The Relationship Between Pulmonary Spiculation and Central Fibrosis, Along With Other Clinicopathological Features, in Invasive Lung Adenocarcinoma and Its Prognostic Significance.

The Relationship Between Pulmonary Spiculation and Central Fibrosis, Along With Other Clinicopathological Features, in Invasive Lung Adenocarcinoma and Its Prognostic Significance - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Search Log in Primary site navigation Search Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice Thorac Cancer . 2026 Jun 5;17(11):e70319. doi: 10.1111/1759-7714.70319 The Relationship Between Pulmonary Spiculation and Central Fibrosis, Along With Other Clinicopathological Features, in Invasive Lung Adenocarcinoma and Its Prognostic Significance Hanbo Le 1 , Yongkui Zhang 1 , Heping Wang 2 , Yuanyuan Wen 3 , Liyong Qian 3 , Hui He 3, ✉ Author information Article notes Copyright and License information 1 Department of Cardio‐Thoracic Surgery, Lung Cancer Research Center, Zhoushan Hospital of Zhejiang Province, Zhoushan, Zhejiang, China 2 Radiological Diagnostic Center, Zhoushan Hospital of Zhejiang Province, Zhoushan, Zhejiang, China 3 Department of Pathology, Zhoushan Hospital of Zhejiang Province, Zhoushan, Zhejiang, China * Correspondence: Hui He ( hehui577@126.com ) ✉ Corresponding author. Revised 2026 Apr 12; Received 2025 Dec 10; Accepted 2026 May 22; Collection date 2026 Jun. © 2026 The Author(s). Thoracic Cancer published by John Wiley & Sons Australia, Ltd. This is an open access article under the terms of the http://creativecommons.org/licenses/by-nc/4.0/ License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes. PMC Copyright notice PMCID: PMC13238589  PMID: 42244481 ABSTRACT Background Spiculation is a critical radiological feature in invasive adenocarcinoma (IAC). This study explores its pathological association with central fibrosis subtypes to refine prognostic evaluation. Materials and Methods A total of 327 cases of IAC were divided into spiculated ( n = 158) and non‐spiculated ( n = 169) groups. Associations between spiculation, central fibrosis, intratumoral fibrosis, and other factors were analyzed. Multivariate logistic regression was used to identify contributors to spiculation formation. Survival was assessed using Kaplan–Meier curves and Cox regression. Propensity score matching (PSM) was utilized to minimize confounding variables. Results The spiculated group showed higher rates of pleural indentation, central fibrosis > 5 mm, and increased intratumoral fibrosis, collagen content in both the original and matched cohorts (all p < 0.05), while spiculation was associated with elastin content in the original cohort ( p = 0.020), this correlation disappeared after PSM. Multivariate logistic regression identified central fibrosis, collagen content, pleural indentation, tumor size, and lymph node metastasis as independent predictors of spiculation. Spiculation was associated with poorer survival in the original cohort ( p < 0.001), but this significance disappeared after PSM ( p = 0.860). In Cox analysis, spiculation was not an independent prognostic factor. Alveolar collapse > 5 mm was an independent protective factor, while intratumoral fibrosis and lymph node metastasis were adverse prognostic factors. Conclusion Spiculation is closely associated with central fibrosis. It shows a stronger correlation with collagen content than with elastin content. While reflecting underlying aggressiveness, spiculation lacks independent prognostic value. Integrating spiculation with central fibrosis is essential for precise risk stratification. Keywords: alveolar collapse, central collagenous fibrosis, central fibrosis, elastin fiber, pulmonary spiculation Although spiculation in IAC reflects tumor aggressiveness, its prognostic value is associated with central fibrosis subtypes. Spiculation driven by collagenous hyperplasia correlates with poorer outcomes, whereas spiculation associated with extensive alveolar collapse serves as a protective factor. Integrating radiological spiculation with specific fibrotic phenotyping enables precise risk stratification. 1. Introduction As the utilization of CT imaging becomes more widespread and public awareness of health screenings increases, the detection rate of lung nodules has consistently risen. Invasive lung adenocarcinoma is the most prevalent type of lung cancer [ 1 ]. Effective identification of malignant invasive adenocarcinoma (IAC) through CT not only enhances patient survival rates, but also substantially alleviates the societal burden associated with late‐stage cancer treatment [ 2 , 3 , 4 ]. In the radiological evaluation of pulmonary nodules, pulmonary spiculation has long been acknowledged as a hallmark of malignancy. Pulmonary spiculation is characterized by fine, short, radial, unbranched lines that emanate from the tumor periphery and extend into the surrounding lung parenchyma, visible on the lung window, yet unattached to the pleura [ 5 , 6 ]. Prior pathological investigations have attributed spiculation to fibrous bands resulting from the tumor's fibroblastic response or infiltration of neoplastic cells into adjacent bronchovascular or lymphatic pathways [ 5 , 6 ]. The underlying mechanism is principally linked to collagen fiber proliferation within the nodule, with no reported association with elastic fibers [ 7 ]. However, our preliminary clinical observations have revealed a correlation between pulmonary spiculation in the IAC and the presence of central fibrosis (CF), potentially involving the proliferation of elastic fibers. CF, also known as a central scar, frequently occurs in the core or subpleural region of lung adenocarcinoma. It is categorized into two distinct forms: one is characterized by fibrous bands primarily composed of collagen fibers and fibroblast proliferation, often undergoing hyalinization, and the other is primarily composed of elastic fibers, commonly referred to as alveolar collapse (AC) [ 8 , 9 , 10 ] Alveolar collapse denotes a significant reduction in alveolar air volume, leading to tumor cell degeneration and an accumulation of elastic fibers at the tumor center, manifesting as diffuse fibrosis [ 9 ]. While early studies linked central fibrosis to poor prognosis [ 8 , 10 , 11 ], recent perspectives have associated it with more favorable outcomes [ 12 , 13 , 14 ]. This contradiction likely arises from the imprecise understanding and classification of CF, with previous research primarily focusing on collagen‐dominant fibrosis and often overlooking the role of elastic fiber‐composed fibrosis [ 13 ]. As previously highlighted by Takise et al. [ 15 ] and Shimosato et al. [ 8 ], the extent of collagenization in central fibrosis serves as a vital prognostic marker for lung adenocarcinoma, and its clinical importance even outweighs tumor size. However, these early studies on CF in lung adenocarcinoma were conducted prior to the establishment of the current standardized lung adenocarcinoma classification system [ 8 , 9 , 10 , 11 , 15 ]. Consequently, the study cohorts may have included early‐stage lesions such as minimally invasive adenocarcinoma (MIA). Given the significant differences in stromal remodeling and biological behavior between MIA and IAC [ 16 , 17 ], such sample heterogeneity might have masked the true fibrotic characteristics inherent to IAC. More importantly, these investigations primarily focused on the relationship between fibrosis and clinical outcomes (e.g., lymph node metastasis and survival rates) [ 10 , 11 ], failing to explore the radiopathologic mechanistic links between internal fibrotic patterns and external radiological manifestations, such as pulmonary spiculation. Furthermore, from a pathological assessment perspective, earlier studies predominantly defined CF based solely on its presence or absence [ 8 , 15 ]. However, small foci of CF measuring approximately 1–3 mm are frequently observed in lung