CDK4/6 inhibitor-related interstitial lung disease in breast cancer: a single-center case series and literature-based patient-level review

Abstract

Background

Cyclin-dependent kinase 4/6 (CDK4/6) inhibitors are widely used in the treatment of breast cancer. CDK4/6 inhibitor-related interstitial lung disease (ILD) is uncommon but potentially serious, and its clinical manifestations, radiologic patterns, and outcomes remain incompletely characterized. This study aimed to characterize the clinical presentation, radiologic patterns, diagnostic evaluation, treatment, and outcomes of CDK4/6 inhibitor-related ILD using a Chinese single-center case series and an updated literature-based patient-level review.

Methods

Seven patients with clinically diagnosed CDK4/6 inhibitor-related ILD were retrospectively identified at our center between November 2022 and February 2026. Clinical data, chest imaging, bronchoscopy and bronchoalveolar lavage fluid (BALF) findings when available, treatment, and outcomes were collected. An updated patient-level literature review was performed through May 30, 2026 using PubMed, Web of Science, and Google Scholar.

Results

A total of 60 cases were analyzed, including seven institutional cases and 53 published cases. The median interval from CDK4/6 inhibitor initiation to ILD onset or diagnosis was 139 days among 59 cases with available latency data, and 24 cases (40.7%) occurred within 3 months. Thirty patients had received palbociclib, 22 had received abemaciclib, and eight had received ribociclib. Fatal outcomes occurred in nine patients. The most frequently reported symptoms were cough and dyspnea, whereas severe cases progressed to hypoxemic respiratory failure. Chest computed tomography (CT) findings were heterogeneous, including ground-glass opacities (GGOs), interlobular septal thickening, patchy consolidations, and radiologic patterns classified as nonspecific interstitial pneumonia (NSIP), organizing pneumonia (OP), hypersensitivity pneumonitis (HP)-like, and acute interstitial pneumonia/acute respiratory distress syndrome (AIP/ARDS)-like. Among eight patients with an AIP/ARDS-like pattern, five died and three had non-fatal outcomes. In the institutional cases, no pathogen was detected in available BALF cultures, and lymphocytic predominance was observed in cases with cell differential data.

Conclusions

CDK4/6 inhibitor-related ILD should be considered when new respiratory symptoms or new CT abnormalities occur during CDK4/6 inhibitor therapy. Because its presentation is nonspecific, evaluation should focus on high-resolution CT (HRCT) assessment, exclusion of infection, tumor progression and radiation-related lung injury, and severity-based decisions regarding drug interruption or discontinuation. BALF is most useful when infection remains a major competing diagnosis.

Highlight box.

Key findings

• Cyclin-dependent kinase 4/6 (CDK4/6) inhibitor-related interstitial lung disease (ILD) showed a broad clinical spectrum, ranging from mild radiologic abnormalities to fatal respiratory failure. In the pooled summary of 60 cases, fatal outcomes were overrepresented among patients with an acute interstitial pneumonia/acute respiratory distress syndrome (AIP/ARDS)-like pattern, although this pattern was not uniformly fatal.

What is known and what is new?

• CDK4/6 inhibitor-related ILD is uncommon but potentially serious.

• This study adds a Chinese single-center case series and an updated literature-based patient-level review, summarizing clinical presentation, high-resolution computed tomography patterns, bronchoalveolar lavage fluid findings, treatment, and outcomes.

What is the implication, and what should change now?

• Clinicians should not attribute new pulmonary opacities during CDK4/6 inhibitor therapy to drug toxicity without excluding infection, cancer progression, lymphangitic carcinomatosis and radiation pneumonitis. Patients with diffuse bilateral ground-glass opacity, worsening oxygenation, or an AIP/ARDS-like pattern require early pulmonology involvement and active management.

Introduction

Cyclin-dependent kinase 4/6 (CDK4/6) inhibitors have become important therapeutic agents for hormone receptor-positive, human epidermal growth factor receptor 2-negative breast cancer. When combined with endocrine therapy, these agents improve clinical outcomes and are generally well tolerated. Common adverse events include myelosuppression, gastrointestinal symptoms, and hepatic dysfunction, most of which can be managed with dose adjustment, temporary interruption, or supportive care (1). In contrast, CDK4/6 inhibitor-related interstitial lung disease (ILD) is uncommon but clinically important because a subset of patients may develop rapidly progressive hypoxemic respiratory failure and fatal outcomes (2-4).

Drug-related ILD is a heterogeneous pulmonary adverse reaction and usually lacks specific clinical, laboratory, radiologic, or pathologic features (5-7). In patients with breast cancer, the diagnosis of CDK4/6 inhibitor-related ILD is particularly challenging because infectious pneumonia, tumor progression, lymphangitic carcinomatosis, radiation pneumonitis, exacerbation of pre-existing ILD, and lung injury related to other anticancer therapies may present with overlapping symptoms and computed tomography (CT) findings (4,6). Therefore, CDK4/6 inhibitor-related ILD should not be diagnosed solely on the basis of new or worsening pulmonary opacities during treatment. Instead, diagnosis requires comprehensive assessment of the temporal relationship with drug exposure, high-resolution CT (HRCT) findings, microbiologic evaluation, oncologic status, clinical course after drug interruption or discontinuation and/or treatment, and multidisciplinary clinical judgment.

Chest CT is central to the evaluation of suspected drug-related ILD. Previous studies have described several radiologic patterns of drug-induced ILD, including nonspecific interstitial pneumonia (NSIP), organizing pneumonia (OP), hypersensitivity pneumonitis (HP)-like pattern, and acute interstitial pneumonia/acute respiratory distress syndrome (AIP/ARDS)-like pattern (5,6). However, the clinical significance of these patterns in CDK4/6 inhibitor-related ILD remains incompletely defined because most available evidence comes from case reports, small case series, and heterogeneous real-world reports.

In this study, we described seven patients with clinically diagnosed CDK4/6 inhibitor-related ILD from a Chinese center and performed an updated literature-based patient-level review of published cases through May 2026. We focused on practical questions encountered in clinical care: when CDK4/6 inhibitor-related ILD should be suspected, how alternative diagnoses were evaluated, what CT and bronchoalveolar lavage fluid (BALF) findings were observed, and which clinical features were seen in fatal cases. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2026-1560/rc).

