0.433
RV	4 (10.5%)	4 (13.3%)	3 (17.6%)	0.682	0.416
TLC	5 (13.2%)	8 (26.7%)	3 (17.6%)	0.075	0.883
DLCOa	14 (43.8%)	18 (72.0%)	10 (62.5%)	0.012	0.193
DLCOa/VA	14 (43.8%)	11 (44.0%)	5 (31.3%)	0.643	0.444

^*Total patients with baseline PFTs = 41.

^†Total patients with 1 y post-aHSCT PFTs = 31.

^‡Total patients with 2 y post-aHSCT PFTs = 17.

aHSCT indicates allogeneic hematopoietic stem cell transplantation; DL_COa, carbon monoxide diffusing capacity adjusted for hemoglobin; FEF_25%-75%, forced expiratory flow at midexpiration; FEV₁, Forced expiratory volume in 1 second; FVC, forced vital capacity; PFT, pulmonary function test; RV, residual volume; TLC, total lung capacity; VA, alveolar volume.

Thirty-one patients (75.6%) had follow-up PFTs at 11.9 ± 2.2 months post-aHSCT; of these, 17 had follow-up PFTs at a mean of 28.5 ± 5.9 months post-aHSCT. From baseline to first reassessment of lung function post-aHSCT, significantly lower FEV₁ (P = 0.007), FVC (P = 0.007), TLC (P = 0.013), but similar FEV₁/FVC values were observed (Table 2). Although lung volumes decreased over the first year post-aHSCT, mean PFT data were largely within the normal range. DL_COa significantly decreased between baseline and 1 year post-aHSCT (Tables 2 and 3), but when adjusted for VA, there was no significant decline in DL_COa/VA.

The only significant change in PFT parameters between baseline and 2 years post-aHSCT was an increase in RV (P = 0.041; Table 2); no other PFT data demonstrated evidence of obstructive lung disease. Patients who had refractory disease (n = 4) had significantly lower FEV₁, FVC, and TLC 1 year post-aHSCT, compared with patients with complete remission or nonmalignant disease.

Effect of Clinical Characteristics and Conditioning on Post-aHSCT Lung Function

Using repeated measures models, we studied the effect of pertinent pre-aHSCT and peri-aHSCT clinical data on 1 year post-aHSCT lung function (Table 4). Age (younger than or equal to 8 or older than 8 y) at transplant, preaHSCT CMV serology, human leukocyte antigen match, hematopoietic stem cell source (bone marrow, cord blood, or peripheral blood), and related versus unrelated donor status did not statistically impact post-aHSCT PFTs.

TABLE 4.

Effect of Age, CMV Status, Conditioning, and GVHD on 1 Year Post-aHSCT PFTs^*

PFT Parameter	Age ≤8 y	Pre-aHSCT CMV Status	TBI	BU/CY	GVHD†
FEV1	0.939	0.457	0.493	0.863	0.891
FVC	0.392	0.186	0.237	0.699	0.876
FEV1/FVC	0.688	0.094	0.215	0.531	0.509
FEF25%-75%	0.527	0.050	0.340	0.940	0.653
RV	0.846	0.264	0.688	0.486	0.014
TLC	0.309	0.054	0.795	0.620	0.180
DLCOa	0.887	0.850	0.601	0.157	0.417
DLCOa/VA	0.167	0.743	0.122	0.019	0.434

aHSCT indicates allogeneic hematopoietic stem cell transplantation; CMV, cytomegalovirus; CR, complete remission; DL_COa, carbon monoxide diffusing capacity adjusted for hemoglobin; FEF_25%-75%, forced expiratory flow at midexpiration; FEV₁, Forced expiratory volume in 1 second; FVC, forced vital capacity; GVHD, graft-versus-host disease; PFT, pulmonary function test; RV, residual volume; TBI, total body irradiation; VA, alveolar volume.

^*No significant effects at 2 y post-aHSCT (data not shown).

^†Presence of GVHD.

Statistically significant values are in bold.

There was no statistically significant effect of total body irradiation (TBI) conditioning on 1 year posttransplant PFTs (Table 4). Ten patients received busulfan/cyclophosphamide (BU/CY) conditioning and had significantly lower DL_COa/VA at 1 year posttransplant PFTs, compared with patients who received other conditioning (P = 0.019; Table 4). Four BU/CY patients developed infectious pulmonary complications posttransplant and 1 went on to develop pulmonary fibrosis (+12 mo).