adenocarcinoma and are generally considered to lack clinical significance [ 11 , 13 ]. This limitation makes it difficult to accurately reflect the complex stromal heterogeneity within IAC. Meanwhile, CF has been largely disregarded by imaging specialists in prior research, and existing imaging studies on spiculation frequently include a variety of heterogeneous tumors, such as squamous cell carcinoma, in which CF is rarely present [ 8 ]. The combination of coarse pathological evaluation and mixed imaging samples has collectively limited the accuracy of research findings regarding pulmonary spiculation. Therefore, the specific relationship between radiological spiculation and central fibrosis subtypes remains to be further elucidated. In light of these considerations, our study focused on a relatively homogeneous cohort of invasive non‐mucinous adenocarcinoma to mitigate the confounding effects observed in prior research. Diverging from previous investigations that primarily examined the relationship between spiculation and nonspecific intratumoral fibrosis [ 5 , 7 ], this study directs its attention toward the structural subtypes of CF. By employing a semiquantitative histomorphological approach, we aimed to provide a more granular analysis of the pathological substrates underlying pulmonary spiculation. We hypothesize that the formation of pulmonary spiculation in IAC is not driven exclusively by isolated collagen proliferation. Instead, it is also modulated by specific subtypes of central fibrosis, appearing either as collagenous fiber hyperplasia or as elastin‐rich alveolar collapse. Furthermore, we posit that the clinical significance of spiculation, particularly regarding its capacity to indicate tumor invasiveness or prognostic risk, may be contingent upon these distinct underlying morphological characteristics of CF. Nevertheless, systematic radiopathologic evidence characterizing these specific associations remains insufficient in the existing literature. This study aimed to meticulously investigate the relationship between pulmonary spiculation and CF. By incorporating propensity score matching (PSM) and multivariable logistic regression, we sought to identify independent pathological drivers of spiculation while evaluating their latent impact on patient prognosis. From a translational medicine perspective, investigating the radiopathologic nexus between spiculation and specific fibrotic patterns holds promise for enhancing the differential diagnosis of pulmonary nodules. This approach enables noninvasive preoperative inference of the tumor's stromal remodeling state, thereby providing a robust rationale for individualized surgical strategies, such as the selection between sublobar resection and standard lobectomy. Furthermore, our findings may provide theoretical support for the development of artificial intelligence (AI)‐assisted diagnostic models, ultimately enhancing the clinical utility of future intelligent systems in interpreting complex lesions. 2. Material and Methods 2.1. Study Participants This retrospective study was approved by the Clinical Research Ethics Committee of Zhoushan Hospital (No. 2022446) with a waiver for informed consent, analyzed the clinical, pathological, and CT data of IAC patients who underwent surgical resection between January 2015 and December 2017. The inclusion criteria were as follows: (1) diagnosis of invasive non‐mucinous adenocarcinoma, (2) solitary primary tumor, (3) availability of all relevant microscopic sections for review, and (4) absence of prior adjuvant or neoadjuvant therapy. The exclusion criteria were as follows: (1) history of other cancers; (2) tumors with a diameter exceeding 3 cm based on gross pathological examination; (3) incomplete or conservative resections; (4) lack of lymph node dissection; (5) tumors diagnosed as mucinous adenocarcinoma or other adenocarcinoma subtypes (e.g., fetal, enteric, or colloid adenocarcinoma); and (6) incomplete imaging data. The flow diagram of the selection process is shown in Figure 1 . A total of 327 patients met the inclusion criteria. The minimum follow‐up duration for the cohort was 79 months, with the final follow‐up conducted on July 31, 2024. FIGURE 1. Open in a new tab Schedule of patient enrolment. 2.2. Minimization of Selection Bias via PSM To reduce potential confounding factors and selection bias between patients with and without pulmonary spiculation, we performed 1:1 PSM. Propensity scores were estimated using a multivariable logistic regression model that incorporated five well‐established clinicopathological covariates: tumor size, histological differentiation, lymph node metastasis, pleural invasion, and spread through air spaces (STAS). 2.3. CT Scan Preoperative chest CT was conducted using a Toshiba Aquilion 64‐slice spiral CT machine. Each subject was positioned supine on the scanning table to ensure that the scanning baseline was perpendicular to the longitudinal axis of the body. Scanning was performed during breath‐holding at full inspiration, encompassing the area from the lung apex to the lung base. The scanning parameters were set to a voltage of 120 kV, current of 110 mA, and collimation of 0.625 × 64 mm, with both slice thickness and spacing maintained at 3 mm. For target reconstruction, a minimum slice thickness of 1 mm and an interval of 0.5 mm were used, employing high‐resolution algorithms (HRCT) and standard reconstruction methods. All CT images were assessed using lung (window width, 1500 HU; window level, −500 HU) and mediastinal windows (window width, 400 HU; window level, 40 HU). Two senior radiologists conducted a blinded review documenting the presence of spiculation and pleural indentation, achieving consensus through discussion when the initial opinions differed. Spiculation was defined as fine, short, radial, and unbranched lines extending from the nodule edge into the surrounding parenchyma and unattached to the pleura [ 5 ]. Pleural indentation was categorized into three types: Type I: A single linear shadow between the tumor and pleura on lung windows, potentially with an associated triangular/funnel‐shaped shadow; type II: Two or more linear shadows between the tumor and pleura, resembling a rabbit ear or V‐sign, with a small triangular shadow near the visceral pleural surface; and type III: Pleural displacement toward the nodule, resulting in indentation or notching, observed on mediastinal windows (Figure 2 ) [ 18 , 19 ]. FIGURE 2. Open in a new tab Radiologic manifestations of lung adenocarcinoma with spiculation and various types of pleural indentation. (A) Lung window CT image displaying classic spiculation with intact pleural morphology and no significant indentation. (B) Spiculation accompanied by type I pleural indentation, characterized by a thin linear opacity connecting the nodule and the pleura. (C) Spiculation with type II pleural indentation, exhibiting two or more linear traction streaks bridging the nodule and the pleura. (D) spiculation associated with type III pleural indentation, featuring focal pleural retraction that forms a characteristic notch or depression. 