Methods

Study population

This study was a single-center retrospective case series. Seven patients who were clinically diagnosed with CDK4/6 inhibitor-related ILD between November 2022 and February 2026 at Peking University People’s Hospital were included. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Peking University People’s Hospital, and individual consent was waived due to the retrospective nature of the study.

Clinical data were retrospectively collected from medical records, including demographic characteristics, CDK4/6 inhibitor exposure, combined endocrine therapy, relevant oncologic and pulmonary history, clinical manifestations, laboratory findings, chest imaging findings, bronchoscopy and BALF findings when available, treatment, clinical outcomes, and Naranjo Adverse Drug Reaction Probability Scale score. Clinical outcomes were assessed at the last available follow-up.

Diagnostic criteria and assessment

In this study, CDK4/6 inhibitor-related ILD was defined as newly developed or worsening pulmonary disease during treatment with a CDK4/6 inhibitor, including ILD, pneumonitis-like changes, pulmonary fibrosis, bronchiolitis, and pulmonary infiltrative abnormalities reported in previous studies (3). Given the complex causes of pulmonary abnormalities in patients with breast cancer, the diagnosis was made by comprehensive clinical adjudication rather than by a single diagnostic test.

Alternative causes, including infectious pneumonia, lymphangitic carcinomatosis, radiation pneumonitis, tumor progression, exacerbation of pre-existing ILD, and lung injury related to other anticancer therapies, were considered according to available clinical, radiologic, microbiologic, bronchoscopic, and oncologic information. All institutional cases were jointly reviewed by two senior pulmonologists. The final clinical diagnosis was reached by consensus after evaluating the temporal relationship with CDK4/6 inhibitor exposure, chest CT evolution, available microbiologic results, oncologic status, treatment response, and Naranjo score.

Chest CT patterns were categorized according to previously reported radiologic patterns of drug-induced ILD, including AIP/ARDS-like pattern, NSIP, OP, acute eosinophilic pneumonia, and HP-like pattern (5,7). For literature cases, HRCT features and patterns were classified according to the descriptions provided in the original reports. When a formal ILD pattern was not reported, descriptive CT findings such as ground-glass opacity (GGO) were recorded as reported.

The causality between CDK4/6 inhibitor exposure and ILD was assessed using the Naranjo Adverse Drug Reaction Probability Scale (8). Scores of ≥9, 5–8, 1–4, and ≤0 were interpreted as definite, probable, possible, and doubtful causality, respectively. The severity of drug-related ILD was graded according to the Common Terminology Criteria for Adverse Events, version 5.0.

Corticosteroid dose classification

Corticosteroid doses were categorized according to methylprednisolone-equivalent daily doses. Pulse-dose therapy was defined as 7.5–30 mg/kg/day, high-dose therapy as >1 mg/kg/day, moderate-dose therapy as 0.5–1 mg/kg/day, low-dose therapy as <0.5 mg/kg/day, and long-term maintenance therapy as 2.5–15 mg/day. For pooled analyses, moderate-dose and low-dose corticosteroids were combined into one category. Because corticosteroid intensity was likely determined by ILD severity, analyses involving corticosteroid regimens were considered descriptive and exploratory.

Literature-based patient-level review

An updated literature-based patient-level review was performed through May 30, 2026 using PubMed, Web of Science, and Google Scholar to identify published cases of CDK4/6 inhibitor-related ILD. The search terms included “CDK4/6 inhibitor”, “cyclin-dependent kinase 4/6 inhibitor”, “abemaciclib”, “palbociclib”, “ribociclib”, “interstitial lung disease”, “pneumonitis”, “pneumonia”, “drug-induced interstitial lung disease”, “DILD”, “pulmonary toxicity”, and “lung injury”. The reference lists of included articles were also manually reviewed to identify additional eligible reports.

Eligible reports for the pooled patient-level analysis were case reports or case series describing suspected or clinically diagnosed CDK4/6 inhibitor-related ILD and providing sufficient patient-level information, including CDK4/6 inhibitor exposure, time to ILD onset or diagnosis, clinical presentation, radiologic findings, treatment, and clinical outcome. Reviews without original patient-level data, duplicate reports, reports with insufficient individual patient-level information, and cases in which a CDK4/6 inhibitor could not be identified as the suspected culprit drug were excluded from the primary analysis. Conference abstracts were excluded unless sufficient patient-level information was available for extraction of the core variables. Cohort studies reporting only aggregate data without sufficient extractable case-level information for the core variables were not included in the pooled patient-level analysis but were considered as contextual evidence in the Discussion.

For each eligible case, the following data were extracted when available: age, CDK4/6 inhibitor type, combined endocrine therapy, time from CDK4/6 inhibitor initiation to ILD onset or diagnosis, clinical symptoms, chest CT findings, bronchoscopy or BALF findings, treatment, and clinical outcome. Time-to-onset data were extracted as originally reported. When time was reported in months or years, it was converted approximately using 30 days per month and 365 days per year for descriptive analysis. A 12-month onset was categorized as >3 months to ≤1 year.

The updated literature review identified 53 published patient-level cases eligible for the primary literature table. Together with the seven institutional cases, 60 cases were included in the pooled patient-level summary.

Outcomes and statistical analysis

This study was a retrospective case series combined with a literature-based patient-level review; therefore, statistical analyses were mainly descriptive. Continuous variables were summarized as median with range or interquartile range, and categorical variables were summarized as number and percentage. Patients in the pooled summary were divided into fatal and non-fatal groups according to reported clinical outcomes. Non-fatal outcomes included recovered and clinically stable cases.

Drug-specific case counts were summarized descriptively. No inferential comparison across CDK4/6 inhibitor types was performed because this selected pooled patient-level sample lacked exposed-population denominators and was susceptible to differences in drug utilization, duration of market availability, publication, and reporting practices. Exploratory comparisons of time-to-onset category, HRCT pattern, and glucocorticoid treatment regimen between fatal and non-fatal cases were performed using the Fisher-Freeman-Halton exact test for R × 2 contingency tables.

Variables with missing data were analyzed using available cases, and missing data were reported in the corresponding tables or footnotes. Given the small number of fatal cases, heterogeneous reporting across published cases, and potential publication bias inherent to case reports and case series, P values were interpreted cautiously and were not considered confirmatory. Multivariable analysis was not performed because of the limited number of fatal events and incomplete reporting of key variables across literature cases.

