Transplant related mortality(TRM) associated with allogeneic stem cell transplantation (HSCT) has necessitated risk-stratification that allows practitioners to make informed choices regarding patient selection for HSCT. Although pulmonary function testing (PFT) is obtained prior to HSCT at most institutions, and incorporated into common comorbidity indices (e.g. HCT-CI) there is no standard protocol to determine suitability for HSCT specific to lung disease. Also, there is a lack of data regarding the impact of tobacco use prior to HSCT. Recently, the association between smoking and kidney disease has gained attention. Tobacco-mediated kidney disease may be prognostic for renal failure post-HSCT. We studied the impact of tobacco use and lung function prior to HSCT on 30- and 100-day mortality as well as the frequency of post-HSCT respiratory and renal failure.


We reviewed consecutive patients undergoing allogeneic HSCT at our institution from 2004 -2014 for acute myeloid leukemia (AML), acute lymphoid leukemia (ALL), multiple myeloma (MM), non-Hodgkin lymphoma (NHL) and myelodysplastic syndrome (MDS). Smoking was categorized as never (0 pack years), low dose (<10 pack years) and high dose (≥10 pack years). FEV/FVC was categorized <0.7, 0.7-0.8 and ≥0.8 of predicted. Association of categorical variables with outcomes and potential confounding variables was assessed using Chi-square tests. Adjusted estimates of association were calculated using multivariable logistic regression. Outcomes were 30 day mortality, 100 day mortality, ventilator and dialysis dependence.


Of 373 subjects who underwent allogeneic HSCT, 315 (84%) had complete smoking histories (42% female, 58% male). AML (44.1%) accounted for the most frequent diagnosis followed by NHL (22.2%), ALL (14.3%), MDS (12.1%) and MM (7.3%). The 30 day mortality was 6.4%, 100 day mortality was 16.5%. 23.5% of patients required mechanical ventilation while 12.4% required dialysis. 46.7% of patients were smokers; 31.8% high dose smokers, 14.9% low dose smokers. There was an 11% increase in ventilator-dependency among high-dose smokers compared to none or low-dose (p=0.032), with a trend toward higher incidence of 100-day mortality with high dose smokers (31%) compared with low dose smokers (23.4%) and never-smokers (19.1%), (p=0.083). High dose smokers were more likely to have reduced pulmonary function with FEV/FVC < 70% (p=0.0004). Patients with FEV/FVC < 70% had a 14% increase in 100-day mortality with an almost 20% increased chance of ventilator dependency (p=0.022) and 17.6% increased incidence of dialysis dependency (p=0.032). DLCO did not correlate to any endpoint with statistical significance. Patients with MDS had the highest proportion of high dose smokers (47.4%; p=0.0023) and highest percentage of patients with FEV/FVC < 70% (18.4%, p=0.0037). Patients with MDS also had the highest 30-day mortality (18.4%, p=0.023), 100-day mortality (31.6%, p=0.0012), ventilator dependency (47.4%, p=0.0014) and dialysis dependency (31.6%, p=0.0036).


Our results differ from prior studies looking at the utility of FEV1 and DLCO in pre-HSCT PFTs as predictive markers; neither had prognostic value. We found that high dose smoking history and reduced FEV1/FVC were significantly associated with post-HSCT ventilator dependence, dialysis dependence and mortality. These results underscore the need for future studies that examine the relationship between smoking, lung function and renal failure post-HSCT. Our study did not demonstrate that low dose smoking predicts worse outcomes post-HSCT, but future studies may help provide patients with more accurate pre-HSCT counseling. Patients with MDS were the largest population of high dose-smokers and had the highest proportion of 100-day mortality, ventilator and dialysis dependency.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.