2.4. Histologic Examination The gross tumor size was measured on unfixed specimens. All the specimens were fixed in 3.7% neutral formalin after surgical resection. To achieve an approximate spatial correlation between radiologic and pathologic findings, pathologists referenced the cross‐sectional CT images to guide the optimal sectioning plane during gross sampling. Tumors with a diameter of less than 3 cm were completely excised. The tumor and surrounding tissue were then cut into tissue samples approximately 2 × 2 cm in size and 0.2–0.3 cm in thickness. If the tumor was located within 2 cm of the visceral pleura, the tumor and adjacent pleura were included in the tissue block (one end representing the tumor and the other end representing the pleura). The samples were then placed in embedding cassettes. The samples were then paraffin‐embedded, dehydrated, sectioned, and stained with hematoxylin and eosin (H&E). Two pathologists (Yuanyuan Wen and Hui He) who were blinded to the radiological images and patient characteristics conducted all histological analyses. Their decisions were reached through a consensus. The analysis was performed on the archived tissue sections. IAC of the lung was diagnosed and classified according to the 2021 WHO criteria, defined as a malignant epithelial tumor with glandular differentiation and/or mucin production, exhibiting an invasive component measuring > 5 mm in greatest dimension. The invasive component is characterized by any histological pattern other than a lepidic growth pattern (e.g., acinar, papillary, micropapillary, or solid patterns) or the presence of tumor cells infiltrating the myofibroblastic stroma. Histological grading is based on the predominant histological pattern combined with the presence of high‐grade patterns, which include solid, micropapillary, complex glandular, and cribriform. The system classifies tumors into three grades: grade 1 (well‐differentiated; predominantly lepidic with or without < 20% high‐grade pattern), grade 2 (moderately differentiated; predominantly papillary or acinar with or without < 20% high‐grade pattern), and grade 3 (poorly differentiated; tumors with ≥ 20% high‐grade pattern) [ 20 ]. 2.5. The Victoria Blue Elastin Fiber Staining and Evaluation The extracellular matrix of normal lung tissue is primarily composed of elastic and collagen fibers, which are also present in the interstitium of lung tumors. In the Victoria blue staining method, elastic fibers were stained deep blue, whereas collagen fibers appeared pink. Pathological sections of the lung tumors were processed according to the instructions provided with the Victoria Blue Elastic Fiber Kit (Beijing Solarbio Science & Technology Co. Ltd.). The assessment followed the methodology described by Yamashita et al., and was adjusted based on practical circumstances [ 21 ]. Intratumoral fibrosis, as well as individual elastic and collagen fiber contents, was graded into three levels based on their respective proportions within the tumor area: Grade 1 (< 20%), grade 2 (20%–40%), and grade 3 (> 40%). Intratumoral fibrosis was specifically evaluated as the combined presence of both elastic and collagen fibers. The proportions of elastic and collagen fibers were assessed in five different, nonoverlapping random fields of view at 200× magnification for each tumor. For each specimen, the fibrous components were semiquantitatively graded (grades 1–3) across these five distinct fields, and the median grade was selected to represent the overall fibrotic landscape of the tumor. 2.6. Assessment of Central Fibrosis Central fibrosis refers to the presence of extensive fibrotic areas in the central region or beneath the pleura in lung adenocarcinomas. It is currently classified into two types: one is characterized by the proliferation of collagen fibers and fibroblasts, which may even display translucency [ 11 ], and the other is predominantly composed of elastic fibers [ 9 ] (Figure 3 ). In this study, a 5 mm cutoff was used to categorize the cases into the following groups [ 11 , 13 ]: central fibrosis > 5 and ≤ 5 mm, central collagenous fibrosis > 5 and ≤ 5 mm, and alveolar collapse > 5 and ≤ 5 mm. FIGURE 3. Open in a new tab Imaging‐pathological correlation of invasive adenocarcinoma. (A–C) Derived from the same patient, who was ultimately diagnosed with poorly differentiated invasive adenocarcinoma. (A) Lung window CT image displaying spiculation. (B) Pathological section exported from K‐viewer software (1× HE), with the dashed box demarcating a central fibrotic area greater than 5 mm. (C) Elastin fiber staining of the tumor, achieved with Victoria Blue dye, where blue corresponds to elastin fibers and red to collagen fibers (1×). This figure highlights central collagenization, with collagen fibers predominating over elastin fibers. (D–F) Derived from the same patient and ultimately diagnosed with moderately differentiated invasive adenocarcinoma. (D) Also exhibits spiculation. (E) A pathological slide highlights a central fibrotic area greater than 5 mm within the dashed box (1× HE). (F) Reveals alveolar collapse, accompanied by a relative abundance of elastin fibers (1×). (B, C, E, F) Scanned by digital pathology slice scanner (KF‐PRO‐005, KFBIO, Ningbo, CHN), and then analyzed with K‐viewer software (KFBIO, Ningbo, CHN). 2.7. Statistical Analysis Categorical data were analyzed using the chi‐square test or Fisher's exact test, andthe results were expressed as frequencies and percentages. Continuous variables were compared using Student's t ‐test. PSM was implemented to balance clinicopathological characteristics between patients with and without pulmonary spiculation, as previously described. A nearest‐neighbor matching algorithm was employed with a stringent caliper width of 0.1; furthermore, exact matching was mandated for the presence of STAS. The balance of covariates was evaluated using the standardized mean difference (SMD), where an SMD < 0.1 indicated an adequate balance. All PSM‐related analyses were conducted using the “MatchIt” package in R software. The chi‐square test or Student's t ‐test was utilized to analyze correlations between pulmonary spiculation and other pathological features (e.g., central fibrosis) in both the original and PSM‐matched cohorts. In the original cohort, variables showing statistical significance in the univariate analysis were candidates for the multivariable logistic regression model. To ensure model stability and address potential multicollinearity among pathological features, collinearity diagnostics were performed using variance inflation factors (VIFs). A backward stepwise procedure based on the likelihood ratio test (backward: LR) was then employed to identify independent predictors of pulmonary spiculation. The entry and removal criteria were set at p < 0.05 and p > 0.10, respectively. The model's robustness was evaluated by ensuring that all retained variables maintained a VIF < 5. Survival curves were estimated using the Kaplan–Meier method, and differences were compared via the log‐rank test across both cohorts. Additionally, univariate and multivariate Cox regression analyses were conducted in the original cohort to evaluate survival‐related variables. For the multivariate analysis, a backward stepwise elimination method (backward: LR) was employed to identify independent prognostic indicators. The results were presented as hazard ratios (HRs) with 95% confidence intervals (CIs). Statistical analyses were conducted using SPSS version 22.0 and R version 4.5.3. Two‐tailed p ‐values < 0.05 were considered statistically significant. 