All statistical analyses were performed using R software, version 4.3.3. A two-sided P value <0.05 was considered statistically significant for exploratory analyses.

Results

Basic clinical characteristics

From November 2022 to February 2026, seven patients were clinically diagnosed with CDK4/6 inhibitor-related ILD at our center (Table 1). All patients were female, with ages ranging from 57 to 71 years. One patient had pre-existing interstitial pneumonia before initiation of CDK4/6 inhibitor therapy; this case was included because new or worsening pulmonary abnormalities developed during treatment and alternative causes were clinically evaluated. Six patients received CDK4/6 inhibitors in combination with endocrine therapy. Patient 6 was treated with palbociclib, whereas the remaining patients received abemaciclib. Most patients had a history of thoracic radiotherapy and/or lung metastasis before CDK4/6 inhibitor treatment.

Table 1. Baseline characteristics, cancer treatment, ILD management, and outcomes of the seven institutional cases.

Case	Age (years)/sex	CDK4/6 inhibitor	Endocrine therapy	Time from CDK4/6 inhibitor initiation to ILD onset/detection, months	Thoracic/breast radiotherapy	Lung metastasis before ILD	Pre-existing ILD	Smoking history	Presentation	CTCAE grade	Other adverse events	ILD management	Radiologic course	Clinical outcome	Naranjo ADR score
1	57/F	Abemaciclib	FUL	21	Yes, left	Yes	No	No	Cough	2	Anemia; leukopenia	CDK4/6 inhibitor discontinued; prednisone	Improved at day 49	Recovered	7
2	71/F	Abemaciclib	EXE	8	Yes, right	No	Yes	No	Cough	3	Pancytopenia	CDK4/6 inhibitor discontinued	Improved at day 36	Recovered	7
3	65/F	Abemaciclib	FUL	6	No	Yes	No	No	Respiratory distress	4	Pancytopenia	CDK4/6 inhibitor discontinued; MPS; IVIG	Transient improvement; death at day 3	Death	7
4	63/F	Abemaciclib	None	2	Yes, left	Yes	No	No	Cough; fever; dyspnea	4	Anemia	CDK4/6 inhibitor discontinued; MPS; IVIG	Improved at day 16	Recovered	6
5	70/F	Abemaciclib	None	2	No	No	No	No	Cough; fever; dyspnea	3	Pancytopenia	CDK4/6 inhibitor discontinued; prednisone	Improved at day 28	Recovered	7
6	59/F	Palbociclib	EXE	30	Yes, right	Yes	No	No	Cough; fever	3	DVT; thrombocytopenia; leukopenia; hepatic injury	CDK4/6 inhibitor discontinued	Stable	Stable	6
7	61/F	Abemaciclib	LET	2	Yes, right	No	No	No	Asymptomatic	1	Pancytopenia; diarrhea	Continued CDK4/6 inhibitor under close monitoring	Stable	Stable	5

ADR, adverse drug reaction; CTCAE, Common Terminology Criteria for Adverse Events; DVT, deep vein thrombosis; EXE, exemestane; F, female; FUL, fulvestrant; ILD, interstitial lung disease; IVIG, intravenous immunoglobulin; LET, letrozole; MPS, methylprednisolone.

The median time from initiation of CDK4/6 inhibitor therapy to symptom onset or radiologic detection was 6 months. The most common respiratory manifestations were cough (n=5), dyspnea (n=3), and fever (n=3). Patient 7 was asymptomatic and was identified during routine radiologic follow-up. Other CDK4/6 inhibitor-related adverse events were observed in all patients, with hematologic abnormalities, mainly varying degrees of myelosuppression, being the most common. Patient 6 developed elevated transaminases and deep vein thrombosis. Five patients experienced grade 3 or higher adverse events.

Laboratory, radiologic, and bronchoscopic findings

Laboratory and radiologic findings are summarized in Table 2. Arterial blood gas analysis on admission showed hypoxemia or respiratory failure in five patients, requiring supplemental oxygen or mechanical ventilation. Hematologic abnormalities were observed in all cases, and pancytopenia developed in four patients during disease progression. C-reactive protein levels appeared to be higher in patients with grade 3 or higher ILD, with a median value of 39.4 mg/L.

Table 2. Laboratory, oxygenation, and chest CT findings in the seven institutional cases.

Case	Respiratory support	FiO2	PaO2, mmHg	PaCO2, mmHg	WBC, ×109/L	Neutrophils, ×109/L	Hb, g/L	Platelets, ×109/L	CRP, mg/L	Autoimmune serology	Infectious work-up	Dominant CT findings	HRCT patter
1	Room air	21%	90.5	39.3	2.97	1.82	93	156	NA	Negative	No pathogen detected	Patchy consolidation; interlobular septal thickening	OP
2	Nasal cannula	29%	65.8	38.4	1.14	0.73	71	80	23.9	Negative	No pathogen detected	GGO; interlobular septal thickening	NSIP
3	Mechanical ventilation	100%	76	32	2.90	2.46	91	93	121.5	ANA 1:40	No pathogen detected	Diffuse bilateral ground-glass opacities	AIP/ARDS-like
4	Non-invasive ventilation	40%	132	71	4.39	3.14	97	201	39.4	Negative	No pathogen detected	Extensive ground-glass opacities	AIP/ARDS-like
5	Room air	21%	57.4	36.6	1.97	1.32	72	128	54.9	Ro-52 positive	No pathogen detected	Diffuse bilateral centrilobular ground-glass nodules	HP-like
6	Room air	21%	57.9	37.6	3.29	2.03	115	353	10.7	Negative	No pathogen detected	Ground-glass interstitial changes	NSIP
7	Room air	21%	85.1	38.8	2.24	1.48	111	75	1.0	ANA 1:40	No pathogen detected	Centrilobular ground-glass nodules and patchy GGO	HP-like

AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome; CRP, C-reactive protein; CT, computed tomography; FiO₂, fraction of inspired oxygen; GGO, ground-glass opacity; Hb, hemoglobin; HP, hypersensitivity pneumonitis; HRCT, high-resolution computed tomography; NA, not available; NSIP, nonspecific interstitial pneumonia; OP, organizing pneumonia; PaCO₂, arterial partial pressure of carbon dioxide; PaO₂, arterial partial pressure of oxygen; WBC, white blood cell count.