GVHD and Post-aHSCT Pulmonary Function

GVHD occurred in 53.7% of patients (n = 22). Thirteen patients (31.7%) had acute GVHD (aGVHD), 4 (9.8%) developed chronic GVHD (cGVHD), and 5 (12.2%) had both aGVHD and cGVHD. Eighteen of 22 patients with GVHD had measurement of PFTs at 1 year post-aHSCT and 11 had PFTs at 2 years post-aHSCT. There were no statistically significant changes in PFT parameters related to GVHD at 1 and 2 years post-aHSCT, except for RV. The average RV in patients with GVHD at 1 year post-aHSCT was 91.5 ± 29.7%, compared with preaHSCT RV of 86.6 ± 44.7% (P = 0.014; Table 4), which suggests more air trapping and possible obstruction. At 2 years post-aHSCT, the mean RV increased to 105 ± 17.9%, but this increase was not statistically significant. The average RV for patients without GVHD (n = 11, 91.7 ± 33.8%) was similar to the RV for patients with GVHD at 1 year post-aHSCT. The RV for patients without GVHD decreased at 2 years post-aHSCT (n = 6, 73.8 ± 13.8%), although this was not statistically significant.

Of patients who had both aGVHD and cGVHD (n = 5), the RV at 1 year post-aHSCT was 119 ± 17%, compared with 65 ± 35.6% at baseline (P = 0.009). The RV at 2 years post-aHSCT (120 ± 13%) remained elevated similar to 1 year post-aHSCT, but was significantly increased compared with baseline (P = 0.009). This pattern suggests development of post-aHSCT air trapping and may suggest that development of pulmonary obstruction related to GVHD may be persistent and gradual over time.

Baseline Lung Function as a Predictor of Mortality

Overall post-aHSCT mortality was 26.9% (n = 11) at 2 years. Time to post-aHSCT death was 546.3 ± 374.5 days. Of the 4 patients (9.7%) who died within 1 year post-aHSCT, 2 died of disease progression, 1 died of disease progression and fungemia, and 1 succumbed to bacterial sepsis and aGVHD of the liver and skin. All 4 patients had abnormal DL_COa/VA before aHSCT, but otherwise normal PFTs. After 1 year post-aHSCT, 1 patient died of unknown cause (lost to follow-up) and 6 patients died of disease progression. Of these 7 patients, only 2 had abnormal PFTs aside from abnormal DL_COa/VA. In our cohort of pediatric aHSCT recipients, only an abnormal pre-aHSCT DL_COa/VA (univariate analysis) was found to be a statistically significant predictor of post-aHSCT mortality (P = 0.038). No other abnormalities in baseline PFT parameters, including DL_COa uncorrected for VA, were predictive of post-aHSCT mortality.

DISCUSSION

The most important observation of this study is that a low pre-aHSCT DL_COa/VA, but no other PFT abnormalities, was found to be highly predictive of mortality by 2 years. Among childhood aHSCT survivors, a significant decline in lung function was seen 1 year after transplantation, but recovered to a large degree by 2 years after transplantation. Although lung function significantly improved between 1 and 2 years post-aHSCT in our patient cohort (Table 2), post-aHSCT PFTs remained below baseline values, but not significantly. Overall, our data and the more recent literature support the finding that lung function declines post-aHSCT, but tends to recover over time.^6–9 Unlike other studies, which reported the experience of both autologous and allogeneic SCT recipients, we report only allogeneic SCT recipients over a 5-year period, as this group represents the vast majority of SCT-related morbidity and mortality. Two important limitations of our study are the relatively small sample size and retrospective nature of the study, which also made it difficult to correlate respiratory symptoms with pre/post-aHSCT PFTs. Although the literature on pediatric aHSCT recipients and lung function is growing, many of the referenced articles reflect patient sample sizes similar to our study.

Although abnormal PFTs tend to recover after pediatric aHSCT, a few centers have found residual pulmonary dysfunction in survivors at 2 to 10 years post-aHSCT.^6,10–12 At 1 year, our patients exhibited lower FEV₁, FVC, and TLC, whereas FEV₁/FVC, FEF_25%-75%, RV, and DL_COa/VA remained unchanged. Unlike earlier studies, which showed a greater decline in FVC than FEV₁,^10,12 our cohort demonstrated a proportional decline in FEV₁ and FVC supporting post-aHSCT lung obstruction, not restriction.

