Computed tomography (CT)-guided percutaneous transthoracic core needle biopsy (TCNB) has become an indispensable tool in the diagnostic evaluation of thoracic masses and pulmonary nodules. This minimally invasive technique allows for precise lesion targeting under real-time tomographic guidance, enabling high diagnostic accuracy while maintaining a favorable safety profile. Reported diagnostic yields for TCNB range from 85% to 95%, depending on lesion size, location, and operator experience [1, 2, 3]. Nevertheless, despite technical advancements, non-diagnostic results and sampling errors remain a concern, particularly in heterogeneous lesions that contain necrotic, hemorrhagic, or fibrotic components [4]. These challenges underscore the importance of identifying imaging markers that reflect the underlying tissue viability and vascularity at the biopsy site. Furthermore, procedural safety relies not only on careful imaging guidance but also on thorough pre-procedural patient assessment, including evaluation of coagulation parameters and platelet counts, as abnormalities in these laboratory indices have been associated with increased risk of bleeding or other complications [5]. Being aware of potential adverse events (such as pneumothorax, hemorrhage, or hemoptysis) and knowing their incidence are essential for optimizing patient selection, minimizing procedural risk, and interpreting post-biopsy imaging changes in clinical practice. Given these challenges, quantitative imaging metrics such as CT attenuation may provide additional insight into tissue characteristics, potentially guiding biopsy targeting and predicting malignant versus non-malignant pathology.

However, recent large-scale and multicenter studies have shown that the diagnostic performance of CT-guided transthoracic biopsy may vary substantially across different institutions and patient populations. Factors such as operator experience, lesion characteristics, and institutional practice patterns have been reported to influence diagnostic accuracy. In a recent multicenter study, Basiri et al. [6] reported a diagnostic accuracy of 82.9% for CT-guided percutaneous transthoracic needle biopsy, which was lower than the rates commonly reported in single-center series. This discrepancy highlights the limitations of conventional biopsy assessment and the need for adjunct quantitative imaging-based parameters to improve diagnostic confidence.

CT attenuation, expressed in Hounsfield units (HUs), provides a quantitative measure of tissue density based on the degree of X-ray beam attenuation. HU values are defined on a relative scale, with air assigned –1000 HU, distilled water 0 HU, and dense cortical bone approximately +1000 HU [7]. The attenuation coefficient primarily depends on the physical density and effective atomic number of the tissue. Moreover, iodinated contrast media, blood volume, and perfusion markedly influence measured attenuation, a relationship demonstrated in perfusion/dynamic contrast-enhanced CT (DCE-CT) and dual‑energy CT studies [8, 9, 10]. Viable, cellular, or vascularized tissues generally exhibit higher attenuation, whereas necrotic, cystic, or poorly perfused regions demonstrate lower HU values [11].

During CT-guided biopsy, both the pre-procedural scan used for needle planning and the post-biopsy scan for immediate complications provide opportunities to assess local attenuation characteristics. Mechanical disruption by the needle, micro-hemorrhages, and transient changes in perfusion can cause measurable alterations in HU at the biopsy site. However, the direction and magnitude of these attenuation changes may vary depending on lesion vascularity, internal composition, and the timing of post-procedure imaging. These variations may indirectly reflect the vascular and cellular composition of the sampled region. For example, tissues rich in viable cells and capillary networks may demonstrate a greater post-biopsy increase in attenuation due to localized hemorrhage and contrast redistribution, whereas necrotic or avascular regions may show minimal change or even decreased attenuation [12].

Despite the extensive use of CT for procedural guidance, the potential diagnostic information embedded in these subtle HU variations remains underexplored. A systematic evaluation of pre- and post-biopsy HU changes in relation to histopathological findings may provide valuable insight into tissue viability, sampling adequacy, and diagnostic yield. Understanding this relationship could improve target selection during biopsy, reduce non-diagnostic sampling, and enhance the role of imaging biomarkers in interventional radiology.

Previous studies have focused on static CT attenuation values, lesion size, or procedural factors affecting diagnostic yield and the complication rates of CT-guided transthoracic biopsy. However, dynamic changes in tissue attenuation during the biopsy itself have not been systematically evaluated. To the best of our knowledge, this is the first retrospective study to investigate pre- to post-biopsy HU changes along the needle tract as a quantitative, contrast-free, intra-procedural biomarker for predicting malignancy in thoracic lesions. Importantly, this approach was designed to be applicable in routine clinical settings where rapid on-site cytopathology evaluation is not available, aiming to support biopsy targeting and specimen adequacy assessment using readily obtainable imaging data. By analyzing attenuation change ($\mathrm{\Delta }$HU) rather than single time-point measurements, the present study aims to provide real-time insight into tissue viability and vascularity at the biopsy target, potentially improving sampling adequacy and diagnostic confidence without altering standard biopsy protocols.