Autoimmune serologic testing was performed, including antinuclear antibody, anti-extractable nuclear antigen antibodies, rheumatoid factor, and antineutrophil cytoplasmic antibodies. Some patients also underwent testing for myositis-specific antibodies and scleroderma-related antibodies. Based on clinical evaluation and serologic results, autoimmune disease-related ILD was considered unlikely.

Chest CT findings were categorized into radiologic patterns compatible with drug-related ILD: OP pattern in one patient, NSIP pattern in two patients, HP-like pattern in two patients, and AIP/ARDS-like pattern in two patients (Figure 1). BALF analysis was performed in six patients. No pathogen was detected in available BALF cultures, and lymphocytic predominance was observed in cases with cell differential data, with lymphocytes accounting for 35% to 77% of total cells. Three patients had an increased CD4/CD8 ratio in BALF lymphocytes, as shown in Table 3.

Figure 1.

Representative chest CT patterns of CDK4/6 inhibitor-related ILD. (A,B) 57-year-old female (Patient 1) with abemaciclib therapy duration of 21 months. (A) Baseline CT demonstrates new right lung patchy consolidation with exacerbation of left-sided radiation pneumonitis and emergent left pleural effusion. (B) Significant resolution of right lung consolidation and partial regression of left lung infiltrates/effusion after 49-day glucocorticoid therapy following drug cessation. (C,D) 63-year-old female (Patient 4) at 2-month abemaciclib exposure. (C) Initial CT reveals diffuse left lung GGOs. (D) Marked attenuation of GGOs following 16-day combined corticosteroid and IVIG therapy. (E,F) 70-year-old female (Patient 5) under 2-month abemaciclib monotherapy. (E) Initial CT showing bilateral centrilobular GGO nodules. (F) Near-complete resolution after 4-week corticosteroid regimen. CDK4/6, cyclin-dependent kinase 4/6; CT, computed tomography; GGO, ground-glass opacity; ILD, interstitial lung disease; IVIG, intravenous immunoglobulin.

Table 3. Bronchoalveolar lavage fluid findings in the seven institutional cases.

Case	BALF microbiologic findings	Macrophages, %	Lymphocytes, %	Eosinophils, %	Neutrophils, %	BALF CD4/CD8 ratio
1	No pathogen detected	63	35	0.5	1.5	3
2	No pathogen detected	22	77	0.5	0.5	1.22
3	No pathogen detected	33	64	2.5	0.5	NA
4	Not performed	NA	NA	NA	NA	NA
5	No pathogen detected	NA	NA	NA	NA	6.88
6	No pathogen detected	46	49	2	3	9.52
7	No pathogen detected	50	50	0	0	NA

BALF, bronchoalveolar lavage fluid; NA, not available.

Treatment and outcomes of institutional cases

CDK4/6 inhibitors were discontinued in six patients after ILD was suspected. Four patients received glucocorticoid therapy at different doses, and two of them also received intravenous immunoglobulin. Four patients recovered, two remained clinically stable, and one died. Patient 3 was a representative severe case. This 65-year-old woman developed hypoxemic respiratory failure after 6 months of abemaciclib therapy. Emergency chest CT showed diffuse bilateral GGOs. Despite immediate treatment with high-dose methylprednisolone (500 mg/day) and intravenous immunoglobulin, only transient radiologic improvement was observed (Figure 2), and the patient died of refractory respiratory failure 3 days later. Patient 7 continued CDK4/6 inhibitor therapy under close monitoring and remained clinically stable.

Figure 2.

Clinical course and radiologic changes in a representative severe case. A 65-year-old female (Patient 3) with metastatic breast cancer. (A) Normal baseline chest CT prior to abemaciclib initiation. (B) Rapid progression to diffuse bilateral GGOs with acute hypoxemic respiratory failure at 6-month therapy. (C) Day 2 chest radiograph showing “white lung” pattern despite high-dose methylprednisolone pulse therapy and IVIG. (D) Transient radiological improvement on Day 3. The patient expired from refractory respiratory failure on Day 4. CT, computed tomography; GGO, ground-glass opacity; IVIG, intravenous immunoglobulin.

Literature-based patient-level review

The updated literature-based patient-level review identified 53 published cases of CDK4/6 inhibitor-related ILD that were eligible for the primary literature table. Detailed patient-level information and the corresponding source reference crosswalk are provided in Table 4.

Table 4. Source references for included published cases.

