Prognostic value of interstitial lung abnormalities in patients with cirrhosis: A retrospective cohort study

To the Editor: Liver cirrhosis is the end stage of various chronic liver diseases characterized by liver tissue fibrosis and impaired liver function.[1] This condition has now become a substantial global health burden, with significant morbidity and mortality.[1–3] Identifying the risk factors that influence the occurrence, development, and prognosis of cirrhosis is important. Recently, the concept of interstitial lung abnormalities (ILAs), which are considered an early stage of pulmonary fibrosis, was proposed. Given that ILAs, as precursors to interstitial lung disease (ILD), share a fibrotic nature with cirrhosis and considering the documented association between ILDs and cirrhosis, there may be a potential link between ILAs and cirrhosis, but this link remains unexplored in the current literature. Therefore, in this study, we aimed to explore (1) the proportion of patients with ILAs in the population with liver cirrhosis; (2) the mortality of cirrhotic patients with or without ILAs; and (3) whether the presence of ILAs is an independent predictor of mortality in patients with cirrhosis according to a retrospective cohort in China. This study was a retrospective cohort study, which strictly adhered to the ethical principles of the Declaration of Helsinki and was approved by the Ethics Committee of West China Hospital, Sichuan University (Approval No. 20231742). As the study involved only retrospective analysis of medical record data with no additional interventions or risks, the ethics committee granted a waiver of informed consent. Patients had signed a general consent form upon hospital admission regarding the use of their medical records for research purposes. Between August 2011 and May 2023, 4522 cirrhosis patients admitted to the Department of Gastroenterology or Infectious Diseases at West China Hospital were initially enrolled in this study. All patients met the following inclusion criteria: cirrhosis was confirmed by vibration-controlled transient elastography and backed by biochemical markers such as the international normalized ratio, bilirubin, and albumin; age ≥18 years; hospitalization during the period from August 2011 to May 2023; and completion of a chest CT examination. The exclusion criteria included individuals with interstitial lung disease; those diagnosed with hepatocellular carcinoma or portal vein thrombosis; those lost to follow-up; and those with missing covariates exceeding 10%. Finally, 4022 cirrhosis patients were included in the study. The study flowchart is shown in Supplementary Figure 1, https://links.lww.com/CM9/C694. Clinical data were collected from electronic records. ILAs were evaluated by chest computed tomography, and the diagnostic criteria included the following: (1) features of ILA, including ground-glass abnormalities, reticular abnormalities, traction bronchiectasis, honeycombing, and nonemphysematous cysts; (2) involvement of at least 5% of the lung zone, delineated into upper, middle, and lower zones by the levels of the inferior aortic arch and right inferior pulmonary vein; and (3) the absence of suspected interstitial lung disease. Moreover, ILAs can be divided into nonsubpleural ILAs, subpleural nonfibrotic ILAs, and subpleural fibrotic ILAs. Nonparametric variables are expressed as medians with interquartile ranges and analyzed using the Mann–Whitney U test. Parametric variables are reported as mean ± standard deviation and compared with Student’s t test. Categorical variables are summarized as frequencies and percentages and assessed with the chi-square or Fisher’s exact test. The associations between ILAs and all-cause mortality in cirrhosis patients were analyzed via a Cox regression model, and the models were adjusted for age, sex, body mass index, smoking status, drinking status, hypertension, diabetes, ischemic heart disease, chronic kidney disease, malnutrition, and the Child–Pugh score. A detailed description of the methods can be found in the Supplementary Materials, https://links.lww.com/CM9/C694. In this study, a total of 4022 patients with cirrhosis were included; these patients had an average age of 54 years, and 2380 (59.2%) were male. A total of 749 (18.6%) subjects were diagnosed with ILAs. Baseline characteristics were compared between the non-ILA group and the ILA group. Participants with ILAs were more likely to be older and have higher total bilirubin levels, prothrombin times, glutamyl transpeptidase levels, and aspartate aminotransferase levels. Additionally, higher proportions of autoimmune hepatitis and liver transplantation history were observed in the ILA group [Supplementary Table 1, https://links.lww.com/CM9/C694]. Although patients with ILAs had a comparable prevalence of comorbidities to that of non-ILA group, they had more advanced Child–Pugh grade, a greater incidence of hepatic encephalopathy, and lower level of serum albumin (P <0.05). No significant differences were observed in ascites, peritonitis, total bilirubin, or coagulation disorders [Supplementary Table 2, https://links.lww.com/CM9/C694]. During the 1 year of follow-up, 667 (16.6%) deaths were recorded, and patients with ILAs had higher mortality rates (156/749 [20.8%] vs. 511/3273 [15.6%]; χ2 = 12.0; P <0.001) than did those without ILAs [Supplementary Table 3, https://links.lww.com/CM9/C694]. According to the fully adjusted