The median age of all patients included in the study was 65 years (range: 19–85). When grouped according to TCNB findings, the median ages were 67 (32–85) years in the malignancy group, 66 (21–78) years in the necrosis group, and 58 (43–75) years in the inflammation group, showing a statistically significant difference between the groups (p = 0.027). In terms of gender distribution, 78.3% of the patients in the malignancy group, 72% in the necrosis group, and 90.5% in the inflammation group were male, with no statistically significant difference between the groups (p = 0.297). Detailed demographic, laboratory, and radiological characteristics are summarized in Table 1.

Procedure-related complications occurred in 50 patients in the malignancy group, 6 in the necrosis group, and 7 in the inflammation group, with no statistically significant difference among groups (p = 0.228) (Table 1). Pneumothorax was the most frequent complication, observed with 23.3% of malignant, 8% of necrotic, and 23.8% of inflammatory lesions. Hemorrhage occurred in 13.3% of malignant and 12% of necrotic lesions, while no hemorrhage was observed in the inflammatory group. Hemoptysis was noted in 3.3%, 4%, and 4.8% of cases, respectively. The coexistence of hemorrhage and pneumothorax was rare, observed only in 1.7% of malignant and 4.8% of inflammatory lesions.

There were no statistically significant differences between the groups in laboratory parameters, including hemoglobin, platelet count, and coagulation indices (international normalized ratio [INR], prothrombin time [PT], activated partial thromboplastin time [aPTT]), with p = 0.06, p = 0.19, p = 0.57, p = 0.66, and p = 0.78, respectively (Table 1).

Radiological attenuation (HU) values measured before and after TCNB differed significantly among the diagnostic groups, with malignant lesions showing higher pre- and post-biopsy HU and positive $\mathrm{\Delta }$HU compared with necrotic and inflammatory lesions (p 0.001) (Table 1).

The demographic, laboratory, and radiological findings of patients diagnosed with malignancy based on TCNB results are presented in Table 2. Age did not differ significantly among histopathological subgroups (p = 0.11). Male predominance was observed across most histopathological subgroups, and gender distribution differed significantly among tumor types (p = 0.015). Procedure-related complications were seen in 7 patients with small cell carcinoma, 19 with adenocarcinoma, 13 with squamous cell carcinoma, 4 with distant metastases, 4 with neuroendocrine tumors, and 3 with lymphoma (p = 0.641).

No significant intergroup differences were observed among the malignant subgroups in terms of laboratory parameters, including hemoglobin, platelet count, INR, PT, and aPTT (p = 0.09, p = 0.17, p = 0.38, p = 0.54, and p = 0.67, respectively) (Table 2).

Pre- and post-biopsy HU values varied across subgroups, with the highest attenuation observed in small cell carcinoma, and the lowest in squamous cell carcinoma. $\mathrm{\Delta }$HU tended to be higher in small cell carcinoma and adenocarcinoma, although these differences were not statistically significant (p = 0.47).

The diagnostic performance of pre-biopsy HU, post-biopsy HU, and $\mathrm{\Delta }$HU values for differentiating malignant from non-malignant lesions was evaluated by ROC curve analysis (Fig. 3). Pre-biopsy HU showed fair discriminative ability, with an AUC of 0.637 (95% CI: 0.546–0.727, p = 0.006). Post-biopsy HU demonstrated higher diagnostic accuracy, with an AUC of 0.886 (95% CI: 0.834–0.937, p 0.001). $\mathrm{\Delta }$HU exhibited the strongest diagnostic power, with an AUC of 0.956 (95% CI: 0.927–0.986, p 0.001). The optimal $\mathrm{\Delta }$HU cut-off value was determined using the Youden index derived from ROC curve analysis. Using a $\mathrm{\Delta }$HU cut-off value of 1.5 HU, the model predicted malignancy with 94% sensitivity and 78% specificity.

Fig. 3.

ROC curve analysis of HU before and after TCNB and HU change in patients with and without malignancy. Using a $\mathrm{\Delta }$HU cut-off value of 1.5 HU, sensitivity and specificity for predicting malignancy were 94% and 78%, respectively. ROC, receiver operating characteristic.