Literature case no.	Source ID	Source citation	CDK4/6 inhibitor	Outcome
1	S1	Gong J, Cho M, Yu KW, et al. A single institution experience with palbociclib toxicity requiring dose modifications. Breast Cancer Res Treat. 2018;168(2):381-387. doi:10.1007/s10549-017-4606-9.	Palbociclib	Recovered
2	S2	Jazieh KA, Budd GT, Dalpiaz N, Abraham J. Can CDK4/6 inhibitors cause fatal lung injury? Expert Rev Anticancer Ther. 2019;19(11):917-919. doi:10.1080/14737140.2019.1674651.	Palbociclib	Recovered
3	S3	Fatal palbociclib-related interstitial pneumonitis [Internet]. Fortune Journals. Available from: http://www.fortunejournals.com/articles/fatal-palbocyclibrelated-interstitial-pneumonitis.html.	Palbociclib	Death
4	S4	Felip E, Llobera L, Perez-Mana C, Quintela D, Guasch I, Margeli M, et al. New Drugs, Old Toxicities: Pneumonitis Related to Palbociclib - A Case Report. Breast Care (Basel). 2020;15(5):548-552. doi:10.1159/000504618.	Palbociclib	Recovered
5	S5	Sarkisian S, Markosian C, Ali Z, Rizvi M. Palbociclib-Induced Pneumonitis: A Case Report and Review of the Literature. Cureus. 2020;12(6):e8929. doi:10.7759/cureus.8929.	Palbociclib	Recovered
6	S6	Stoffaës L, Dumazet A, Deslee G, Mohammad Muttaqi Z, Didier K, Botsen D. Sarcoidosis-like reaction during palbociclib treatment for metastatic breast cancer: A case report. Breast J. 2020;26(6):1239-1241. doi:10.1111/tbj.13800.	Palbociclib	Recovered
7	S7	Okura F, et al. [A Case of Interstitial Pneumonitis Induced by Palbociclib]. Gan To Kagaku Ryoho. 2020. PMID:32541183.	Palbociclib	Death
8	S8	David S, Ho G, Day D, et al. Enhanced toxicity with CDK 4/6 inhibitors and palliative radiotherapy: Non-consecutive case series and review of the literature. Transl Oncol. 2021;14(1):100939. doi:10.1016/j.tranon.2020.100939.	Palbociclib	Death
9	S9	Mathew N, Joel A, Andrews AG, John AO, Singh A. CDK 4/6 inhibitor induced lung injury: a case report and review of literature. Ecancermedicalscience. 2021;15:1245. doi:10.3332/ecancer.2021.1245.	Palbociclib	Death
10	S10	Rivera-Sánchez FE, Moreno-Licea C, Mellado-Orellana R, González-Santiesteban S, de la Peña-López R. Pneumonitis, a common adverse event of cyclin-dependent kinase 4/6 inhibitors: a case report. Int J Sci Rep. 2022;8(3):78-80. doi:10.18203/issn.2454-2156.IntJSciRep20220401.	Palbociclib	Recovered
11	S11	Algwaiz G, Badran AA, Elshenawy MA, Al-Tweigeri T. Ribociclib-Induced Pneumonitis: A Case Report. Breast Care (Basel). 2021;16(3):307-311. doi:10.1159/000507647.	Ribociclib	Recovered
12	S12	Rajendran A, Kunoor A, Pushpa Ragahavan R, Keechilat P. Ribociclib-associated organising pneumonia. BMJ Case Rep. 2021;14(12):e245187. doi:10.1136/bcr-2021-245187.	Ribociclib	Recovered
13	S13	Niciforovic D. Cyclin-dependent kinase 4/6 inhibitors - Ribociclib - Induced ILD/Pneumonitis - The importance of early diagnosis: A case report. Open J Clin Med Case Rep. 2023;9(41). doi:10.52768/2379-1039/2162.	Ribociclib	Recovered
14	S14	Komoda A, et al. [A Case of Liver Abscess during Treatment for Abemaciclib-Induced Interstitial Lung Disease]. Gan To Kagaku Ryoho. 2022. PMID:35046375.	Abemaciclib	Recovered
15	S15	Mitarai Y, Tsubata Y, Hyakudomi M, Isobe T. Drug-Induced Eosinophilic Pneumonia as an Adverse Event of Abemaciclib. Cureus. 2022;14(1):e21741. doi:10.7759/cureus.21741.	Abemaciclib	Recovered
16	S16	Okayasu K, Kawasaki T, Kumagai J, Miyazaki Y. Clinicoradiological course of abemaciclib-induced pneumonitis with histology findings. BMJ Case Rep. 2023;16(5):e254349. doi:10.1136/bcr-2022-254349.	Abemaciclib	Recovered
17	S17	Yamashiro H, Morii N. Abemaciclib-induced lung damage leading to discontinuation in brain metastases from breast cancer: A case report. World J Clin Cases. 2023;11(35):8425-8430. doi:10.12998/wjcc.v11.i35.8425.	Abemaciclib	Recovered
18	S18	Ogata H, Koga C, Yanagihara T, et al. Relapsing eosinophilic pneumonia in a patient with recurrent breast cancer receiving abemaciclib plus endocrine therapy. Respirol Case Rep. 2024;12(3):e01320. doi:10.1002/rcr2.1320.	Abemaciclib	Recovered
19	S19	Al-Ghabban A, Al-Shibany A, Al-Shangiti K, Aseafan M, Latif K, Al-Tweigeri T. Adjuvant abemaciclib-induced pneumonitis: A case report and review of the literature. World Acad Sci J. 2021;3(3). doi:10.3892/wasj.2021.99.	Abemaciclib	Recovered
20	S20	Burrows JT, Lantz R. Chronic Use of Abemaciclib Leading to Breathlessness and Demise: A Case Report. Cureus. 2024;16(7):e64774. doi:10.7759/cureus.64774.	Abemaciclib	Death
21	S21	Wilgus ML, Rosenberg J. CDK4/6 inhibitor-induced pneumonitis: a case report and review of the literature. Proc UCLA Health. 2020;24.	Abemaciclib	Recovered
22	S22	Kaburaki S, Tanaka T, Murata A, et al. The Histopathology of Abemaciclib-Induced Interstitial Lung Disease: A First Case Report With Transbronchial Lung Cryobiopsy. Breast Cancer (Auckl). 2024;18:11782234241301314. doi:10.1177/11782234241301314.	Abemaciclib	Recovered
23	S23	Natarajan S, Natarajan A. Acute on chronic dyspnea in a patient with metastatic breast cancer on abemaciclib. Chest. 2020;158(4):A1150. doi:10.1016/j.chest.2020.08.1050.	Abemaciclib	Recovered
24	S24	Funasaka C, Naito Y, Kusuhara S, et al. Clinical features of CDK4/6 inhibitor-related interstitial lung disease in patients with breast cancer: a case series study. Jpn J Clin Oncol. 2023;53(2):105-114. doi:10.1093/jjco/hyac168.	Palbociclib	Recovered
25	S24		Palbociclib	Recovered
26	S24		Palbociclib	Recovered
27	S24		Palbociclib	Death
28	S24		Palbociclib	Recovered
29	S24		Palbociclib	Recovered
30	S24		Abemaciclib	Recovered
31	S24		Palbociclib	Recovered
32	S24		Abemaciclib	Recovered
33	S24		Palbociclib	Recovered
34	S24		Abemaciclib	Recovered
35	S24		Palbociclib	Recovered
36	S24		Abemaciclib	Recovered
37	S24		Abemaciclib	Recovered
38	S24		Abemaciclib	Recovered
39	S24		Palbociclib	Recovered
40	S24		Palbociclib	Recovered
41	S24		Palbociclib	Death
42	S25	Ilhan N, Dogan A, Erolmez HN, et al. The Incidence and Clinical Characteristics of Interstitial Lung Disease Associated with CDK4/6 Inhibitors in Breast Cancer Patients: A Retrospective Multicenter Study. Medicina. 2025;61(3):549. doi:10.3390/medicina61030549.	Ribociclib	Recovered
43	S25		Ribociclib	Recovered
44	S25		Palbociclib	Death
45	S25		Ribociclib	Recovered
46	S25		Palbociclib	Recovered
47	S25		Palbociclib	Recovered
48	S25		Palbociclib	Recovered
49	S25		Palbociclib	Recovered
50	S25		Palbociclib	Recovered
51	S25		Ribociclib	Recovered
52	S26	Pina-Cabral T, Cavaco P, Miranda H, Fonseca V, Martins A. Palbociclib-Induced Interstitial Lung Disease in Metastatic Breast Cancer: A Case Report and Clinical Considerations. Case Rep Oncol. 2025;18(1):1619-1626. doi:10.1159/000549439.	Palbociclib	Recovered
53	S27	Mustafa Z, Ahmad M, Alkhatib M, et al. Drug-Induced Pneumonitis Secondary to Ribociclib in a Patient With Metastatic Breast Cancer. Cureus. 2026;18(4):e106954. doi:10.7759/cureus.106954.	Ribociclib	Recovered

Source IDs are used only for this table and are independent of the numbered references in the main manuscript.