3. Results 3.1. The Relationship Between Pulmonary Spiculation and Central Fibrosis Along With Other Clinical Pathological Features in the Original Cohort In the original cohort, a total of 327 cases of IAC were included, consisting of 125 male and 202 female patients, with ages ranging from 34 to 84 years (mean ± SD: 61.63 ± 9.13 years). Tumor diameters varied between 0.6 and 3 cm (mean ± SD: 1.57 ± 0.58 cm). Among these cases, 158 showed spiculation, whereas 169 did not. Table 1 provides an overview of the associations between spiculation, central fibrosis, and other clinicopathological features. Compared with the non‐spiculated group, the spiculated group had larger tumor diameters, higher probabilities of pleural indentation signs II and III, and a higher probability of central fibrosis > 5 mm, alveolar collapse > 5 mm, and central collagenous fibrosis > 5 mm ( p < 0.05). The spiculated group also exhibited higher degrees of intratumoral fibrosis, elastic fiber formation, and collagen fiber accumulation than the non‐spiculated group. Furthermore, the spiculated group had higher rates of lymph node metastasis, pleural invasion, necrosis, and occurrence of STAS than the non‐spiculated group. No statistically significant differences in sex, age, or tumor location were observed between the spiculated and non‐spiculated groups. TABLE 1. Univariate analysis of pulmonary spiculation and central fibrosis with other clinicopathological features in the original cohort. Total ( N = 327) Spiculation χ 2 or t p Present ( n = 158) Absent ( n = 169) Sex Male 125 (38.23%) 61 (38.61%) 64 (37.87%) 0.019 0.891 Female 202 (61.77%) 97 (61.39%) 105 (62.13%) Age, mean ± SD 61.63 ± 9.13 62.04 ± 8.23 61.25 ± 9.94 0.791 0.43 Tumor location Left upper lobe 93 (28.44%) 48 (30.38%) 45 (26.63%) 1.318 0.858 Left lower lobe 41 (12.54%) 20 (12.66%) 21 (12.43%) Right upper lobe 110 (33.64%) 54 (34.18%) 56 (33.14%) Right middle lobe 29 (8.87%) 12 (7.59%) 17 (10.06%) Right lower lobe 54 (16.51%) 24 (15.19%) 30 (17.75%) Tumor size, mean ± SD 1.57 ± 0.58 1.81 ± 0.56 1.35 ± 0.52 7.692 < 0.001 Tumor size ≤ 1 cm 68 (20.80%) 13 (8.23%) 55 (32.54%) 45.786 < 0.001 > 1, ≤ 2 cm 198 (60.55%) 97 (61.39%) 101 (59.76%) > 2, ≤ 3 cm 61 (18.65%) 48 (30.38%) 13 (7.69%) Pleural indentation Absent 116 (35.47%) 26 (16.46%) 90 (53.25%) 55.841 < 0.001 I 76 (23.24%) 38 (24.05%) 38 (22.49%) II 84 (25.69%) 58 (36.71%) 26 (15.38%) III 51 (15.60%) 36 (22.78%) 15 (8.88%) Central fibrosis ≤ 5 mm 242 (74.01%) 95 (60.13%) 146 (86.39%) 29.06 < 0.001 > 5 mm 85 (25.99%) 63 (39.87%) 23 (13.61%) Alveolar collapse ≤ 5 mm 262 (80.12%) 114 (72.15%) 148 (87.57%) 12.194 < 0.001 > 5 mm 65 (19.88%) 44 (27.85%) 21 (12.43%) Central collagenous fibrosis ≤ 5 mm 307 (93.88%) 140 (88.61%) 167 (98.82%) 14.821 < 0.001 > 5 mm 20 (6.12%) 18 (11.39%) 2 (1.18%) Intratumoral fibrosis Grade 1 137 (41.90%) 34 (21.52%) 103 (60.95%) 59.686 < 0.001 Grade 2 88 (26.91%) 48 (30.38%) 40 (23.67%) Grade 3 102 (31.19%) 76 (48.10%) 26 (15.38%) Elastic fiber grading Grade 1 226 (69.11%) 98 (62.03%) 128 (75.74%) 7.783 0.02 Grade 2 47 (14.37%) 26 (16.46%) 21 (12.43%) Grade 3 54 (16.51%) 34 (21.52%) 20 (11.83%) Collagen fibers grading Grade 1 218 (66.67%) 77 (48.73%) 141 (83.43%) 48.266 < 0.001 Grade 2 70 (21.41%) 47 (29.75%) 23 (13.61%) Grade 3 39 (11.93%) 34 (21.52%) 5 (2.96%) Lymph node metastasis Present 23 (7.03%) 21 (13.29%) 2 (1.18%) 18.307 < 0.001 Absent 304 (92.97%) 137 (86.71%) 167 (98.82%) Pleural invasion Present 33 (10.09%) 28 (17.72%) 5 (2.96%) 19.614 < 0.001 Absent 294 (89.91%) 139 (87.97%) 164 (97.04%) Histological differentiation Grade 1 71 (21.71%) 21 (13.29%) 50 (29.59%) 18.354 < 0.001 Grade 2 223 (68.20%) 113 (71.52%) 110 (65.09%) Grade 3 33 (10.09%) 24 (15.19%) 9 (5.33%) Necrosis Present 37 (11.31%) 32 (20.25%) 5 (2.96%) 24.34 < 0.001 Absent 290 (88.69%) 126 (79.75%) 164 (97.04%) STAS Present 44 (13.46%) 34 (21.52%) 10 (5.92%) 17.069 < 0.001 Absent 283 (86.54%) 124 (78.48%) 159 (94.08%) Open in a new tab 3.2. Correlation Between Spiculation and Pathological Features After PSM After 1:1 PSM, a balanced cohort comprising 190 patients (95 matched pairs) was established. Statistical analysis revealed no significant differences between the two groups in terms of demographic characteristics (age, sex) and tumor location ( p > 0.05). Additionally, the key prognostic indicators including tumor size, lymph node status, histological differentiation, and STAS were well‐matched between the groups ( p = 0.978–1.000, SMD < 0.1), as shown in Table 2 . This indicates that the two groups were highly comparable in terms of baseline tumor burden and biological invasiveness. TABLE 2. Association between pulmonary spiculation and clinical pathological features after propensity score matching. Spiculation χ 2 or t p SMD Present ( n = 95) Absent ( n = 95) Sex Male 34 (35.79%) 36 (37.89%) 0.09 0.764 0.044 Female 61 (64.21%) 59 (62.11%) Age, mean ± SD 62.29 ± 8.29 61.65 ± 9.47 0.497 0.62 0.072 Tumor location Left upper lobe 34 (35.79%) 29 (30.53%) 2.602 0.626 0.236 Left lower lobe 10 (10.53%) 9 (9.47%) Right upper lobe 26 (27.37%) 31 (32.63%) Right middle lobe 10 (10.53%) 6 (6.32%) Right lower lobe 15 (15.79%) 20 (21.05%) Tumor size, mean ± SD 1.62 ± 0.50 1.54 ± 0.55 1.074 0.284 0.156 Tumor size ≤ 1 cm 12 (12.63%) 12 (12.63%) 0.044 0.978 0.031 > 1, ≤ 2 cm 69 (72.63%) 70 (73.68%) > 2, ≤ 3 cm 14 (14.74%) 13 (13.68%) Pleural indentation Absent 17 (17.89%) 44 (46.32%) 18.055 < 0.001 0.648 I 26 (27.37%) 20 (21.05%) II 33 (34.74%) 20 (21.05%) III 19 (20.00%) 11 (11.58%) Central fibrosis ≤ 5 mm 60 (63.16%) 79 (83.16%) 9.676 0.003 0.463 > 5 mm 35 (36.84%) 16 (16.84%) Alveolar collapse ≤ 5 mm 65 (68.42%) 81 (85.26%) 7.572 0.01 0.407 > 5 mm 30 (31.58%) 14 (14.74%) Central collagenous fibrosis ≤ 5 mm 90 (94.74%) 93 (97.89%) 0.444 a 0.168 > 5 mm 5 (5.26%) 2 (2.11%) Intratumoral fibrosis Grade 1 29 (30.53%) 50 (52.63%) 10.608 0.005 0.486 Grade 2 30 (31.58%) 25 (26.32%) Grade 3 36 (37.89%) 20 (21.05%) Elastic fiber grading Grade 1 55 (57.89%) 66 (69.47%) 2.773 0.25 0.243 Grade 2 20 (21.05%) 15 (15.79%) Grade 3 20 (21.05%) 14 (14.74%) Collagen fibers grading Grade 1 60 (63.16%) 77 (81.05%) 7.63 0.022 0.409 Grade 2 24 (25.26%) 13 (13.68%) Grade 3 11 (11.58%) 5 (5.26%) Lymph node metastasis Present 3 (3.16%) 2 (2.11%) 1 a 0.066 Absent 92 (96.84%) 93 (97.89%) Pleural invasion Present 3 (3.16%) 5 (5.26%) 0.721 a 0.105 Absent 92 (96.84%) 90 (94.74%) Histological differentiation Grade 1 17 (17.89%) 16 (16.84%) 0.037 0.982 0.028 Grade 2 72 (75.79%) 73 (76.84%) Grade 3 6 (6.32%) 6 (6.32%) Necrosis Present 10 (10.53%) 5 (5.26%) 1.81 0.179 0.196 Absent 85 (89.47%) 90 (94.74%) STAS Present 6 (6.32%) 6 (6.32%) 0 1 < 0.001 Absent 89 (93.68%) 89 (93.68%) Open in a new tab Note: Data are presented as n (%) or mean ± SD. a Indicates p ‐values calculated using Fisher’s exact test due to small expected cell counts (expected count < 5). Abbreviations: PSM, propensity score matching; SMD, standardized mean difference; SD, standard deviation; STAS, spread through air spaces. After matching for clinical and prognostic factors, the pulmonary spiculation sign remained closely related to several distinct pathological features. Specifically, the spiculation group exhibited a markedly higher prevalence of central fibrosis (> 5 mm: 36.8% vs. 16.8%, p = 0.003) and alveolar collapse (> 5 mm: 31.6% vs. 14.7%, p = 0.010). The spiculation group demonstrated a significantly higher intratumoral fibrosis grade ( p = 0.005) and increased collagen fiber grading ( p = 0.022). Additionally, this group was more likely to exhibit the pleural indentation sign ( p < 0.001). In contrast, there was no statistically significant difference in elastic fiber grade ( p = 0.250) between the two groups, suggesting that it may play a secondary role in the formation of the spiculation sign. While necrosis ( p = 0.179) and central collagenous fibrosis ( p = 0.444) did not reach statistical significance, this may be attributable to limited event rates in the sample, resulting in insufficient statistical power. 