In this retrospective study, we investigated changes in attenuation ($\mathrm{\Delta }$HU) after CT-guided TCNB of thoracic lesions and analyzed their correlation with histopathological outcomes. Malignant lesions consistently exhibited a positive $\mathrm{\Delta }$HU, indicating an increase in attenuation post-biopsy, whereas necrotic lesions demonstrated a negative $\mathrm{\Delta }$HU and inflammatory lesions showed minimal change. These results suggest that $\mathrm{\Delta }$HU measurement could be used intra-procedurally to predict sample adequacy before histological confirmation, potentially reducing the need for repeat biopsies in thoracic lesions. This quantitative approach may potentially complement artificial intelligence (AI)-based imaging tools or radiomics, providing additional information in estimating malignancy probability.

Our findings align with prior reports demonstrating high diagnostic yields for CT-guided TCNB [1, 15]. Whereas earlier studies emphasized lesion size, location, and operator experience as predictors of diagnostic success, our results highlight the additional value of quantitative HU measurements. By integrating $\mathrm{\Delta }$HU analysis, interventional radiologists may obtain an objective assessment of tissue characteristics that extends beyond conventional planning.

To our knowledge, no previous study has systematically quantified HU variation before and after CT-guided transthoracic biopsy in relation to histopathological outcomes. Existing reports have mainly focused on attenuation as a marker of procedural complications rather than diagnostic yield. Zhou et al. [16] reported a nonlinear association between lung needle path CT attenuation and immediate pneumothorax risk, with lower HU values associated with higher complication rates. Although these studies focused primarily on complications rather than histopathology, they underscore the relevance of tissue attenuation as a marker of local tissue properties. Despite this potential, no previous studies have specifically examined changes in needle tract attenuation as a predictor of biopsy success in pulmonary nodules.

The strong discriminative performance of $\mathrm{\Delta }$HU suggests that these attenuation changes reflect intrinsic tissue characteristics, such as viability and vascularity, rather than being mere procedural observations. While these biological interpretations remain speculative and require confirmation through direct pathological or perfusion imaging correlation, the observed diagnostic accuracy is promising. As our study included a single-center cohort, these findings have not yet been externally validated. However, we utilized predefined imaging variables and standardized measurement and analysis techniques to reduce potential overfitting. Future multicenter studies are warranted to confirm the generalizability of $\mathrm{\Delta }$HU as a diagnostic biomarker.

The integration of pre-procedural PET-CT or contrast-enhanced MRI further strengthens the clinical applicability of our findings. Prior work has shown that PET/CT fusion imaging combined with intra-procedural CT can significantly improve diagnostic yield, particularly in heterogeneous or metabolically active lesions [15, 17]. In our cohort, PET-CT or MRI-guided targeting likely enhanced the detection of viable, vascularized regions, which corresponded with higher $\mathrm{\Delta }$HU values observed in malignant lesions. Notably, small cell carcinoma and adenocarcinoma subtypes exhibited the highest $\mathrm{\Delta }$HU, suggesting that increased cellularity or perfusion may contribute to post-biopsy attenuation changes, although inter-subtype differences were not statistically significant.

Age distribution differed significantly among the diagnostic groups, with patients in the malignancy group being older on average than those in the inflammatory group. This finding may reflect the epidemiological tendency for malignancy to occur at older ages, whereas inflammatory or necrotic lesions can present across a wider age range. While gender distribution did not differ significantly among groups, awareness of these demographic patterns may aid clinicians in interpreting biopsy results in the context of patient characteristics.

Laboratory parameters, including hemoglobin, platelet count, INR, PT, and aPTT, were largely within normal limits across all histopathological groups, reflecting careful pre-procedural patient assessment. The absence of significant differences among groups suggests that $\mathrm{\Delta }$HU variations are primarily attributable to intrinsic tissue characteristics rather than coagulation abnormalities. Complication rates were consistent with the literature, with pneumothorax and hemorrhage being the most common adverse events [13]. Importantly, complication frequencies did not differ significantly among histopathological categories or malignancy subtypes, supporting the overall safety of CT-guided TCNB when standard precautions are observed.

Although the concept of $\mathrm{\Delta }$HU as a marker of tissue viability and vascularity has not been directly validated previously, our study demonstrates its potential incremental value by quantitatively distinguishing malignant from necrotic or inflammatory lesions. These findings suggest that $\mathrm{\Delta }$HU may provide complementary information to conventional imaging and help guide more accurate sampling during CT-guided biopsy. From a physiological standpoint, the observed $\mathrm{\Delta }$HU differences may be related to tissue perfusion or micro-hemorrhage. While direct perfusion measurements were not performed, these hypothetical interpretations are supported by studies demonstrating correlations between CT perfusion/DCE parameters and microvessel density, attenuation thresholds predictive of necrosis, and characteristic HU evolution in hemorrhage [9, 10, 11]. Literature evidence indicating a correlation between perfusion parameters and HU dynamics suggests that $\mathrm{\Delta }$HU can serve as an indirect marker of tissue vascularity and viability. Future studies incorporating quantitative perfusion imaging could further validate this mechanistic link.

