Contrast-enhanced Imaging in Peripheral Pulmonary Lesions: The Role in US-guided Biopsies

Abstract

Purpose

To compare the tissue adequacy and diagnostic accuracy of US-guided biopsies of peripheral pulmonary lesions (PPLs) with and without contrast agents.

Materials and Methods

A retrospective study was conducted at four medical centers in patients with PPLs who underwent US-guided percutaneous transthoracic needle biopsy (PTNB) between January 2017 and October 2022. The patients were divided into contrast-enhanced US (CEUS) and US groups based on whether prebiopsy CEUS evaluation was performed. Tissue adequacy and the diagnostic accuracy of PTNB, stratified by lesion size, were analyzed and compared between groups. A propensity score matching (PSM) analysis was conducted using the nearest-neighbor matching method.

Results

A total of 1027 lesions were analyzed, with 634 patients (mean age, 59.4 years ± 13.0 [SD]; 413 male) in the US group and 393 patients (mean age, 61.2 years ± 12.5; 270 male) in the CEUS group. The CEUS group produced more acceptable samples than the US group (98.2% vs 95.7%; P = .03) and achieved higher diagnostic accuracy (96.9% vs 94.2%; P = .04), with no evidence of a difference in sensitivity (96.7% vs 94.0%; P = .06). PSM and stratified analyses (n = 358 per group) indicated higher tissue adequacy (99.0% vs 95.7%; P = .04) and diagnostic accuracy (98.5% vs 92.9%; P = .006) in the CEUS group compared with the US group for 2–7-cm PPLs but not for lesions larger than 7 cm.

Conclusion

PTNB with prebiopsy CEUS evaluation demonstrated significantly better tissue adequacy and diagnostic accuracy compared with US guidance alone for PPLs ranging from 2 to 7 cm, with similar biopsy performance achieved between groups for lesions larger than 7 cm.

Keywords: Contrast Material, Thoracic Diseases, Ultrasonography, Image-Guided Biopsy

© RSNA, 2024

Summary

Contrast-enhanced US in percutaneous needle biopsy for peripheral pulmonary lesions demonstrated better tissue adequacy and diagnostic accuracy compared with conventional US guidance for lesions between 2 and 7 cm but not for those larger than 7 cm.

Key Points

• ■ Percutaneous needle biopsy with contrast-enhanced US (CEUS) demonstrated significantly higher tissue adequacy (99.0% vs 95.7%) and diagnostic accuracy (98.5% vs 92.9%) than conventional US for peripheral pulmonary lesions (PPLs) measuring 2–7 cm; however, values were similar between CEUS and conventional US for PPLs larger than 7 cm.

• ■ These results suggest that CEUS can serve as a supplement in biopsy for 2–7-cm PPLs but is not compulsory for those larger than 7 cm.

• ■ The workflow of US-guided biopsy for PPLs can be simplified by stratified management.

Introduction

In the past decade, there have been substantial advancements in the field of radiology, leading to improved detection rates and diagnostic accuracy for focal lung lesions. However, with the advent of the precision medicine era, there is a growing need for more detailed information to guide personalized patient treatment and predict patient prognosis. Biopsy plays a crucial role by providing pathologic diagnosis with cellular- and molecular-level information, including immunohistochemistry and molecular phenotype.

The biopsy of lung lesions has played an increasingly important role in the diagnosis and treatment of malignant lung tumors. This is particularly true for some peripheral pulmonary lesions (PPLs) that are challenging to detect using bronchoscopy. Obtaining sufficient tissue for these lesions has posed a clinical challenge (1,2). Over the past few years, percutaneous transthoracic needle biopsy (PTNB) guided by CT has been widely recommended and used for image-guided biopsy procedures (3,4). However, owing to the better optimization of target lesion positioning relative to ribs and intercostal spaces, real-time US stands out for its lower complication rates, shorter operation times, and similar or slightly higher diagnostic accuracy in comparison with CT and other methods (4–7).

Contrast-enhanced US (CEUS) has gained widespread use and has been extensively studied in various organs over the past decade (8,9). CEUS offers the unique ability to display microvascular and tissue blood perfusion, thereby improving imaging quality and enabling the detection of lesions with high sensitivity and specificity. CEUS has also demonstrated important advantages in interventional procedures (3,10,11). The latest guidelines from the European Federation of Societies for Ultrasound in Medicine and Biology (9) recommend the use of CEUS during lung biopsy to visualize the necrotic part of the lesion and improve the success rate of puncture. However, additional data-driven evidence is required to support these guidelines and provide more comprehensive details for their implementation into clinical practice.

