Abstract
A recent study associated early proliferative or fibrotic diffuse alveolar damage with mortality in acute respiratory distress syndrome. We highlight four interpretive concerns: phenotype status was assigned at biopsy after follow-up began, creating potential immortal-time bias; a single biopsy cannot measure pathological velocity; sparse and imbalanced phase-by-timing cells limit component-specific inference; and concurrent physiological measurements complicate causal interpretation. Until delayed-entry survival and phase-specific sensitivity analyses are reported, the phenotype is better viewed as early advanced-phase histology than evidence of accelerated pathology.
Keywords: Acute respiratory distress syndrome, Diffuse alveolar damage, Immortal-time bias, Lung biopsy, Survival analysis
To the Editor
We read with interest Chang et al.'s study of the chronological phases of diffuse alveolar damage (DAD) in acute respiratory distress syndrome (ARDS) [1]. The authors are to be commended for assembling a substantial biopsy-confirmed cohort and examining whether histological timing carries prognostic information. They defined "rapid progression" as proliferative DAD within 7 days or fibrotic DAD within 21 days after ARDS diagnosis and reported associations with 60-day mortality and lower biopsy-day arterial carbon dioxide tension (PaCO2). Several issues warrant clarification.
First, survival was measured from ARDS diagnosis, whereas rapid status was assigned only at biopsy, performed a mean of 5.4 days after diagnosis in the rapid group and 11.3 days in the non-rapid group. All included patients had to survive until biopsy, so the later-biopsied non-rapid group accrued about 5.9 additional days of guaranteed survival before classification. Starting follow-up at diagnosis while assigning groups using later information may introduce immortal-time bias, which would be expected to favor the later-biopsied non-rapid group and could bias the hazard ratio away from the null, although the net direction after accounting for other selection mechanisms cannot be determined without reanalysis [2, 3]. Similar biopsy intervals among survivors and non-survivors do not resolve this risk-set problem. A left-truncated Cox model using ARDS diagnosis as the time scale and biopsy date as delayed entry, supplemented by an analysis re-anchored at biopsy or a restricted biopsy-window analysis within the proliferative phase around the 7-day threshold, would better align eligibility, classification, and follow-up.
Second, a single biopsy provides a cross-sectional snapshot, not an intra-individual measure of pathological velocity. DAD features overlap temporally and may coexist within one lung [4, 5]. The composite is also structurally imbalanced: the rapid group contained 27 proliferative and 21 fibrotic cases, whereas the non-rapid group contained 36 proliferative, 2 fibrotic, and 4 exudative cases. The fibrotic non-rapid cell was, therefore, limited to two patients; its 28-day and 60-day mortality estimates (0/2 and 1/2) illustrate the resulting instability, precluding any reliable within-phase inference. All four exudative cases were necessarily non-rapid and had no 60-day deaths. In the final cohort, 21 of 23 fibrotic cases represents 91.3%; we would welcome clarification of the denominator used for the reported 84.0%. Within the fibrotic stratum, the prespecified 21-day cutoff produced nonsignificant mortality contrasts, whereas a 10-day cutoff selected by receiver operating characteristic analysis in the same cohort produced a significant 28-day, but not 60-day, contrast. The prespecified and data-driven definitions should be distinguished. Reporting events and exact confidence intervals in the phase-by-status table, phase-specific estimates, an interaction test, and a sensitivity analysis excluding exudative cases would clarify whether the pooled association is consistent across components.
Third, pathological ascertainment may vary across sampling approaches. The Methods do not specify whether a unified blinded re-review was performed, how a dominant phase was assigned when features coexisted, or whether interobserver agreement was quantified. DAD was detected in 66.9% of open lung biopsies but 21.2% of transbronchial cryobiopsies [1]. Because modality is linked to sampling adequacy, patient selection, and calendar period [6], reporting modality by progression group and an open-lung-biopsy-only sensitivity analysis would be informative. If feasible, dual blinded review with an agreement statistic would further strengthen reproducibility [7].
Finally, biopsy-day PaCO2 is a concomitant marker, not an antecedent predictor. Ventilatory ratio includes PaCO2 in its numerator [8]; entering both variables in one model creates mathematical coupling. Separate models with collinearity diagnostics would be preferable. Table 6 labels exponentiated logistic-regression coefficients as relative risks, although they are odds ratios. With 48 of 90 patients classified as rapid, this distinction is material; the estimates should be relabelled or risk ratios estimated with modified Poisson regression [9]. Similar tidal volumes and numerically lower driving pressure and mechanical power in the rapid group do not support a specific ventilator-mediated mechanism. Multiple competing physiological and treatment-related explanations remain, and longitudinal, time-weighted measurements would better distinguish them.