multivariate Cox regression models [Figure 1], the hazard ratio (HR) of patients with ILAs was 1.275 (95% CI, 1.061–1.531; P = 0.009) compared with that of patients without ILAs. During long-term follow-up, the median observation period was 2.3 (1.0–5.1) years, and 1610 (40.0%) deaths were recorded. The cause of death was septic shock in 469 (29.1%) cases, acute-on-chronic liver failure in 277 (17.2%), cardiogenic shock (including heart failure) in 269 (16.7%), multiple organ dysfunction in 224 (13.9%), gastrointestinal bleeding in 134 (8.3%), and respiratory failure in 104 (6.5%). Similarly, patients with ILAs had higher mortality (364/749 [48.6%] vs. 1246/3273 [38.1%]; χ2 = 28.1; P <0.001) than those without ILAs. Moreover, ILAs significantly elevated all-cause mortality (HR: 1.348, 95% CI: 1.196–1.519, P <0.001). Notably, these associations remained significant in patients with alcoholic cirrhosis, primary biliary cirrhosis, and other causes (P <0.05) but not in those with viral or autoimmune cirrhosis [Supplementary Table 4, https://links.lww.com/CM9/C694].Figure 1: Cox regression models were used to analyze the associations between the classification of interstitial lung abnormalities (ILAs) and mortality in patients with cirrhosis. Models were adjusted for age, sex, body mass index, smoking status, drinking status, hypertension status, diabetes status, ischemic heart disease status, chronic kidney disease status, malnutrition status, and Child–Pugh score. CI: Confidence interval; HR: Hazard ratio; ILAs: Interstitial lung abnormalities.The present study revealed five features of ILAs, namely, ground-glass abnormalities, reticular abnormalities, traction bronchiectasis, honeycombing, and nonemphysematous cysts, and all of these features were positively related to long-term all-cause mortality in the adjusted Cox proportional hazard models (P <0.05, Figure 1). Moreover, ILAs presenting with honeycombing had the largest adjusted HR during the 1-year follow-up (HR: 1.996, 95% CI: 1.221–3.263, P = 0.006) and long-term follow-up (HR: 1.601, 95% CI 1.149–2.231, P = 0.005). Kaplan–Meier curve analysis [Supplementary Figure 2, https://links.lww.com/CM9/C694] revealed that subcategories of ILAs had a significantly lower cumulative survival rate than did non-ILAs at the 1-year follow-up (non-ILAs vs. nonsubpleural ILAs vs. subpleural nonfibrotic ILAs vs. subpleural fibrotic ILAs: 83.9% vs. 73.5% vs. 80.8% vs. 75.6%, log rank χ2: 19.37, P = 0.001) and long-term follow-up (non-ILAs vs. nonsubpleural ILAs vs. subpleural nonfibrotic ILAs vs. subpleural fibrotic ILAs: 45.5% vs. 27.7% vs. 28.3% vs. 24.6%, log rank χ2: 55.26, P <0.001). Among the different age, sex, and disease severity subgroups [Supplementary Figure 3, https://links.lww.com/CM9/C694], the associations between ILAs and mortality were consistent (P for interaction >0.05). In this study, there were 205 inconsistent ILA diagnoses in the first round of inspection by specialized CT readers, and the kappa coefficient reached 0.829, which indicates that the diagnostic consistency of ILAs was almost perfect. In the sensitivity analysis, we excluded 205 samples with diagnostic disagreement; the association between ILA and mortality remained consistent in the cirrhotic population compared with the analysis including these samples [Supplementary Table 5, https://links.lww.com/CM9/C694]. Moreover, after the competitive risks of liver transplantation were excluded, the associations between ILAs and mortality were consistent [Supplementary Table 6, https://links.lww.com/CM9/C694]. By employing a relatively large sample size and adequate follow-up periods, this study explored the relationship between ILA and all-cause mortality in patients with cirrhosis. The present study demonstrated that the proportion of ILAs in patients with cirrhosis was 18.6%, and cirrhotic patients with ILAs had 27.5% and 34.8% greater risks of one-year and long-term mortality, respectively, than those without ILAs. In our previous study involving a large sample of Chinese individuals undergoing periodic health check-ups, the prevalence of ILAs was 2.1%, and the most common classification was subpleural nonfibrotic ILAs (81.7%).[4] Notably, this study revealed that 18.6% of patients with cirrhosis had ILAs, and 423/749 (56.5%) had subpleural nonfibrotic ILAs. Compared with the general population, patients with cirrhosis had an 8-fold greater prevalence of ILA; thus, cirrhosis and ILA are comorbidities worthy of attention. In conclusion, we observed a significant relationship between ILA and long-term mortality among patients with cirrhosis in a retrospective cohort from China. Importantly, patients with different classifications of ILA and etiologies of liver cirrhosis are at increased risk of mortality. Therefore, targeting ILA and developing risk stratification, treatment evaluation and prognosis assessment strategies for patients with liver cirrhosis may yield clinical benefits. Acknowledgments The authors express their appreciation to Yizhou Li, Jingxi Ma, Jia Feng, Jiawen Li, Yu Cheng, Yi Yao, and Jiaxin Bai for their contributions to the CT report and to Yonggang Zhang for his guidance on data analysis. Conflicts of interest None. Funding This study was supported by the Funds from the International Cooperation and Exchange of the National Natural Science Foundation of China (No. W2411076) and National Natural Science Foundation of China Young Scientists Fund (No. 62506248).