Clinically, the combination of $\mathrm{\Delta }$HU measurement and careful pre-procedural planning offers several advantages. Quantitative HU assessment may help refine target selection, provide immediate feedback on sampling adequacy, and reduce non-diagnostic biopsy rates in heterogeneous lesions containing necrosis or fibrosis. Wider adoption of $\mathrm{\Delta }$HU measurement could streamline workflow, minimize repeat procedures, and support decision-making in centers without on-site cytopathology. While PET-CT or MRI guidance may improve lesion targeting and $\mathrm{\Delta }$HU assessment, the added cost and workflow implications should be considered. In routine practice, $\mathrm{\Delta }$HU measurement using standard CT alone could provide a cost-effective adjunct in centers without advanced imaging resources. When integrated with standard laboratory assessment and procedural precautions, $\mathrm{\Delta }$HU may serve as an adjunctive biomarker with the potential to optimize diagnostic yield and procedural safety.

Limitations of this study include the single-center design and relatively small sample sizes in certain histological subgroups, which may limit generalizability and reduce statistical power, particularly for analyses of less common tumor types or histopathological subgroups. Manual ROI placement for HU measurement, though performed independently by two experienced radiologists, introduces potential observer variability. The potential role of automated or semi‑automated HU measurement tools could not only reduce interobserver variability but also streamline workflow and enhance real-time biopsy guidance. Future validation studies are warranted to assess the accuracy and clinical utility of such tools. Additionally, subtle micro-air or hemorrhagic changes may still influence HU measurements despite the exclusion of overt artifacts. Finally, variability in the timing of post-biopsy CT acquisition could affect $\mathrm{\Delta }$HU, suggesting that standardized post-procedure imaging protocols are warranted.

Our study demonstrates that $\mathrm{\Delta }$HU measured during CT-guided TCNB is a reliable imaging biomarker of tissue viability and histopathological outcome in thoracic lesions. When combined with careful laboratory evaluation and advanced imaging guidance, $\mathrm{\Delta }$HU analysis has the potential to enhance diagnostic accuracy, optimize biopsy targeting and improve patient management in interventional radiology practice.

• $\mathrm{\Delta }$HU (attenuation change from pre- to post-biopsy) is a novel quantitative biomarker obtained during CT-guided transthoracic core needle biopsy and has not been previously evaluated in thoracic lesions.

• Malignant lesions show a characteristic positive $\mathrm{\Delta }$HU, reflecting higher tissue viability/vascularity, whereas inflammatory and necrotic lesions demonstrate minimal or negative $\mathrm{\Delta }$HU, respectively, indicating poor perfusion or non-viable tissue.

• $\mathrm{\Delta }$HU demonstrated excellent diagnostic accuracy for malignancy (AUC = 0.956), with 94% sensitivity and 78% specificity, outperforming static HU values alone.

• HU measurement during routine pre- and post-biopsy scans is highly reproducible (interobserver $\alpha$ = 0.849), requiring no protocol changes or contrast administration.

• Dynamic HU analysis may improve biopsy targeting, reduce sampling errors in heterogeneous lesions, and increase diagnostic confidence, especially when necrosis or non-diagnostic tissue is a concern.

• Integration of $\mathrm{\Delta }$HU into the standard TCNB workflow is feasible, low-cost, and provides immediate real-time quantitative feedback.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

MY, FAY and ÖK designed the research study and wrote the first draft. MY, BK and FAL performed the research. MY, AA and AK analysed the data. All authors contributed to the important editorial changes in the manuscript. All authors read and approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

The study was approved by the local ethics committee of Atatürk University (approval no: 26, meeting no: 10, date: 28.11.2025) and conducted in accordance with the Declaration of Helsinki. Ethics approval was obtained prior to data analysis and manuscript preparation. Written informed consent for the biopsy procedure was obtained from all patients as part of routine clinical care. Due to the retrospective nature of the study and the use of anonymized data, the requirement for additional research-specific informed consent was waived by the ethics committee.

Not applicable.

This research received no external funding.

The authors declare no conflict of interest.