The primary objective of our study was to investigate the tissue adequacy and diagnostic accuracy of CEUS as a real-time guidance tool for PTNB in PPLs and compare it with conventional US. Our study aims to provide clinicians with valuable information to support the development of informed decisions regarding the optimal imaging guidance method for PTNB in various clinical scenarios.

Materials and Methods

Patients

This retrospective study was approved by the research and ethics committee of Sun Yat-sen University Cancer Center, and a waiver for written informed consent was granted due to the retrospective nature of the study. The study included 1108 patients from four medical centers who, between January 2017 and October 2022, underwent biopsy for indeterminate PPLs that were detected at CT or other examination and were visible at US. Twenty-eight patients were excluded from the study due to either missing images (n = 15) or a lack of final diagnosis (n = 13). Additionally, lesions smaller than 2 cm (the shortest length of the notch within the biopsy device [12]) were also excluded (n = 53). The remaining 1027 patients were then divided into two groups: the CEUS group and the US group. Further details can be found in the flowchart presented in Figure 1.

Figure 1:

Flowchart of patient selection. CEUS = contrast-enhanced US, PPL = peripheral pulmonary lesion, PSM = propensity score matching.

Prebiopsy Conventional US and CEUS Evaluation

Conventional US evaluation.—Conventional US, including B-mode US and color Doppler flow imaging, was used to scan the PPLs identified at CT or other radiologic assessment techniques. Several sonography systems were used for the conventional US examinations, including Hi Vision Avius (Hitachi Aloka), Resona 7 (Mindray), Aplio i800 (Canon), Logiq E9 (GE HealthCare), and those listed in the following section for the CEUS evaluation. The probe frequency ranged from 2.0 to 5.0 MHz. Details detected at the initial conventional US evaluation were recorded, including the lesion location, lesion size, typical US features, and presence of necrosis.

CEUS evaluation.—In this study, CEUS was performed using Acuson Sequoia (Siemens Healthineers), Logiq E9, and Aplio i800 systems. CEUS was conducted with a low mechanical index ranging from 0.05 to 0.12. The probe frequency used for imaging was set between 3.5 and 5.0 MHz. To prepare the contrast agent, SonoVue (Bracco), a homogeneous microbubble suspension was created by mixing 5 mL of normal saline with the lyophilized powder and shaking the mixture to achieve uniformity. It was administered via an antecubital vein at a reasonable dosage of 1.0–2.0 mL over approximately 3 seconds, followed by a 5-mL saline flush. During contrast agent injection, the perfusion and enhancement patterns of the target lesion were continuously monitored for a minimum 30 seconds, with intermittent observations for at least 3 minutes (9), which provides detailed insights into the vascularization of the target lesion. The CEUS evaluation was performed by senior radiologists on the same day as the PTNB procedure, ensuring timely and accurate assessment of the target lesion.

Biopsy protocol.—The biopsies in this study were guided using four sonography systems: Hi Vision Avius, Resona 7, Aplio i800, and Logiq E9. The probe frequency used during the procedure ranged from 1 to 6.6 MHz. Our center, serving as a national interventional training base, guarantees that our interventional radiologists, along with those from the other center involved in this study, receive comprehensive puncture training and acquire extensive experience in this field. All participating centers in the study strictly implemented a system of interventional US procedural grading. All operators involved in this study were senior physicians at the level of attending physicians or higher, possessing a minimum of 8 years of experience in general US, 2 years of experience CEUS, and 2 years of interventional US procedural experience. In this context, operators were categorized into two groups based on their varying levels of expertise (≤5 years vs >5 years) in US-guided interventional procedures.

Operators used “cognitive fusion” during real time US-guided biopsy, reviewing CEUS images acquired within 24 hours before the procedure. This technique involved simultaneously displaying CEUS and conventional US images in split-screen mode to confirm the area of viable lesion tissue on the conventional US images. The recorded position of sampling was categorized as either “central region” or “peripheral region,” with the division determined by the inner circle. The inner circle was defined as a concentric circle with a diameter half the size of the lesion, as shown in Figure 2.