Together with the seven institutional cases, 60 cases were included in the pooled patient-level summary (Table 5). All patients were female, with ages ranging from 40 to 89 years. The median interval from CDK4/6 inhibitor initiation to ILD onset or diagnosis was 139 days among 59 cases with available latency data. Twenty-four cases (40.7%) occurred within 3 months, 22 cases occurred between >3 months and ≤1 year, and 13 cases occurred after >1 year.

Table 5. Clinical characteristics and outcomes of 60 patients with CDK4/6 inhibitor-related ILD.

Clinical characteristic	Total cases, n	Deaths (n=9)	Non-fatal outcomes (n=51)	P value
CDK4/6 inhibitor type				–
Abemaciclib	22	2	20
Palbociclib	30	7	23
Ribociclib	8	0	8
Time from CDK4/6 inhibitor initiation to ILD onset/diagnosis				0.81
≤3 months	24	4	20
>3 months to ≤1 year	22	4	18
>1 year	13	1	12
HRCT pattern				<0.001
AIP/ARDS-like	8	5	3
NSIP	10	0	10
OP	11	0	11
HP-like	8	1	7
GGO-predominant/unclassified GGO	13	2	11
Others	9	0	9
Not available	1	1	0
Glucocorticoid treatment				0.04
Pulse-dose therapy	5	2	3
High-dose corticosteroids	6	2	4
Moderate- to low-dose corticosteroids	33	5	28
Not used	15	0	15

Drug-specific counts are presented descriptively. No inferential comparison was performed across agents because this selected pooled patient-level sample, derived largely from published case reports and case series, lacked exposed-population denominators and was susceptible to differences in drug utilization, duration of market availability, publication, and reporting practices. P values for time-to-onset category, HRCT pattern, and glucocorticoid treatment regimen were calculated using the Fisher-Freeman-Halton exact test and should be interpreted as exploratory. Time-to-onset data were unavailable for one non-fatal literature case; a 12-month onset was categorized as >3 months to ≤1 year. HRCT information was unavailable for one fatal literature case, which was excluded from the HRCT comparison. Glucocorticoid treatment information was unavailable for one non-fatal literature case. AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome; CDK4/6, cyclin-dependent kinase 4/6; GGO, ground-glass opacity; HP, hypersensitivity pneumonitis; HRCT, high-resolution computed tomography; ILD, interstitial lung disease; NSIP, nonspecific interstitial pneumonia; OP, organizing pneumonia.

Among the 60 cases, 30 patients had received palbociclib, 22 had received abemaciclib, and eight had received ribociclib. Fatal outcomes were reported in nine patients, including seven treated with palbociclib and two treated with abemaciclib; no fatal outcome was reported among ribociclib-treated patients in this selected patient-level sample. These drug-specific counts are presented descriptively and cannot be used to estimate drug-specific incidence or comparative fatality risk because the pooled sample lacked exposed-population denominators and was susceptible to differences in drug utilization, duration of market availability, publication, and reporting practices. The distribution of time-to-onset categories did not differ significantly between fatal and non-fatal cases in the exploratory analysis (P=0.81).

Among cases with available HRCT information, HRCT-pattern distribution differed between fatal and non-fatal cases in the exploratory analysis (P<0.001). Among eight patients with an AIP/ARDS-like pattern, five died and three had non-fatal outcomes. No fatal outcome was observed among patients with NSIP or OP patterns. Fatal outcomes were also reported in patients with HP-like and GGO-predominant or unclassified GGO findings. HRCT information was unavailable for one fatal literature case. These findings suggest that an AIP/ARDS-like pattern was overrepresented among fatal cases, but it was not uniformly fatal.

Glucocorticoid treatment regimens varied substantially across cases. Pulse-dose therapy was used in five patients, high-dose corticosteroids in six patients, moderate- to low-dose corticosteroids in 33 patients, and no glucocorticoid treatment in 15 patients. Glucocorticoid treatment regimen differed between fatal and non-fatal cases in exploratory analysis (P=0.04). However, this finding should be interpreted cautiously because corticosteroid intensity was likely determined by disease severity, and the available data cannot establish the effect of corticosteroid dose on outcome.

Discussion

In patients with breast cancer receiving CDK4/6 inhibitors, new lung opacities may be encountered in routine practice, but their cause is often difficult to determine. Infection, tumor progression, lymphangitic carcinomatosis, radiation-related lung injury, and drug-related ILD may coexist or mimic each other. CDK4/6 inhibitor-related ILD should therefore be approached as a clinical diagnosis supported by treatment timing, CT evolution, exclusion of competing diagnoses, and response to drug interruption, discontinuation, or anti-inflammatory treatment. Although CDK4/6 inhibitors are generally well tolerated, previous reports and the present study indicate that clinically significant ILD can occur during treatment. Pulmonologists and oncologists should remain alert to new respiratory symptoms and new chest imaging abnormalities during CDK4/6 inhibitor therapy.

In the present study, we described seven patients with clinically diagnosed CDK4/6 inhibitor-related ILD from a Chinese center and performed an updated literature-based patient-level review through May 2026. Together with 53 published cases, 60 cases were included in the pooled summary. The clinical manifestations were nonspecific, mainly including cough, dyspnea, fever, and incidentally detected radiologic abnormalities. Chest CT findings were heterogeneous and included GGOs, interlobular septal thickening, patchy consolidations, and radiologic patterns such as NSIP, OP, HP-like pattern, and AIP/ARDS-like pattern. BALF analysis in our institutional cases showed no pathogen detected in available cultures and lymphocytic predominance to varying degrees. The latency from CDK4/6 inhibitor initiation to ILD onset or diagnosis varied widely, with a median of 139 days among cases with available data. Approximately 40% of cases occurred within 3 months, but delayed onset after more than 1 year was also observed. These findings suggest that CDK4/6 inhibitor-related ILD has a broad clinical spectrum and may occur at different stages of treatment.