3.3. Multivariable Logistic Regression Analysis of Independent Predictors for Spiculation Initially, multivariable logistic regression was performed using the Enter method; however, significant multicollinearity was observed among several pathological indicators, with VIFs for intratumoral fibrosis, elastic fiber, and collagenous fiber reaching as high as 29.896. Notably, Central collagenous fibrosis was excluded from the full equation due to severe statistical redundancy (Table S1 ). To eliminate these redundant interferences, a backward stepwise procedure based on the likelihood ratio test (backward: LR) was employed to optimize the model. After systematically removing nonsignificant and collinear variables, the final optimized model demonstrated high stability. The model's robustness was confirmed as all retained independent predictors maintained a VIF < 5 (ranging from 1.05 to 2.42), indicating no significant multicollinearity interference. The multivariable analysis identified five independent predictors of pulmonary spiculation (Table 3 ). Among the pathological features, central fibrosis ( p = 0.010), collagen fiber grading ( p = 0.001), and pleural indentation ( p < 0.001) remained significantly associated with spiculation. Furthermore, clinical indicators including tumor size > 2 cm ( p = 0.017) and lymph node metastasis ( p = 0.038) were also confirmed as independent predictors. TABLE 3. Results of multivariate logistic regression analysis on pulmonary spiculation. Variables B (SD) Exp ( B ) 95% CI p VIF Central fibrosis 0.853 2.347 1.224–4.5 0.01 1.181 Collagen fibers grading (Grade 1) < 0.001 1.194 Collagen fibers grading (Grade 2) 1.234 3.433 1.712–6.884 0.001 Collagen fibers grading (Grade 3) 1.889 6.610 2.251–19.412 0.001 Pleural indentation < 0.001 1.227 Pleural indentation (I) 1.318 3.737 1.833–7.62 < 0.001 Pleural indentation (II) 1.633 5.119 2.445–10.717 < 0.001 Pleural indentation (III) 1.525 4.593 1.936–10.897 0.001 Tumor size (≤ 1 cm) 0.059 1.345 Tumor size (> 1, ≤ 2 cm) 0.602 1.825 0.854–3.903 0.121 Tumor size (> 2, ≤ 3 cm) 1.24 3.455 1.243–9.603 0.017 Lymph node metastasis 1.824 6.196 1.106–34.716 0.038 1.149 Constant −2.405 0.090 < 0.001 Open in a new tab 3.4. Prognostic Impact of Clinicopathological and Morphological Features in the Original Cohort In the original cohort, Kaplan–Meier survival analysis revealed that invasive lung adenocarcinoma with spiculation was associated with a significantly poorer overall survival ( p < 0.001; Figure 4A ). No significant difference in survival was found between the group with central fibrosis ≤ 5 mm and the group with central fibrosis > 5 mm ( p = 0.910; Figure 4B ). In contrast, the group with alveolar collapse > 5 mm showed significantly better overall survival than the group with alveolar collapse ≤ 5 mm ( p = 0.015; Figure 4C ). On the other hand, the group with central collagenous fibrosis > 5 mm demonstrated notably worse overall survival than the ≤ 5 mm group ( p < 0.001; Figure 4D ). When intratumoral fibrosis in IAC was classified into three grades, grade 3 indicated the poorest prognosis, grade 1 was the best, and grade 2 was intermediate ( p < 0.001; Figure 4E ). Additionally, when IAC was graded based on elastin content, the prognostic differences among the three grades were not statistically significant ( p = 0.710; Figure 4F ). In contrast, when categorized by collagen content, grade 3 again presented the worst prognosis; grade 1, the best; and grade 2, intermediate ( p < 0.001; Figure 4G ). Furthermore, when pleural indentation was classified into four types, patients with invasive pulmonary adenocarcinoma had a more favorable prognosis if there was no pleural indentation or if it was classified as type I; however, those with type III pleural indentation experienced the most adverse prognosis ( p = 0.036; Figure 4H ). Moreover, when IAC was classified into three groups based on tumor size (≤ 1, > 1 cm and ≤ 2, > 2 cm, and ≤ 3 cm), the results indicated that larger tumors were correlated with worse prognosis ( p < 0.001; Figure 4I ). When histological types of tumors were categorized based on differentiation, lower levels of differentiation corresponded to poorer prognoses ( p < 0.001; Figure 4J ). Additionally, patients with invasive pulmonary adenocarcinoma accompanied by pleural invasion ( p < 0.001; Figure 4K ) or lymph node metastasis ( p < 0.001; Figure 4L ) had a significantly poorer prognosis. FIGURE 4. Open in a new tab Kaplan–Meier 5‐year survival analysis was conducted for patients with lung invasive adenocarcinoma ≤ 3.0 cm (A) Overall survival was compared based on the presence or absence of pulmonary spiculations. (B) Overall survival for invasive lung adenocarcinoma was also evaluated concerning the size of central fibrosis (CF). (C) Additionally, comparisons were made regarding the size of alveolar collapse (AC). (D) The size of central collagenous fibrosis (CC) was similarly analyzed. Further comparisons were performed based on the degree of intratumoral fibrosis (E), the degree of elastic fibers (F), and the degree of collagenous fibers (G). Other parameters compared included pleural indentation type (H), tumor size (I), histological differentiation (J), pleural invasion (K), and lymph node metastasis (L) in relation to overall survival for patients with lung invasive adenocarcinoma. To further investigate the impact of alveolar collapse on the prognostic value of tumor size, a stratified subgroup analysis was performed (Table 4 ). For patients with tumor diameters between 2.0 and 3.0 cm ( n = 61), Kaplan–Meier analysis revealed that those with alveolar collapse > 5 mm exhibited a significantly better 5‐year overall survival compared to those with ≤ 5 mm (100% vs. 75.0%, Log‐rank p = 0.011, Figure S1 ). In the ≤ 2.0 cm subgroup, a favorable prognosis was observed in the > 5 mm group (100% vs. 93.6%, Log‐rank p = 0.094), although the difference did not reach statistical significance. TABLE 4. Prognostic impact of alveolar collapse on 5‐year overall survival stratified by tumor size. Tumor size (cm) Alveolar collapse Patients ( n ) 5‐year OS rate (%) Log‐rank p ≤ 2 cm ≤ 5 mm 222 93.60% 0.094 > 5 mm 44 100.00% > 2, ≤ 3 cm ≤ 5 mm 40 75.00% 0.011 > 5 mm 21 100.00% Open in a new tab 3.5. Identification of Independent Prognostic Factors via PSM and Multivariable Cox Analysis Univariate Cox regression analysis initially indicated that spiculation was associated with an increased mortality risk (HR = 3.276, p = 0.001; Table 5 ). However, spiculation failed to maintain its prognostic independence in the multivariable Cox model, which was strictly optimized using the backward stepwise elimination method (backward: LR) and further cross‐verified by the full model incorporating all covariates (Table S2 ). This was further validated by KM analysis in a PSM cohort; after balancing baseline clinicopathological characteristics, the prognostic impact of spiculation was no longer observed ( p = 0.860, Figure 5 ). According to the multivariable analysis (Table 5 ), alveolar collapse greater than 5 mm emerged as an independent protective factor for overall survival (HR = 0.081, p = 0.001). Furthermore, intratumoral fibrosis and lymph node metastasis have been identified as adverse prognostic factors. TABLE 5. Univariate and multivariate Cox regression analysis for survival for invasive adenocarcinoma of the lung ≤ 3.0 cm. Variables Univariate analysis Multivariate analysis HR (95% CI) p HR (95% CI) p Gender (male vs. female) 1.285 (0.682–2.42) 0.437 Spiculation (present vs. absent) 3.276 (1.596–6.723) 0.001 Pleural indentation Absent 0.052 Type I 0.445 (0.145–1.364) 0.157 Type II 1.171 (0.525–2.614) 0.700 Type III 2.083 (0.933–4.65) 0.073 Central_fibrosis (> 5 vs. ≤ 5 mm) 0.957 (0.466–1.964) 0.905 Alveolar collapse (> 5 vs. ≤ 5 mm) 0.203 (0.049–0.843) 0.028 0.081 (0.019–0.349) 0.001 Central collagenous fibrosis (> 5 vs. ≤ 5 mm) 4.474 (2.054–9.742) < 0.001 Intratumoral fibrosis Grade 1 < 0.001 < 0.001 Grade 2 2.748 (0.999–7.562) 0.050 1.962 (0.676–5.693) 0.215 Grade 3 5.847 (2.38–14.363) < 0.001 6.006 (2.17–16.622) 0.001 Elastic fiber Grade 1 0.711 Grade 2 0.812 (0.314–2.097) 0.667 Grade 3 0.692 (0.268–1.787) 0.447 Collagen fiber Grade 1 < 0.001 Grade 2 4.726 (2.098–10.643) < 0.001 Grade 3 11.269 (5.055–25.123) < 0.001 Histologic grade Grade 1 < 0.001 Grade 2 1.393 (0.525–3.694) 0.506 Grade 3 7.651 (2.722–21.504) < 0.001 Necrosis (present vs. absent) 7.799 (4.134–14.711) < 0.001 Tumor size ≤ 1 cm 0.002 0.097 > 1, ≤ 2 cm 3.821 (0.899–16.251) 0.069 2.25 (0.498–10.161) 0.292 > 2, ≤ 3 cm 9.274 (2.121–40.559) 0.003 4.169 (0.851–20.419) 0.078 STAS (present vs. absent) 4.711 (2.471–8.985) < 0.001 Lymph node metastasis (present vs. absent) 8.324 (4.205–16.479) < 0.001 3.571 (1.728–7.381) 0.001 Visceral pleural invasion (present vs. absent) 4.606 (2.332–9.097) < 0.001 Open in a new tab FIGURE 5. Open in a new tab Prognostic significance of pulmonary spiculation in the PSM‐matched cohort. Kaplan–Meier survival curves demonstrate that there was no significant difference in prognosis between the spiculated and non‐spiculated groups after propensity score matching ( p = 0.862). 