Figure 2:

(A) Sankey diagram shows the relationship between lesion size and puncture position in the overall cohort. The lesion size was significantly correlated with the puncture location (Mantel-Haenszel χ² = 122.56 [P < .001]; R = 0.346 [P < .001]): The central region was mostly selected as the target for smaller lesions, and the peripheral region for larger ones. (B) Schema shows a cross-sectional view of an idealized spherical lesion with divided regions. The position was recorded as either the “central region” or the “peripheral region,” which were divided by the inner circle. The inner circle was defined as a concentric circle with half the diameter (D) of the lesion.

Following the application of 2% lidocaine local anesthesia, the biopsy needle was introduced and guided in real-time to a position immediately adjacent to the pleural lesion. For the biopsy procedure, a freehand approach was used, with a 16- or 18-gauge self-cutting needle (Bard Magnum disposable core biopsy needle). The number of punctures performed was determined based on desired sample quality and patient endurance, with a maximum of six. On-site lesion pathology assessment was not conducted. After the procedure, specimens were fixed in a 10% formalin solution and sent for pathologic examination. Additional immunohistochemical and molecular genetic studies were performed when necessary. Following the biopsy, all patients were observed for a minimum of 30 minutes.

Pathologic Evaluation and Final Diagnosis

Tissue adequacy.—The tissue adequacy in this study was determined based on whether the biopsy sample volume met the requirements of pathologic examination. If the sample volume was sufficient, allowing for a pathologic diagnosis to be established, the sampling was considered successful. Conversely, if the sample volume was inadequate and a pathologic diagnosis could not be made, the sampling was considered to have failed. All results obtained from the biopsies were interpreted under the supervision of two board-certified tumor pathologists with at least 5 years of experience interpreting thoracic pathology.

Diagnostic accuracy.—For cases where the pathology result indicated malignancy, the biopsy diagnosis was classified as a true positive. However, if the pathologic diagnosis indicated a negative result for malignancy, further laboratory tests, imaging studies, or a follow-up period of at least 12 months were required to verify the biopsy results. If the final diagnosis confirmed the lesion as benign, the biopsy results were considered to be a true negative. On the other hand, if the final diagnosis indicated malignancy despite a negative biopsy result, the biopsy was considered falsely negative (13). In such cases, additional biopsies were performed as needed. Overall, the accuracy of the biopsy results was determined based on their consistency with the pathology findings or the ultimate clinical diagnosis.

Conventional US and CEUS analysis.—Researchers retrieved, meticulously analyzed, and extracted imaging characteristics from images and videos of lesions that were captured and archived according to established protocols during the examinations. Additionally, two experienced radiologists (Z.F.Y. and K.H.G., with 10 and 15 years of US experience, respectively) conducted independent retrospective evaluations of the imaging characteristics of each lesion, without knowledge of clinical and pathologic outcomes.

Lesions with clear boundaries were defined as having clear margins. Internal echoes of the lesions are divided into homogeneous or heterogeneous echo based on whether the echo was uniform. For lesions evaluated at conventional US only, necrosis was identified during the conventional US examination as a decreased echogenicity region without blood flow signals at color Doppler flow imaging within the PPLs. Correspondingly, for the evaluation aided by CEUS, necrotic areas within the lesions were defined as regions that did not exhibit enhancement at any stage of the imaging process. It is important to note that areas perfused with contrast agents cannot be used directly as biopsy targets and should be differentiated from compressed lung tissue by observing differences in contrast agent arrival time (14).

Statistical Analysis

Data are summarized using standard descriptive statistics and frequency tabulation. The Kolmogorov-Smirnov test was used to assess the normal distribution of the data, and the Levene test was used to evaluate the homogeneity of variance. Group differences were analyzed by using the Mann-Whitney U test or χ² test, as appropriate. All statistical tests were two-sided, and P < .05 was considered to indicate a statistically significant difference between groups. To address potential latent biases, a one-to-one propensity score matching (PSM) analysis was performed using the nearest-neighbor matching method with a caliper width set at 0.05 (15–17). Key matching variables included “operator experience” and “number of puncture attempts.” Subsequently, a stratified analysis based on lesion size was conducted within the PSM cohort. The 7-cm cutoff point was used to stratify the analysis, as defined by the American Joint Committee on Cancer’s Cancer Staging Manual (eighth edition) (18); this cutoff distinguishes between T3 and T4 lesions, as it has shown to have a significant impact on survival outcomes. Furthermore, the associations of lesion size and puncture location with biopsy performance were also examined. SPSS 24.0 (IBM) was used for all data analyses.