The pulmonary safety evidence should be considered separately for individual CDK4/6 inhibitors. Abemaciclib-associated ILD has been more extensively characterized, particularly in Japanese post-marketing and real-world studies. ILD-related deaths were reported in the MONARCH 2 and MONARCH 3 trials (9). In Japanese post-marketing surveillance, 82 ILD cases were reported among an estimated 4,700 patients exposed to abemaciclib, including 13 fatal cases; however, these data were based on spontaneous reports and an estimated exposure denominator (10). The NOSIDE study subsequently provided denominator-based, independently adjudicated evidence from routine clinical practice. Among 1,189 Japanese patients with advanced breast cancer who received abemaciclib, 59 patients (5.0%) were adjudicated as having abemaciclib-induced ILD, and eight patients died (0.7% of the exposed cohort). Among the 59 confirmed ILD cases, 64.4% developed within 180 days of treatment initiation. Eastern Cooperative Oncology Group performance status ≥2 [hazard ratio (HR), 5.03; 95% confidence interval (CI): 2.26–11.11] and a history of interstitial pneumonia (HR, 6.49; 95% CI: 3.09–13.70) were identified as independent risk factors (11).

These findings support close monitoring throughout abemaciclib treatment, with particular vigilance during the first 6 months and in patients with impaired performance status or pre-existing interstitial pneumonia. Continued monitoring remains warranted because delayed-onset ILD may also occur. In our institutional series, six of seven patients had received abemaciclib and one patient had pre-existing interstitial pneumonia. However, the predominance of abemaciclib exposure may reflect local prescribing patterns, and the small size of our series does not permit estimation of drug-specific incidence or evaluation of risk factors.

ILD has also been reported with palbociclib and ribociclib, including serious and fatal pulmonary events in clinical trials, pharmacovigilance analyses, and post-marketing reports (2,12-14). Although seven of the nine fatal reports in our selected pooled patient-level sample involved palbociclib and no fatal ribociclib-associated case was identified, these observations cannot establish comparative pulmonary safety. The pooled sample lacked exposed-population denominators and was susceptible to differences in drug utilization, duration of market availability, publication, and reporting practices. Although the extent of agent-specific evidence and monitoring concern differs, the core management principles remain similar across the class: new or worsening respiratory symptoms suggestive of ILD/pneumonitis require prompt evaluation and treatment interruption, with permanent discontinuation for severe disease (7).

Asian populations should not be considered a homogeneous risk category for CDK4/6 inhibitor-related ILD. A Japanese retrospective study reported ILD in 18 of 224 patients treated with CDK4/6 inhibitors. Because eight patients had received both agents, the drug-specific exposure groups were not mutually exclusive: ILD occurred in 12 of 185 patients exposed to palbociclib and six of 47 patients exposed to abemaciclib (3). ILD events have also been reported in Asian subgroup analyses of monarchE and the MONARCH 2 and MONARCH 3 trials (15,16). However, the available Japanese and Chinese data are not directly comparable. A recent nationwide electronic medical record-based study included 3,423 Chinese patients treated with abemaciclib and recorded ILD/pneumonitis in 121 patients (3.5%) overall, including 19 of 619 patients (3.1%) with early breast cancer and 102 of 2,804 patients (3.6%) with advanced breast cancer (17). No new overall safety signals were identified. However, the study did not use the same independent ILD adjudication framework as NOSIDE, systematic information on ILD severity and time to onset was unavailable, and the electronic medical record-based design may have resulted in incomplete capture of adverse events. Therefore, the 5.0% independently adjudicated incidence reported in NOSIDE and the 3.6% recorded prevalence among Chinese patients with advanced breast cancer should not be interpreted as directly comparable incidence estimates.

The apparent variation between Japanese and Chinese reports may reflect differences in diagnostic adjudication, CT surveillance, diagnostic and reporting thresholds, follow-up duration, patient populations, baseline performance status, pre-existing interstitial pneumonia, prior thoracic radiotherapy, prescribing patterns, and pharmacovigilance practices. Genetic background, including possible HLA-related susceptibility, and environmental exposures may also contribute, but these explanations remain unproven. Our institutional series lacked an exposed-population denominator and cannot provide a Chinese incidence estimate. Prospective comparative studies using harmonized surveillance, diagnostic adjudication, follow-up, and reporting criteria are required to determine whether a true Japan-China or broader intra-Asian difference exists.

The diagnosis of CDK4/6 inhibitor-related ILD remains challenging because of the lack of specific clinical manifestations, radiologic features, and biomarkers. In patients receiving CDK4/6 inhibitors, new respiratory symptoms or new chest imaging abnormalities should prompt timely chest CT evaluation and careful differential diagnosis (14). Infection, tumor progression, lymphangitic carcinomatosis, radiation pneumonitis, exacerbation of pre-existing ILD, and lung injury related to other anticancer therapies should all be considered. In our series, most patients had a history of thoracic radiotherapy or lung metastasis, which may complicate the interpretation of pulmonary abnormalities. Therefore, CDK4/6 inhibitor-related ILD should not be diagnosed solely on the basis of new pulmonary opacities during treatment (18). Instead, diagnosis should be based on the temporal relationship with drug exposure, CT evolution, microbiologic evaluation, oncologic assessment, bronchoscopy and BALF findings when available, treatment response, and multidisciplinary clinical judgment.

Chest CT is central to the evaluation of suspected drug-related ILD. Previous studies have shown that common radiologic patterns of anticancer therapy-related ILD include HP-like changes, NSIP, OP, and AIP/ARDS-like pattern (5,6,19). In the present study, among cases with available HRCT information, HRCT-pattern distribution differed between fatal and non-fatal cases in the exploratory analysis. Among eight patients with an AIP/ARDS-like pattern, five died and three had non-fatal outcomes. No fatal outcome was observed among patients with NSIP or OP patterns, whereas deaths were also reported in patients with HP-like and GGO-predominant or unclassified GGO findings. These results suggest that AIP/ARDS-like pattern was overrepresented among fatal cases, but it was not uniformly fatal. Therefore, CT pattern should not be used in isolation to predict prognosis. Rather, diffuse bilateral GGOs, rapidly progressive hypoxemia, extensive consolidation, and an AIP/ARDS-like clinical-radiologic presentation should alert clinicians to a severe acute lung injury phenotype requiring urgent evaluation and active management.