4. Discussion Pulmonary spiculation, an imaging manifestation observed on CT scans, is characterized by irregular, sharp, and linear protrusions at the periphery of the tumor [ 5 ]. Predominantly observed in IAC, this hallmark correlates strongly with adverse clinicopathologic features. Reflecting the tumor's underlying invasive growth pattern, spiculation serves as a critical noninvasive indicator of malignant potential [ 5 , 22 ]. It frequently co‐occurs with pleural indentation [ 23 ], a finding that is corroborated by our study. The pathological mechanisms underlying these phenomena are similar. Current evidence suggests that the formation of both pulmonary spiculation and pleural indentation is primarily attributed to intratumoral fibrosis [ 5 , 6 ]. However, the potential relationship between pulmonary spiculation and central fibrosis remains unclear. Central fibrosis is a prevalent characteristic of lung adenocarcinoma and is commonly categorized as alveolar collapse and central collagenous fibrosis [ 10 ]. In this study, we stratified alveolar collapse and central collagenous fibrosis into two groups based on a 5 mm threshold. The results demonstrated that in invasive lung adenocarcinoma with spiculation, the probability of central fibrosis > 5 mm, alveolar collapse > 5 mm, and central collagenous fibrosis > 5 mm was significantly elevated ( p < 0.05). This robust association largely persisted even after PSM was employed to minimize potential confounders. Although the association between central collagenous fibrosis (> 5 mm) and spiculation did not reach statistical significance following PSM ( p = 0.444), we contend that this does not necessarily imply a lack of biological correlation. This outcome is likely attributable to the relatively low prevalence of this specific subtype within the IAC spectrum (only 6.12%), where the limited sample size may have constrained the statistical power of our quantitative analysis, thereby potentially masking the contribution of central collagenous fibrosis in driving marginal invasive alterations such as spiculation. Such findings underscore that spiculation is far from a stochastic imaging finding; instead, it serves as a faithful radiographic reflection of the intricate fibrotic remodeling occurring within the tumor stroma. While pulmonary spiculation correlates with both central collagenous fibrosis and alveolar collapse, these two fibrotic subtypes exhibit markedly divergent prognostic implications. Univariate Cox analysis identified spiculation and central collagenous fibrosis as predictors of unfavorable clinical outcomes, whereas alveolar collapse emerged as a potential protective factor. This prognostic divergence likely reflects the inherent pathological heterogeneity underlying the spiculation phenotype rather than an intrinsic contradiction in our findings. To gain a more nuanced understanding of this intricate relationship, we further explored the distinct pathophysiological mechanisms of central collagenous fibrosis and alveolar collapse and their differential influences on tumor behavior. Alveolar collapse denotes the aggregation and condensation of extensive regions of elastic fibers within the central area. In contrast to other tissues, normal pulmonary parenchyma contains a substantial quantity of elastic fibers. When hypoxia or necrosis occurs within a neoplasm, tumor cell proliferation is hindered, leading to intratumoral collapse [ 24 , 25 , 26 ]. Based on univariate analysis, spiculation demonstrated a significant association with alveolar collapse across both the primary and post‐PSM cohorts. However, in multivariate logistic regression, alveolar collapse and intratumoral elastic fiber grading did not emerge as independent risk factors for spiculation formation. Alveolar collapse is regarded as a favorable prognostic marker, possibly due to its underlying molecular mechanisms and the tumor microenvironment with relatively lower invasiveness [ 13 ]. While larger tumor diameters are traditionally correlated with poorer outcomes [ 27 , 28 ]. Previous studies, along with our data, indicate that extensive alveolar collapse can lead to an “overestimated” gross tumor size. This increase in size does not necessarily reflect a corresponding enhancement in malignancy potential [ 13 , 14 ]. In our subgroup analysis of tumors between 2.0 and 3.0 cm (Table 4 ), we found that the group with alveolar collapse > 5 mm had a significantly better prognosis than the ≤ 5 mm group (100.00% vs. 75.00%, p = 0.011). Multivariate Cox regression analysis further reinforced this observation, confirming that alveolar collapse > 5 mm is a robust independent protective factor (HR = 0.081, p = 0.001; Table 5 ). Furthermore, although the detection rate of spiculated signs, a radiological feature of invasiveness, tends to increase with the gross tumor size (Table 1 ), significant alveolar collapse seems to modulate the associated risk. Therefore, we propose the existence of a “prognostic balance” between these morphological features: true malignant expansion indicates increased biological invasiveness, while extensive alveolar collapse, which leads to spiculation, acts as a key modifying factor, potentially “counteracting” the prognostic risk typically associated with an increase in tumor diameter. Central collagenous fibrosis denotes the host's proliferative response of connective tissue to tumor invasion, resulting in the characteristic stellate morphology of the tumor. In contrast to noncancerous scars, such as those observed in tuberculosis, this proliferative response is characterized by an abundance of host myofibroblasts and a distinctive collagen structure with a notably high concentration of type V collagen [ 29 ]. Myofibroblasts are known to possess contractile properties; consequently, the extensive presence of these cells within the central collagenous fibrosis accounts for the observed pleural retraction and spiculated features of the lesion [ 30 , 31 ]. Although central collagenous fibrosis may superficially augment tumor size, it is consistently associated with poorly differentiated histological subtypes and lymph node metastasis, which reflects a higher degree of malignancy [ 10 , 11 ]. Additionally, compared to the relatively common finding of alveolar collapse, central collagenous fibrosis is infrequent within the overall spectrum of IAC, typically observed only in cases with prominent stromal fibrosis and high‐grade invasive patterns. Hence, central collagenous fibrosis serves as an indicator of poor prognosis. While univariate analysis suggested a potential association between spiculation and elastin‐rich alveolar collapse, our multivariate logistic regression analysis demonstrated that collagen fiber grading was a robust independent predictor of spiculation formation. This indicates that the development of spiculation is more significantly correlated with collagen content than with elastin accumulation. In addition to being associated with central fibrosis, the occurrence of pulmonary spiculation is also linked to intratumoral fibrosis, elastin fiber grading, collagen fiber grading, tumor differentiation, lymph node metastasis, pleural invasion, tumor necrosis, STAS, and other factors. These findings are consistent with previous studies [ 22 , 32 ], suggesting that pulmonary spiculation serves as a radiological surrogate for increased biological invasiveness and aggressive pathological features in patients. The independent prognostic significance of spiculation disappeared in both PSM analyses and multivariate Cox regression, indicating that spiculation merely reflects the underlying invasive pathological process rather than serving as an independent predictor of survival. This finding strongly supports our hypothesis that the clinical