Results

Patient Characteristics

A total of 1027 patients who underwent PTNB were included in this study, with 634 patients (mean age, 59.4 years ± 13.0 [SD]; 413 male, 221 female) in the US group and 393 patients (mean age, 61.2 years ± 12.5; 270 male, 123 female) in the CEUS group. Table 1 provides a summary of the baseline patient characteristics. There was no evidence of differences observed between the CEUS group and the US group in terms of sex, the proportion of malignant lesions, and the size of the puncture needle. However, it should be noted that the long-axis diameter of the lesion was found to be larger in the CEUS group compared with the US group (7.0 cm ± 2.9 vs 5.7 cm ± 2.6; P < .001). Additionally, a difference in age was detected between the two groups. Furthermore, certain features related to the puncture procedures, including the operator’s experience (P = .002) and the number of attempts to puncture (P < .001), differed between the groups.

Table 1:

Baseline Characteristics of the US and CEUS Groups Before and After Propensity Score Matching

Table 1 also shows a comparison of baseline characteristics following PSM between the propensity score–matched CEUS group (n = 358) and the US group (n = 358). The results demonstrated no evidence of differences in age, sex, the proportion of malignant lesions, number of puncture attempts, size of the puncture needle, and the level of operator experience between the two groups. This suggests that PSM successfully mitigated the differences in demographic characteristics and puncture operation between the two groups. Although disparities in lesion size still remained, they were subsequently thoroughly analyzed and presented using stratified analysis.

Imaging and Sampling Characteristics

In the overall cohort, a comparison between CEUS and conventional US revealed that CEUS was more effective in assisting the detection of necrosis within lesions (242 of 393 [61.6%] vs 75 of 634 [11.8%]; P < .001) and identifying lesions with clear boundaries (219 of 393 [55.7%] vs 291 of 634 [45.9%]; P = .002). Additionally, CEUS identified a greater number of lesions exhibiting heterogeneous echo patterns compared with conventional US (332 of 393 [84.5%] vs 480 of 634 [75.7%]; P = .001). These significant differences between the groups persisted following PSM, as shown in Table 2.

Table 2:

Imaging Characteristics of the US and CEUS Groups Before and After Propensity Score Matching

When performing PTNB, the operators demonstrated a preference for selecting the peripheral region of larger lesions as the target, while the central regions were favored for smaller lesions. This observation indicated a significant correlation between the diameter of the long axis of the lesion and the selected target puncture position, as shown in Figure 2 (Mantel-Haenszel χ² = 122.56; P < .001).

Biopsy Performance

The CEUS group showed significantly higher tissue adequacy compared with the US group (acceptable samples, 98.2% [386 of 393] vs 95.7% [607 of 634]; P = .03). Moreover, the diagnostic accuracy (96.9% [381 of 393] vs 94.2% [597 of 634]; P = .04) of PTNB in the CEUS group was significantly higher than in the US group, while the sensitivity was similar between the groups (96.7% [352 of 364] vs 94.0% [544 of 579]; P = .06). After PSM, there was no evidence of a difference in tissue adequacy (98.0% [351 of 358 patients] vs 96.1% [344 of 358]; P = .12), diagnostic accuracy (96.9% [347 of 358] vs 94.7% [339 of 358]; P = .14), or sensitivity (97.0% [324 of 334] vs 94.1% [318 of 338]; P = .07) of PTNB between the two groups. Table 3 shows a summary of the biopsy performance before and after PSM in both groups.

