Association between pre-HCT pulmonary metatranscriptome and post-HCT clinical outcomes
| Post-HCT outcomes . | Pre-HCT pulmonary metatranscriptome . | P . | |||
|---|---|---|---|---|---|
| Cluster 1 (n = 42) OP-heavy . | Cluster 2 (n = 45) Microbially depleted . | Cluster 3 (n = 84) OP-light . | Cluster 4 (n = 10) Virus enriched . | ||
| Lung injury | |||||
| Cumulative incidence, %* | 19.0 ± 0.4 | 44.4 ± 0.6 | 9.5 ± 0.1 | 30.0 ± 2.4 | <.001 |
| SHR (95% CI), univar | 2.1 (0.8-5.7), P = .13 | 5.6 (2.5-12.8), P < .001 | Ref | 3.5 (0.9-13.0), P = .06 | <.001 |
| SHR (95% CI), multivar | 1.8 (0.7-4.8), P = .23 | 3.8 (1.4-9.9), P = .007 | Ref | 2.6 (0.6-10.7), P = .19 | .052 |
| All-cause mortality | |||||
| Kaplan-Meier estimate, %† | 28.6 ± 7.0 | 37.8 ± 7.2 | 25.0 ± 4.7 | 40.0 ± 15.5 | .511 |
| Cox HR, univar | 1.2 (0.6-2.3), P = .70 | 1.5 (0.8-2.9), P = .19 | Ref | 1.7 (0.6-5.1), P = .31 | .518 |
| Cox HR, multivar | 1.0 (0.5-2.1), P = .90 | 1.4 (0.7-2.7), P = .29 | Ref | 1.5 (0.5-4.3), P = .89 | .685 |
| NRM | |||||
| Cumulative incidence, %* | 16.7 ± 0.3 | 31.1 ± 0.5 | 17.9 ± 0.2 | 30.0 ± 2.4 | .312 |
| SHR (95% CI), univar | 0.9 (0.4-2.3), P = .85 | 1.8 (0.9-3.6), P = .12 | Ref | 1.8 (0.5-6.3), P = .36 | .310 |
| SHR (95% CI), multivar | 0.8 (0.3-2.1), P = .69 | 1.6 (0.8-3.3), P = .17 | Ref | 1.5 (0.4-4.9), P = .53 | .360 |
| Fatal lung injury | |||||
| Cumulative incidence, %* | 4.8 ± 0.1 | 22.2 ± 0.4 | 3.6 ± 0.1 | 20.0 ± 1.8 | .003 |
| SHR (95% CI), univar | 1.4 (0.2-8.3), P = .73 | 6.8 (1.9-24.3), P = .003 | Ref | 6.3 (1.1-38.2), P = .044 | .010 |
| SHR (95% CI), multivar | 1.2 (0.2-6.9), P = .87 | 5.3 (1.5-19.0), P = .01 | Ref | 4.4 (0.8-26.1), P = .10 | .032 |
| Post-HCT outcomes . | Pre-HCT pulmonary metatranscriptome . | P . | |||
|---|---|---|---|---|---|
| Cluster 1 (n = 42) OP-heavy . | Cluster 2 (n = 45) Microbially depleted . | Cluster 3 (n = 84) OP-light . | Cluster 4 (n = 10) Virus enriched . | ||
| Lung injury | |||||
| Cumulative incidence, %* | 19.0 ± 0.4 | 44.4 ± 0.6 | 9.5 ± 0.1 | 30.0 ± 2.4 | <.001 |
| SHR (95% CI), univar | 2.1 (0.8-5.7), P = .13 | 5.6 (2.5-12.8), P < .001 | Ref | 3.5 (0.9-13.0), P = .06 | <.001 |
| SHR (95% CI), multivar | 1.8 (0.7-4.8), P = .23 | 3.8 (1.4-9.9), P = .007 | Ref | 2.6 (0.6-10.7), P = .19 | .052 |
| All-cause mortality | |||||
| Kaplan-Meier estimate, %† | 28.6 ± 7.0 | 37.8 ± 7.2 | 25.0 ± 4.7 | 40.0 ± 15.5 | .511 |
| Cox HR, univar | 1.2 (0.6-2.3), P = .70 | 1.5 (0.8-2.9), P = .19 | Ref | 1.7 (0.6-5.1), P = .31 | .518 |
| Cox HR, multivar | 1.0 (0.5-2.1), P = .90 | 1.4 (0.7-2.7), P = .29 | Ref | 1.5 (0.5-4.3), P = .89 | .685 |
| NRM | |||||
| Cumulative incidence, %* | 16.7 ± 0.3 | 31.1 ± 0.5 | 17.9 ± 0.2 | 30.0 ± 2.4 | .312 |
| SHR (95% CI), univar | 0.9 (0.4-2.3), P = .85 | 1.8 (0.9-3.6), P = .12 | Ref | 1.8 (0.5-6.3), P = .36 | .310 |
| SHR (95% CI), multivar | 0.8 (0.3-2.1), P = .69 | 1.6 (0.8-3.3), P = .17 | Ref | 1.5 (0.4-4.9), P = .53 | .360 |
| Fatal lung injury | |||||
| Cumulative incidence, %* | 4.8 ± 0.1 | 22.2 ± 0.4 | 3.6 ± 0.1 | 20.0 ± 1.8 | .003 |
| SHR (95% CI), univar | 1.4 (0.2-8.3), P = .73 | 6.8 (1.9-24.3), P = .003 | Ref | 6.3 (1.1-38.2), P = .044 | .010 |
| SHR (95% CI), multivar | 1.2 (0.2-6.9), P = .87 | 5.3 (1.5-19.0), P = .01 | Ref | 4.4 (0.8-26.1), P = .10 | .032 |
Cumulative incidence and variance were calculated using the Fine and Gray cumulative incidence function and compared with the Fine and Gray competing risks regression. Competing risks for lung injury, NRM, and fatal lung injury were death, relapse, and death not related to lung injury, respectively. SHR was calculated with the Fine and Gray competing risks regression. All-cause mortality and standard error were estimated with Kaplan-Meier function and compared using the log-rank test. All-cause mortality hazard ratios were calculated using Cox regression. All multivariate models were adjusted for HCT era. Cluster 3 is the reference group for all comparisons. Bold data indicate statistically significant comparisons.
Multivar, multivariate; Ref, reference; univar, univariate.
Data reported as cumulative incidence estimate with variance.
Estimate is reported as Kaplan-Meier estimate with standard error.