relevance of spiculation is not isolated but closely linked to the morphological characteristics of central fibrosis. When spiculation arises from elastin‐rich alveolar collapse, its associated risk appears attenuated; however, when underpinned by collagenous remodeling, it signals a shift toward aggressive disease and poorer outcomes. Consequently, risk stratification for invasive lung adenocarcinoma should evolve beyond the binary assessment of spiculation to incorporate internal fibrotic phenotyping. Through the implementation of PSM, our study sought to mitigate potential selection bias, thereby strengthening the internal validity of our radiologic–pathologic correlation analysis. Our findings regarding the biological substrates of pulmonary spiculation align with existing literature while offering substantive supplementary evidence. Specifically, although spiculation appears to correlate with elastin‐rich central fibrosis, its association with collagen fiber content was observed to be more pronounced. The analysis of the relationship between spiculation and central fibrosis may offer a more precise pathological interpretation for evaluating its prognostic value, potentially contributing to a more refined risk stratification for IAC. Although this study offers valuable clinical insights, several limitations must be acknowledged. First, the retrospective, single‐center design may introduce inherent selection bias. In addition, the lack of external validation across diverse ethnic populations means that the generalizability of our findings on the pathological basis of spiculation remains uncertain. Second, the relatively small sample size may reduce the statistical power needed to detect subtle relationships between spicule morphology and specific stromal remodeling patterns, such as central collagenous fibrosis. Equally important are concerns regarding subjective evaluation and potential spatial mismatches. Despite aligning pathological sampling with the CT scanning plane, differences in slice thickness between CT images and microscopic sections, along with interobserver variability, may still affect the accuracy of our radiopathologic correlations. Finally, the cross‐sectional nature of this study prevented us from observing the dynamic progression of spiculation over time. Moving forward, prospective multicenter studies that incorporate longitudinal imaging and more detailed morphological analysis are needed to establish a more robust theoretical foundation for AI‐assisted radiopathologic diagnostic models. 5. Conclusion In conclusion, this study provides substantive supplementary evidence for the biological basis of pulmonary spiculation in IAC. Our findings indicate that the formation of spiculation is closely associated with central fibrosis. Although spiculation correlates with increased biological invasiveness, its prognostic significance appears to be modulated by specific subtypes of central fibrosis. Notably, the “prognostic balance” created by extensive alveolar collapse may attenuate the malignancy risk typically associated with spiculated lesions. Consequently, we propose that clinical risk stratification for IAC should transition from a binary assessment of spiculation to a more detailed evaluation of internal fibrotic patterns. This proposed radiopathologic framework may offer helpful insights for personalized prognostic assessment and inform the future development of AI‐assisted diagnostic models. Author Contributions Yongkui Zhang: conceptualization, investigation, supervision. Liyong Qian: investigation. Yuanyuan Wen: methodology, investigation. Heping Wang: data curation, methodology, investigation. Hui He: writing – review and editing, investigation, supervision, funding acquisition. Hanbo Le: investigation, funding acquisition, writing – original draft. Funding This work was supported by grants from the Zhoushan Municipal Science and Technology Bureau, Zhejiang Province (2023C31001) and Medical and Health Science and Technology Project of Zhoushan (2024RC01). Ethics Statement This retrospective study, approved by the Clinical Research Ethics Committee of Zhoushan Hospital (No. 2022446) with a waiver for informed consent. Conflicts of Interest The authors declare no conflicts of interest. Supporting information Figure S1: Kaplan–Meier survival analysis according to alveolar collapse size in the 2.0–3.0 cm tumor subgroup. Overall survival was compared between patients with alveolar collapse ≤ 5 and > 5 mm among those with invasive adenocarcinoma measuring 2.0–3.0 cm. Patients with alveolar collapse > 5 mm showed significantly better overall survival than those with collapse ≤ 5 mm, as assessed by the log‐rank test ( p = 0.011). TCA-17-e70319-s002.jpg (344.7KB, jpg) Table S1: Full‐model logistic regression analysis of predictors for pulmonary spiculation using the enter method (including multicollinearity diagnostics). TCA-17-e70319-s001.docx (20.3KB, docx) Table S2: Multivariable Cox proportional hazards regression analysis of potential prognostic factors (enter method) in the original cohort. TCA-17-e70319-s003.docx (20.6KB, docx) Acknowledgments We would like to thank Yun Lei for her helpful technical assistance with the pathological section preparation. Data Availability Statement The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions. References 1. Siegel R. L., Kratzer T. B., Wagle N. S., Sung H., and Jemal A., “Cancer Statistics, 2026,” CA: A Cancer Journal for Clinicians 76, no. 1 (2026): e70043, 10.3322/caac.70043. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 2. Balagurunathan Y., Beers A., McNitt‐Gray M., et al., “Lung Nodule Malignancy Prediction in Sequential CT Scans: Summary of ISBI 2018 Challenge,” IEEE Transactions on Medical Imaging 40, no. 12 (2021): 3748–3761, 10.1109/tmi.2021.3097665. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 3. Zhong D., Sidorenkov G., Jacobs C., et al., “Lung Nodule Management in Low‐Dose CT Screening for Lung Cancer: Lessons From the NELSON Trial,” Radiology 313, no. 1 (2024): e240535, 10.1148/radiol.240535. [ DOI ] [ PubMed ] [ Google Scholar ] 4. de Koning H. J., van der Aalst C. M., de Jong P. A., et al., “Reduced Lung‐Cancer Mortality With Volume CT Screening in a Randomized Trial,” New England Journal of Medicine 382, no. 6 (2020): 503–513, 10.1056/NEJMoa1911793. [ DOI ] [ PubMed ] [ Google Scholar ] 5. Zwirewich C. V., Vedal S., Miller R. R., and Müller N. L., “Solitary Pulmonary Nodule: High‐Resolution CT and Radiologic‐Pathologic Correlation,” Radiology 179, no. 2 (1991): 469–476, 10.1148/radiology.179.2.2014294. [ DOI ] [ PubMed ] [ Google Scholar ] 6. Siegelman S. S., Khouri N. F., Leo F. P., Fishman E. K., Braverman R. M., and Zerhouni E. A., “Solitary Pulmonary Nodules: CT Assessment,” Radiology 160, no. 2 (1986): 307–312, 10.1148/radiology.160.2.3726105. [ DOI ] [ PubMed ] [ Google Scholar ] 7. Huawei Wu X. X., Liu S., and Li H., “Intra‐Tumor Basis and Influential Factors of Pleural Indentation in Peripheral Lung Cancer,” Chinese Journal of Radiology 35, no. 10 (2001): 731–735, 10.3760/j.issn:1005-1201.2001.10.002. [ DOI ] [ Google Scholar ] 8. Shimosato Y., Suzuki A., Hashimoto T., et al., “Prognostic Implications of Fibrotic Focus (Scar) in Small Peripheral Lung Cancers,” American Journal of Surgical Pathology 4, no. 4 (1980): 365–373, 10.1097/00000478-198008000-00005. [ DOI ] [ PubMed ] [ Google Scholar ] 9. Kung I. T., Lui I. O., Loke S. L., et al., “Pulmonary Scar Cancer. A Pathologic Reappraisal,” American Journal of Surgical Pathology 9, no. 6 (1985): 391–400, 10.1097/00000478-198506000-00001. [ DOI ] [ PubMed ] [ Google Scholar ] 10. Okudera K., Kamata Y., Takanashi S., et al., “Small Adenocarcinoma of the Lung: Prognostic Significance of Central Fibrosis Chiefly Because of Its Association With Angiogenesis and Lymphangiogenesis,” Pathology International 56, no. 9 (2006): 494–502, 10.1111/j.1440-1827.2006.01997.x. [ DOI ] [ PubMed ] [ Google Scholar ] 11. Suzuki K., Yokose T., Yoshida J., et al., “Prognostic Significance of the Size of Central Fibrosis in Peripheral Adenocarcinoma of the Lung,” Annals of Thoracic Surgery 69, no. 3 (2000): 893–897, 10.1016/s0003-4975(99)01331-4. [ DOI ] [ PubMed ] [ Google Scholar ] 12. Xu L., Lin S., and Zhang Y., “Differentiation of Adenocarcinoma In Situ With Alveolar Collapse From Minimally Invasive Adenocarcinoma or Invasive Adenocarcinoma Appearing as Part‐Solid Ground‐Glass Nodules (≤ 2 cm) Using Computed Tomography,” Japanese Journal of Radiology 40, no. 1 (2022): 29–37, 10.1007/s11604-021-01183-9. [ DOI ] [ PubMed ] [ Google Scholar ] 13. Shen Y., Li L., Yang Z., Wen Y., Qian L., and He H., “Prognostic Significance of Alveolar Collapse in Lung Invasive Adenocarcinoma and Its Relationship With Tumor Infiltrating Lymphocytes,” Pathology, Research and Practice 254 (2024): 155149, 10.1016/j.prp.2024.155149. [ DOI ] [ PubMed ] [ Google Scholar ] 14. Trejo Bittar H. E., Jerome J. A., Hartman D., Pantanowitz L., Mehrad M., and Dacic S., “Prognostic Significance of Microscopic Size in Peripherally Located Scar‐Associated Clinical Stage I Lung Carcinomas,” Lung Cancer (Amsterdam, Netherlands) 143 (2020): 12–18, 10.1016/j.lungcan.2020.03.004. [ DOI ] [ PubMed ] [ Google Scholar ] 15. Takise A., Kodama T., Shimosato Y., Watanabe S., and Suemasu K., “Histopathologic Prognostic Factors in Adenocarcinomas of the Peripheral Lung Less Than 2 cm in Diameter,” Cancer 61, no. 10 (1988): 2083–2088, 10.1002/1097-0142(19880515)61:10<2083::aid-cncr2820611025>3.0.co;2-u. [ DOI ] [ PubMed ] [ Google Scholar ] 16. Zhu J., Fan Y., Xiong Y., et al., “Delineating the Dynamic Evolution From Preneoplasia to Invasive Lung Adenocarcinoma by Integrating Single‐Cell RNA Sequencing and Spatial Transcriptomics,” Experimental & Molecular Medicine 54 (2022): 2060–2076, 10.1038/s12276-022-00896-9. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 17. Zhang X., Liang B., Huang Y., et al., “Behind the Indolent Facade: Uncovering the Molecular Features and Malignancy Potential in Lung Minimally Invasive Adenocarcinoma by Single‐Cell Transcriptomics,” Advanced Science 10, no. 36 (2023): e2303753, 10.1002/advs.202303753. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 18. Hanbo Cao M. W. and Wang H., “The Value of MSCT in the Diagnosis and Differential Diagnosis of Pleural Indentation of Solitary Pulmonary Nodules (≤2 cm),” Journal of Medical Imaging 27, no. 8 (2017): 1471–1474. [ Google Scholar ] 19. Hsu J. S., Han I. T., Tsai T. H., et al., “Pleural Tags on CT Scans to Predict Visceral Pleural Invasion of Non‐Small Cell Lung Cancer That Does Not About the Pleura,” Radiology 279, no. 2 (2016): 590–596, 10.1148/radiol.2015151120. [ DOI ] [ PubMed ] [ Google Scholar ] 20. Borczuk A. C., “Updates in Grading and Invasion Assessment in Lung Adenocarcinoma,” Modern Pathology 35, no. Suppl 1 (2022): 28–35, 10.1038/s41379-021-00934-3. [ DOI ] [ PubMed ] [ Google Scholar ] 21. Yamashita K., Matsunobe S., Takahashi R., et al., “Small Peripheral Lung Carcinoma Evaluated With Incremental Dynamic CT: Radiologic‐Pathologic Correlation,” Radiology 196, no. 2 (1995): 401–408, 10.1148/radiology.196.2.7617852. [ DOI ] [ PubMed ] [ Google Scholar ] 22. Wen Y., Wu W., Liufu Y., et al., “Differentiation of Granulomatous Nodules With Lobulation and Spiculation Signs From Solid Lung Adenocarcinomas Using a CT Deep Learning Model,” BMC Cancer 24, no. 1 (2024): 875, 10.1186/s12885-024-12611-0. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 23. Long Y., Li Y., Lin L., et al., “Predicting Visceral Pleural Invasion in Part‐Solid and Solid Nodules Using CT Features: A Systematic Review, Meta‐Analysis, and Independent Cohort Validation,” Journal of Computer Assisted Tomography 50 (2025): 426–433, 10.1097/rct.0000000000001831. [ DOI ] [ PubMed ] [ Google Scholar ] 24. Thunnissen E., Motoi N., Minami Y., et al., “Elastin in Pulmonary Pathology: Relevance in Tumours With a Lepidic or Papillary Appearance. A Comprehensive Understanding From a Morphological Viewpoint,” Histopathology 80, no. 3 (2022): 457–467, 10.1111/his.14537. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 25. Yamashiro K., Yasuda S., Nagase A., Hirata T., Nojima T., and Nagashima K., “Prognostic Significance of an Interface Pattern of Central Fibrosis and Tumor Cells in Peripheral Adenocarcinoma of the Lung,” Human Pathology 26, no. 1 (1995): 67–73, 10.1016/0046-8177(95)90116-7. [ DOI ] [ PubMed ] [ Google Scholar ] 26. Eto T., Suzuki H., Honda A., and Nagashima Y., “The Changes of the Stromal Elastotic Framework in the Growth of Peripheral Lung Adenocarcinomas,” Cancer 77, no. 4 (1996): 646–656. [ PubMed ] [ Google Scholar ] 27. Castro P. P., de Perrot M., Chua Y. C., et al., “Importance of Tumor Size in Resectable Stage III‐N2 Non‐Small Cell Lung Cancer,” Journal of Thoracic and Cardiovascular Surgery 164, no. 3 (2022): 629–636, 10.1016/j.jtcvs.2022.02.015. [ DOI ] [ PubMed ] [ Google Scholar ] 28. Li Z., Xu W., Gu T., Cao X., Wu W., and Chen L., “Tumor Size, but Not Consolidation‐To‐Tumor Ratio, Is an Independent Prognostic Factor for Part‐Solid Clinical T1 Non‐Small Cell Lung Cancer,” Thorac Cancer 14, no. 6 (2023): 602–611, 10.1111/1759-7714.14788. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 29. Barsky S. H., Huang S. J., and Bhuta S., “The Extracellular Matrix of Pulmonary Scar Carcinomas Is Suggestive of a Desmoplastic Origin,” American Journal of Pathology 124, no. 3 (1986): 412–419. [ PMC free article ] [ PubMed ] [ Google Scholar ] 30. Seemayer T. A., Schürch W., and Lagacé R., “Myofibroblasts in Human Pathology,” Human Pathology 12, no. 6 (1981): 491–492, 10.1016/s0046-8177(81)80061-5. [ DOI ] [ PubMed ] [ Google Scholar ] 31. Majno G., “The Story of the Myofibroblasts,” American Journal of Surgical Pathology 3, no. 6 (1979): 535–542, 10.1097/00000478-197912000-00006. [ DOI ] [ PubMed ] [ Google Scholar ] 32. Seemann M. D., Seemann O., Luboldt W., et al., “Differentiation of Malignant From Benign Solitary Pulmonary Lesions Using Chest Radiography, Spiral CT and HRCT,” Lung Cancer 29, no. 2 (2000): 105–124, 10.1016/s0169-5002(00)00104-5. [ DOI ] [ PubMed ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials Figure S1: Kaplan–Meier survival analysis according to alveolar collapse size in the 2.0–3.0 cm tumor subgroup. Overall survival was compared between patients with alveolar collapse ≤ 5 and > 5 mm among those with invasive adenocarcinoma measuring 2.0–3.0 cm. Patients with alveolar collapse > 5 mm showed significantly better overall survival than those with collapse ≤ 5 mm, as assessed by the log‐rank test ( p = 0.011). TCA-17-e70319-s002.jpg (344.7KB, jpg) Table S1: Full‐model logistic regression analysis of predictors for pulmonary spiculation using the enter method (including multicollinearity diagnostics). TCA-17-e70319-s001.docx (20.3KB, docx) Table S2: Multivariable Cox proportional hazards regression analysis of potential prognostic factors (enter method) in the original cohort. TCA-17-e70319-s003.docx (20.6KB, docx) Data Availability Statement The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions. Articles from Thoracic Cancer are provided here courtesy of Wiley ACTIONS View on publisher site PDF (2.2 MB) Cite Collections Permalink RESOURCES Similar articles Cited by other articles Links to NCBI Databases Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top

Related documents

• The Study Of Probe-Based Confocal Laser Endomicroscopy Of Distal Airways In Diagnostics Of Lung Diseases#1088881
• Confocal Laser Micro-endoscopy in Chronic Obstructive Pulmonary Disease (COPD) and Lung Transplant Recipients#863740
• Probe-based Confocal Laser Endomicroscopy in Critically Ill COVID-19 Patients#1008214
• Characterization of Apolipoprotein A-I Pathways in Idiopathic Pulmonary Fibrosis#832227
• Human Autologous Lung Stem Cell Transplant for Idiopathic Pulmonary Fibrosis#828404