Bronchoscopy and BALF analysis may be useful in selected patients, mainly to evaluate infection and characterize the inflammatory cell profile. In our institutional cases, no pathogen was detected in available BALF cultures, and lymphocytic predominance was observed in several patients. These findings may support an inflammatory lung injury process, but BALF lymphocytosis is not specific for CDK4/6 inhibitor-related ILD and should not be used as a stand-alone diagnostic marker. The decision to perform bronchoscopy should be individualized according to oxygenation status, bleeding risk, and expected diagnostic yield. Transbronchial lung biopsy should not be considered a routine surveillance procedure, but may be considered in selected cases when the diagnosis remains uncertain and the procedure is clinically feasible. Recent studies have suggested that surfactant protein-D may help differentiate drug-induced pneumonia from bacterial pneumonia (20). In a case report by Okayasu et al., peripheral blood surfactant protein-D levels increased during abemaciclib-related ILD and improved with treatment (21). KL-6 may also be useful in assessing drug-related ILD, particularly in cases with alveolar epithelial injury or fibrotic change. Nevertheless, the diagnostic and prognostic value of SP-D and KL-6 in CDK4/6 inhibitor-related ILD requires further investigation.

The optimal treatment strategy for CDK4/6 inhibitor-related ILD has not been established. For suspected drug-related ILD, drug interruption or discontinuation, supportive care, and corticosteroid therapy according to disease severity remain the main treatment approaches (22). In previously reported cases, many mild-to-moderate cases improved after drug discontinuation with or without corticosteroid treatment. However, severe cases may progress rapidly despite high-dose corticosteroids or additional immunomodulatory therapy. In our series, one patient with severe AIP/ARDS-like disease died of refractory respiratory failure despite early treatment with high-dose methylprednisolone and intravenous immunoglobulin. In the pooled analysis, glucocorticoid treatment regimen differed between fatal and non-fatal cases in exploratory analysis. However, this finding should be interpreted cautiously because corticosteroid intensity was likely determined by disease severity. The available data cannot establish whether a specific corticosteroid dose or regimen improves outcome in severe CDK4/6 inhibitor-related ILD.

Evidence regarding continuation or re-challenge of CDK4/6 inhibitors after ILD is limited (3). In published reports, only a small number of patients were re-challenged after recovery, and the safety of this strategy remains unclear. In our series, one patient with mild disease continued CDK4/6 inhibitor therapy under close monitoring and remained clinically stable. However, this case should not be interpreted as formal re-challenge after drug discontinuation and recovery, nor can it establish the safety of continuing or restarting CDK4/6 inhibitors after ILD. Decisions regarding continuation, interruption, discontinuation, or re-initiation of CDK4/6 inhibitors should be individualized according to ILD severity, radiologic progression, oxygenation status, oncologic benefit, availability of alternative treatment options, and the patient’s overall risk.

The pathogenic mechanisms underlying CDK4/6 inhibitor-related ILD remain incompletely understood. Preclinical studies suggest that CDK4/6 inhibition may affect pulmonary immune and inflammatory responses. In a bleomycin-induced lung injury model, palbociclib enhanced pulmonary macrophage and T-cell recruitment, suggesting that inflammatory cell infiltration may contribute to lung injury (23). At the same time, CDK4/6 inhibitors also have immunomodulatory properties, and their effects on lung inflammation and tissue injury may vary across different experimental models (24). Therefore, findings from animal models cannot be directly extrapolated to human CDK4/6 inhibitor-related ILD, and additional clinical and experimental studies are needed to clarify the underlying mechanisms.

This study has several limitations. First, the institutional component was a single-center retrospective case series of only seven patients and was therefore susceptible to selection bias and limited generalizability. Second, the literature-based component was derived largely from published case reports and case series. Severe, unusual, and fatal cases are more likely to be published, whereas mild or asymptomatic cases may be underreported. Accordingly, the nine fatal outcomes among the 60 selected patient-level cases (15.0%) represent the fatal-outcome proportion observed in this selected sample rather than a population-based mortality or case-fatality estimate. The observed proportion may therefore be higher than that in routine clinical practice. Third, the study included neither a control group nor exposed-population denominators and therefore could not estimate the incidence of ILD or compare the relative pulmonary risk of individual CDK4/6 inhibitors. Drug-specific case counts and fatal outcomes are descriptive only. Fourth, the published cases originated from different countries, healthcare systems, and reporting periods. Diagnostic adjudication, microbiologic testing, CT surveillance, HRCT classification, corticosteroid dosing, supportive care, follow-up duration, and outcome reporting were not standardized. HRCT patterns in the literature cases were based on the classifications or descriptions reported in the original publications rather than centralized image review. The limited number of fatal outcomes precluded multivariable adjustment. Consequently, the observed differences in HRCT-pattern distribution and glucocorticoid regimens may have been influenced by disease severity and other unmeasured confounding. Fifth, although the institutional cases were reviewed by experienced pulmonologists, residual diagnostic uncertainty could not be completely eliminated, particularly in patients with prior thoracic radiotherapy, lung metastasis, pre-existing pulmonary abnormalities, or incomplete microbiologic evaluation. The pooled findings should therefore be interpreted as descriptive and hypothesis-generating rather than confirmatory.

Conclusions

CDK4/6 inhibitor-related ILD is rare, but it should be considered when new respiratory symptoms or new CT abnormalities appear during treatment. In this patient-level summary, clinical presentation and CT findings were heterogeneous, and fatal outcomes were more common in AIP/ARDS-like cases but were not limited to this pattern. Evaluation should prioritize HRCT, exclusion of infection, tumor progression and radiation-related lung injury, and severity-based decisions about drug interruption or discontinuation. Larger studies with standardized diagnostic evaluation are needed to define incidence, risk factors and the safety of continuation or re-challenge.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Peking University People’s Hospital, and individual consent was waived due to the retrospective nature of the study.

Data Sharing Statement

Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2026-1560/dss

jtd-18-9-1041-dss.pdf ^{(92.4KB, pdf)}

DOI: 10.21037/jtd-2026-1560