We did not find significant changes in pediatric PFTs in association with GVHD, similar to other published reports.^6,9,12 However, we found that patients who had aGVHD that progressed to cGVHD demonstrated a significant increase in 1 and 2 year post-aHSCT RV compared with baseline (Table 4). We confirmed that patients with refractory disease had significantly worse PFTs post-aHSCT compared with patients in complete remission or patients with nonmalignant disease.^6,11,12 We did not find any significant correlation between age, CMV status, or earlier pulmonary-toxic chemotherapy and pre/post-aHSCT lung function parameters, morbidity, or mortality. In a study of 90 pediatric aHSCT recipients, Lee et al¹³ reported on the respiratory failure and mortality 5 of 6 patients with suspected or confirmed CMV pneumonia. Although rates of post-aHSCT CMV antigenemia were similar between our 2 cohorts (24.8% vs. 24.4%), the Korean cohort had higher pre-aHSCT CMV seropositivity of 90% compared with 31.7% in our cohort, which often predicts risk of post-HSCT CMV reactivation and disease.¹³ In the reported deaths, only 1 of 6 cases of CMV pneumonia were confirmed, which suggests other factors may have contributed. Despite 1 confirmed and 1 suspected case of CMV pneumonia in our cohort, no deaths or respiratory failure ensued. Evaluation of pre-aHSCT PFTs in the Korean cohort may provide insight to the high degree of mortality related to CMV pneumonia.

We did not find a significant difference in post-aHSCT PFTs in the 25 patients (61%) who received TBI in our retrospective cohort, similar to earlier reports.^1,6,7,9–12 However, our data showed a significant decline in DL_COa/VA (Table 4) in patients who received BU/CY conditioning 1 year post-aHSCT. BU has been found to contribute to abnormal lung function after pediatric aHSCT.^6,10,11 One caveat to our finding is the lack of significant decline in DL_COa, compared with DL_COa/VA, which would suggest that decline in diffusing capacity in our cohort was dependent on loss of lung volume and VA.

In assessing baseline lung function as a predictor for mortality, patients with a pretransplant DL_COa/VA <80% had significantly decreased overall survival (P = 0.038). Perhaps surprisingly, we did not find a significant association of DL_COa with death. Pulmonologists debate whether or not DL_COa should be corrected for VA. DL_COa is affected by VA, which is in turn dependent on chest size. Thus, DL_COa in younger children will be lower than that in adults; a direct comparison without correction for VA could give misleading results. Furthermore, patients with restrictive lung disease tend to have underestimated DL_COa and overestimated DL_COa/VA, making such measures difficult to interpret and predict clinical importance of lung impairments.¹⁴ Some experts recommend comparing DL_COa/VA with TLC to optimize clinical interpretation in patients with known restrictive lung disease.¹⁵ It is important to note that without adjustment for VA, it is difficult to determine whether abnormally low DL_COa is due to a deficit in gas exchange, low lung volume, or a combination of both.¹⁶ Despite the difficulty in interpreting DL_COa, and DL_COa/VA, our findings support the concept of adjusting DL_COa for VA in aHSCT.^14–19

In a large adult cohort, reduced DL_COa and alveolararterial oxygen gradient were associated with increased 1 year post-HCST mortality.² Studies in adult⁴ and pediatric⁶ patients have demonstrated reductions in DL_CO over time, but these results have not correlated with posttransplant mortality. Parimon et al¹⁹ described the use of a predictive pre-aHSCT “lung function score,” based on pretransplant FEV₁ and DL_COa. Although these authors described predominantly an adult cohort, higher pre-aHSCT lung function scores derived from lower FEV₁ and DL_COa values correlated with increased risk of post-aHSCT death. Ginsberg et al⁷ reported a similar finding, but did not adjust DL_CO for hemoglobin or VA.

Hoffmeister et al¹⁸ demonstrated that 20% of evaluable pediatric aHSCT survivors had low DL_COa and DL_COa/VA, suggesting underlying pulmonary vascular disease, which may affect long-term lung function. These results stress the importance of using both DL_COa and DL_COa/VA in pre-aHSCT and post-aHSCT monitoring of lung function, and imply that DL_COa/VA may help identify a subgroup of pre-aHSCT patients at risk for pulmonary dysfunction and mortality. Although some pulmonary data has failed to demonstrate a difference in the ability of DL_COa or DL_COa/VA to predict abnormal gas exchange or desaturations with exertion, 1 study demonstrated that patients with known abnormal gas exchange were more likely to have lower DL_COa/VA than DL_COa.¹⁷ We argue that given our data and recent findings, DL_COa should be corrected for VA due to the intense treatment many aHSCT recipients receive before aHSCT, and its potential to better identify at-risk patients when used with other PFT parameters.

The American Society for Blood and Marrow Transplantation and the European Group for Blood and Marrow Transplantation have published joint guidelines on long-term monitoring of lung function after HSCT.²⁰ However, these guidelines do not reflect the potential importance in correcting DL_CO for hemoglobin and VA. Our data suggests that correction for VA may be important in using DL_CO as a pre-aHSCT risk factor, and that utilization of “lung function scores” may be misinterpreted without appropriate adjustment of DL_CO. A multicenter, prospective study of lung function in pediatric aHSCT patients could improve current clinical practice and patient outcomes, test previously described predictive tools,^3,7,8,19 and provide resolution to the conflicting literature on interpretation of DL_CO with associated correction in hemoglobin and VA.