Diagnostic Factors Associated with Sarcoidosis in Patients Referred for EBUS-TBNA Due to Mediastinal Lymphadenopathy
Round 1
Reviewer 1 Report
Comments and Suggestions for Authors
Dear colleagues,
The manuscript addresses an important and clinically relevant question: the identification of pre-procedural clinical, laboratory and radiological factors associated with a definitive diagnosis of sarcoidosis in patients undergoing first-time EBUS-TBNA for mediastinal lymphadenopathy. Given that endobronchial ultrasound-guided transbronchial needle aspiration is widely implemented in routine respiratory practice but does not invariably yield a definitive diagnosis at first attempt, the development of predictive models based on readily available variables is of practical significance.
Overall, the study is well conceived and focuses on a clinically meaningful endpoint. The attempt to construct point-based diagnostic scoring models is commendable and aligns with current trends towards risk stratification and evidence-informed decision-making. However, several methodological aspects require clarification to enhance transparency, reproducibility and interpretability.
Strengths
Clinical relevance: The study addresses a frequent diagnostic dilemma in respiratory and thoracic medicine.
Use of routinely available variables: The inclusion of demographic, laboratory and CT parameters increases potential applicability in everyday practice.
Model development: The creation of two point-based scoring systems represents a practical approach to translating statistical findings into clinical tools.
Clear clinical message: Younger age (≤55 years), female sex, absence of a pulmonary mass >10 mm, normal white blood cell count and typical clinical features of sarcoidosis were shown to be associated with a higher likelihood of sarcoidosis.
Major Comments
1. Diagnostic Criteria
The manuscript would benefit from a more explicit definition of the diagnostic criteria used for both sarcoidosis and malignancy.
For sarcoidosis, it should be clearly stated whether diagnosis required:
cytological or histological evidence of non-caseating granulomas;
exclusion of infectious granulomatous diseases (e.g. microbiological testing for mycobacteria and fungi);
supportive clinical and radiological findings;
longitudinal follow-up confirming disease stability or evolution consistent with sarcoidosis.
For malignancy, clarification is required regarding:
cytological or histological confirmation;
use of immunohistochemistry where appropriate;
reliance on subsequent biopsy, surgical pathology or clinical follow-up in cases with non-diagnostic EBUS-TBNA results.
Clear specification of these criteria is essential to ensure internal validity and to allow readers to assess the robustness of the reference standard.
2. Inclusion and Exclusion Criteria
The inclusion and exclusion criteria should be described in greater detail. Specifically:
Inclusion criteria should define:
the minimum size threshold for mediastinal lymphadenopathy on CT;
age limits, if any;
confirmation that all procedures were first-time EBUS-TBNA.
Exclusion criteria should specify:
prior established diagnosis of sarcoidosis or malignancy;
history of treated cancer;
active infectious disease;
incomplete clinical or follow-up data;
immunosuppressed states, if applicable.
Without this information, selection bias cannot be adequately assessed.
3. Methodological Transparency
Further detail is required regarding:
Sample size and distribution of final diagnoses.
Statistical methods used for variable selection (e.g. univariable screening followed by multivariable logistic regression).
Assessment of model performance, including:
area under the receiver operating characteristic curve (AUC);
sensitivity, specificity and predictive values;
calibration metrics.
Internal validation procedures (e.g. cross-validation or bootstrapping).
The statement that the models demonstrated “clinically relevant discriminatory performance” should be supported by quantitative data.
4. Definition of the Final Diagnosis
The manuscript should clarify how the “definitive diagnosis” was established in cases where initial EBUS-TBNA was non-diagnostic. It is important to describe whether patients underwent repeat biopsy, surgical mediastinoscopy, or structured radiological and clinical follow-up.
Minor Comments
The rationale for the age cut-off of 55 years should be explained (clinical convention vs. data-driven threshold).
The definition of “clinical features typical of sarcoidosis” should be explicitly listed.
Consider discussing potential confounders, such as smoking status or ethnicity, if data were available.
Conclusions
This study addresses a relevant diagnostic challenge and proposes pragmatic scoring systems based on easily obtainable pre-procedural variables. The findings are clinically plausible and potentially useful in guiding patient counselling and procedural planning.
However, clarification of diagnostic standards, inclusion and exclusion criteria, and statistical methodology is necessary to strengthen the scientific rigour of the work. Prospective, preferably multicentre, external validation is required before widespread clinical implementation of the proposed scoring system.
Subject to satisfactory revision addressing the points above, the manuscript may represent a valuable contribution to the literature on the diagnostic approach to mediastinal lymphadenopathy and suspected sarcoidosis.
Best wishes,
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for Authors
Dear colleagues! The manuscript addresses a clinically relevant question. Given the high prevalence of sarcoidosis among patients referred for EBUS-TBNA and the limited possibilities of EBUS in some cases, a simple pre-procedural scoring system that estimates the probability of sarcoidosis is a valuable concept. The idea behind this study is interesting and has practical potential.
However, there is a fundamental methodological concern regarding the definition of the study cohort that needs to be addressed before the results can be considered reliable.
The logistic regression analyses (Tables 4, 6, 7) were performed on all 274 patients, which means that 63 patients without a confirmed disease diagnosis were included in the "no sarcoidosis" reference group. Among these 63 patients, 37 had no follow-up data at all. It is unknown how many of these patients actually had sarcoidosis — some may have been lost to follow-up precisely because they were asymptomatic, which is not uncommon in sarcoidosis. Including these patients in the reference group introduces a misclassification bias: if even a proportion of them had undiagnosed sarcoidosis, the odds ratios and the scoring system derived from them may be substantially distorted.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Comments and Suggestions for Authors
Dear colleagues,
All my recommendations were added.
Reviewer 2 Report
Comments and Suggestions for Authors
Dear colleagues, Thank you for the answer; I think it will be enough for the publication of this paper.