The authors appropriately acknowledge the associational nature of their findings. These caveats, however, do not resolve temporal misalignment, the sparse composite structure, or the interpretation of concurrent physiological measurements. Until delayed-entry and component-specific analyses are reported, "rapid progression" is better interpreted as early detection of advanced-phase histology among selected patients who survived to biopsy, rather than directly observed pathological acceleration. Clarifying these boundaries is essential before this uncommon cohort informs prognostic enrichment or trial eligibility criteria.
Acknowledgements
Not applicable.
Abbreviations
- ARDS
Acute respiratory distress syndrome
- DAD
Diffuse alveolar damage
- PaCO2
Partial pressure of carbon dioxide in arterial blood
Author contributions
WM and XL performed the literature search, conceived the arguments, and drafted the manuscript. WL critically reviewed and revised the manuscript for important intellectual content. All authors read and approved the final manuscript and agree to be accountable for its content.
Funding
This work was supported by the National Natural Science Foundation of China (82160952) and the Basic Research Program of Guizhou Provincial Department of Science and Technology (qiankehe jichu [2020]1Y364).
Data availability
No datasets were generated or analyzed during the current study.
Declarations
Ethics approval and consent to participate
Not applicable.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Chang KW, Huang CH, Leu SW, Chiu TH, Fang HY, Wu HP, et al. The relationship between the chronological phases of diffuse alveolar damage and clinical outcomes in acute respiratory distress syndrome. J Intensive Care. 2026. 10.1186/s40560-026-00911-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Fu EL, Harhay MO, Schneeweiss S, Desai RJ, Hernán MA. Starting right: aligning eligibility and treatment assignment at time zero when emulating a target trial. BMJ. 2026;392:e084909. 10.1136/bmj-2025-084909. [DOI] [PubMed] [Google Scholar]
- 3.Hernán MA, Sterne JAC, Higgins JPT, Shrier I, Hernández-Díaz S. A structural description of biases that generate immortal time. Epidemiology. 2025;36(1):107–14. 10.1097/ede.0000000000001808. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Thille AW, Esteban A, Fernández-Segoviano P, Rodriguez JM, Aramburu JA, Vargas-Errázuriz P, et al. Chronology of histological lesions in acute respiratory distress syndrome with diffuse alveolar damage: a prospective cohort study of clinical autopsies. Lancet Respir Med. 2013;1(5):395–401. 10.1016/s2213-2600(13)70053-5. [DOI] [PubMed] [Google Scholar]
- 5.Lee JTH, Barnett SN, Roberts K, Ashwin H, Milross L, Cho JW, et al. Integrated histopathology, spatial and single cell transcriptomics resolve cellular drivers of early and late alveolar damage in COVID-19. Nat Commun. 2025;16(1):1979. 10.1038/s41467-025-56473-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Li Q, Lafrance D, Liberman M, Leduc C, Charbonney E, Titova P, et al. Transbronchial lung cryobiopsies, transbronchial forceps lung biopsies, and surgical lung biopsies in mechanically ventilated patients with acute hypoxemic respiratory failure: a retrospective cohort study. J Intensive Care Med. 2024;39(10):985–93. 10.1177/08850666241247145. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Vanbelle S, Engelhart CH, Blix E. A comprehensive guide to study the agreement and reliability of multi-observer ordinal data. BMC Med Res Methodol. 2024;24(1):310. 10.1186/s12874-024-02431-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Sinha P, Calfee CS, Beitler JR, Soni N, Ho K, Matthay MA, et al. Physiologic analysis and clinical performance of the ventilatory ratio in acute respiratory distress syndrome. Am J Respir Crit Care Med. 2019;199(3):333–41. 10.1164/rccm.201804-0692OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Sawada K, Hagiwara Y, Matsuyama Y. Choosing between modified Poisson and log-binomial regression: evidence for the superiority of modified Poisson regression under heterogeneous risk ratios. Am J Epidemiol. 2025. 10.1093/aje/kwaf232. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
No datasets were generated or analyzed during the current study.