Radiological bronchiectasis (RBE) is only a radiographic finding, and identified by an international consensus about RBE recently.[2] With the deepening of research, more and more attention was received from RBE in various lung diseases, such as RBE in patients with TB[17], cystic fibrosis,[23] lung cancer,[24] COPD,[25] et al. However, the research on RBE in patients after TBI is still limited. In the present study, the incidence rate of RBE in patients after TBI was 42.33% in Wuhan, China, which was similar to about 40% in India.[17]
Growing evidence suggests that inflammatory response and nutrition status were associated with clinical BE (CBE).[26–33] Scholars from China revealed that Compared to healthy controls, patients with mild-to-moderate CBE had higher levels of inflammatory biomarkers, such as tumor necrosis factor(TNF)-alpha, fibrinogen, ESR, CRP, and ceruloplasmin,[26, 27] while lower levels of albumin, pre-albumin, BMI, and fat-free mass index(FFMI) were observed in the CBE.[27, 28] Additionally, greater systemic levels of inflammation are directly to higher scores of disease severity in patients with CBE,[26, 29–32] and greater nutritional parameters may be associated with a less severity CBE in females.[33] However, limited studies described the mechanism and risk factors of RBE, because of few or no symptoms or exacerbations in patients with RBE.[2, 21] A study focused on the bronchial gene expression alterations in patients with RBE reported a novel mechanism of BE initiation: an increase in the ciliary process and a reduction of the Wnt signaling pathway.[34] It’s worth noting that there is no literature about the potential mechanisms and risk factors for RBE after TBI. China is a country with a high burden of TB, and exploring the potential mechanisms and risk factors for RBE after TBI has significant significance for the formulation of personalized diagnosis and treatment plans for patients after TB infection.
In the present study, the levels of SII, NHR, MHR, PHR, MRR, CAR, CPR, and COUNT scores in the RBE group were significantly higher than those in the NRBE group, while the levels of PNI and HALP scores were significantly lower. In addition, lower PNI, lower HALP score, higher Lg (SII), and higher CAR were independent risk factors for RBE in patients after TBI. These results were roughly consistent with previous studies,[26–28, 34] which suggested that inflammation processes and low nutrition status emerged in the early stages of BE, and RBE may eventually develop into CBE as the greater inflammation status and lower nutrition status continues. In addition, our team previously revealed that TB patients with RBE had lower survival time compared to TB patients without RBE. Hence, the NIRS was constructed to identify the high-risk population of RBE to carry out early intervention and reduce the additional burden of TBI. The present results show that NIRS could predict RBE with AUC of 0.757, 0.754, and 0.761 in the development cohort, internal validation cohort, and external validation cohort, respectively. However, our team thinks that the predictive performance is unsatisfactory, and hence, INRS and clinical factors were combined to construct a higher predictive nomogram model.
The present results showed that was Age, smoker, COPD, albumin, RDW.CV and INRS ≥ 1.86 were independent risk factors for RBE after TBI. Previous studies indicated that the incidence rate of CBE increases with age,[35] which was roughly consistent with our result about age. One possible reason is that the functional decline of various organs and tissues leads to a loss of balance between nutrient supply and consumption and basic diseases, such as hypertension, coronary heart disease, and diabetes, break the stability of the internal environment to increase the incidence of malnutrition. Another possible reason is that immunosenescence brings more inflammatory reactions.[36] A survey of post-TB bronchiectasis from India found that smoking prevalence was higher in post-TB bronchiectasis than in post-TB without bronchiectasis (27.27% vs. 12.12%).[37] Additionally, smokers could increase the risk of RBE in patients with autosomal dominant polycystic kidney disease.[38] Previous literature may explain the phenomenon that smoking increases the incidence rate of RBE, by which nicotine can increase the release of TNF-a, a key inflammatory cytokine.[39] In addition, Increasingly, there is evidence that COPD and bronchiectasis are underpinned by similar pathophysiological processes, especially centered on the functions of the neutrophil and impaired immunity. Subjects with COPD present different forms of impaired immunity that facilitate the persistent bronchial inflammation in the lower airways, progressively damaging the bronchial wall through the release of bacterial and inflammatory proteolytic products, leading to the formation of bronchiectasis.[40] Similarly, lower albumin is an independent risk factor for RBE after TBI, which was roughly consistent with a previous study in which lower levels of albumin were observed in the CBE.[27] RDW-CV was proven to have significant associations with poor prognosis in anemic TB patients.[35] So far, there is no research exploring the relationship between RDW.CV and BE, the present study found for the first time that high levels of RDW.CV increases the risk of incidence rate of RBE in patients after TBI.
Finally, the prediction nomogram model was successfully constructed to identify the high-risk population of RBE after TBI. The prediction nomogram model showed an excellent predictive value, with AUC of 0.839, 0.837, and 0.829in the development cohort, internal validation cohort, and external validation cohort, respectively. In addition, the INRS-based nomogram model also showed a strong accuracy and clinical utility in development, internal validation, and external validation cohorts. The INRS-based nomogram model has many advantages. First, the prediction nomogram model not only explains the mechanism of RBE occurrence but also reflects the clinical characteristics of RBE, thus providing targeted guidance to patients after TBI. For instance, we can persuade smoking cessation, optimize nutritional status, and improve hyper-inflammatory states, thereby minimizing the incidence rate of RBE in patients after TBI. Secondly, variables in our prediction nomogram model are accessible in most hospitals, even primary hospitals, and it’s of great significance for poor areas, especially areas with high TB burden. Last but not least, our prediction nomogram model has a remarkable performance power in predicting RBE due to TBI. When patients are at greater risk of RBE, they should be advised to go to a high-level hospital for further treatment, which will reduce the incidence rate of RBE, time cost, and financial loss, and improve the quality of life.
However, the current study is not exempt from some limitations. Firstly, the present study is retrospective, there may be some selection bias in the study population, and further validation is needed through prospective research. Secondly, the impact of anti-tuberculosis drugs on inflammation indicators and nutrition indexes has not been ruled out. Thirdly, the variables of time to diagnosis of RBE and body mass index (BMI) was not included in the logistic regression because of the lack of data in the retrospective analysis. In addition, while we verified the nomogram to predict RBE in an independent validation set, the sample size in the external validation cohort was relatively small, and the external validation was only performed at a single medical center.
Our study is the first to establish an INRS by combining PNI, HALP score, Lg (SII), and CAR levels. We demonstrated HINS’s superior prediction ability, which may be a potential predictive biomarker for RBE in patients after TBI. We also developed an INRS-based nomogram model combined with several clinical parameters. This model is a convenient and efficient tool for predicting RBE in patients after TBI and helps clinicians better manage the follow-up treatment of patients after TBI.