Table 3:

Sampling Success Rates and Diagnostic Accuracies of the US and CEUS Groups Before and After Propensity Score Matching

Figure 3 shows the results of the intergroup analyses stratified by lesion size. For lesions ranging from 2 to 7 cm, CEUS showed a significant improvement in tissue adequacy (before PSM: 99.1% [217 of 219 patients] vs 96.1% [447 of 465 patients] [P = .03]; after PSM: 99.0% [195 of 197 patients] vs 95.7% [244 of 255 patients] [P = .04]) and diagnostic accuracy (before PSM: 98.6% [216 of 219] vs 92.7% [431 of 465] [P = .001]; after PSM: 98.5% [194 of 197] vs 92.9% [237 of 255] [P = .006]) compared with US. However, for lesions larger than 7 cm, there was no evidence of differences between groups in tissue adequacy (before PSM: 97.1% [169 of 174 patients] vs 94.7% [160 of 169 patients] [P = .25]; after PSM: 96.9% [156 of 161 patients] vs 97.1% [100 of 103 patients] [P = .93]) or diagnostic accuracy (before PSM: 94.8% [165 of 174] vs 98.2% [166 of 169] [P = .09]; after PSM: 95.0% [153 of 161] vs 99.0% [102 of 103] [P = .08]).

Figure 3:

Bar graphs show sampling success rates and diagnostic accuracies stratified by lesion size between groups before and after propensity score matching (PSM). (A) Sampling success rates between groups before PSM. (B) Diagnostic accuracies between groups before PSM. (C) Sampling success rates between groups after PSM. (D) Diagnostic accuracies between groups after PSM. CEUS = contrast-enhanced US.

Figure 4 shows the performance of puncture location in terms of tissue adequacy and diagnostic accuracy, stratified by lesion size after PSM. The results indicated no evidence of a difference in the accuracy rate of PTNB, regardless of whether the puncture was performed in the central or peripheral region of the tumor. Representative cases showing US findings and biopsy procedures are presented in Figure 5.

Figure 4:

Bar graphs show the relationship between biopsy performance and puncture target regions stratified by lesion size after propensity score matching. (A) Sampling success rates of different puncture target regions. (B) Diagnostic accuracies of different puncture target regions.

Figure 5:

Prebiopsy evaluation and needle path. (A, B) Images in a 64-year-old male patient with 1-month cough history and a right lower lung lesion detected at CT. (A) Grayscale US image (right) shows an irregular hypoechoic mass, while contrast-enhanced US image (left) during the arterial phase (at 23 seconds after contrast material injection) indicates nonuniform enhancement within the lesion without persistent nonenhancement areas (*). (B) On a grayscale image, the US-guided biopsy needle path (arrowheads) avoids the area with persistent nonenhancement, confirming invasive adenocarcinoma at histopathologic diagnosis. (C, D) Images in a 61-year-old female patient after colon cancer surgery with a lung mass detected at PET/CT. (C) Grayscale US image (left) and color Doppler flow image (right) were used to assess mass characteristics, revealing an irregular hypoechoic lesion with peripheral blood flow signals in color Doppler flow imaging mode. (D) On a grayscale US image, the biopsy needle path (arrowheads), initially through the central region of the lesion, showed necrotic and fibrous tissue in the initial histologic examination, without cancer evidence. A subsequent change in guidance and path confirmed metastatic adenocarcinoma.

False-Negative Results

A total of 49 patients (4.8%) were found to have false-negative biopsy results. Among these, 35 patients (71%) had non–small cell lung cancer, five (10%) had small cell lung cancer, eight (16%) had other types of malignant lesions, and one had an inflammatory pseudotumor. The false-negative rates were 3.1% (12 of 393 biopsies) in the CEUS group and 5.8% (37 of 634 biopsies) in the US group. Of the 49 patients with false-negative results, 19 were confirmed at clinical follow-up, while 31 were confirmed at secondary biopsy. Among those who underwent a secondary biopsy, nine underwent CEUS assessment followed by biopsy, 10 underwent conventional US-guided PTNB, and 12 underwent bronchoscopic biopsy. Detailed information regarding the false-negative results and corresponding explanations are presented in Table 4.

Table 4:

Explanations and Features of False-Negative Results in the US and CEUS Groups

Discussion

In this study, we demonstrated that CEUS offers performance advantages over conventional US imaging in identifying lesion perfusion features. However, we found that this improvement observed in CEUS did not lead to a significant enhancement in the tissue adequacy or diagnostic accuracy after PSM. Nevertheless, our findings indicate that CEUS can enhance the tissue adequacy and diagnostic accuracy of PTNB for PPLs measuring between 2 and 7 cm. Additionally, experienced operators can achieve consistent diagnostic accuracy when using either US or CEUS as a guide for PTNB in PPLs larger than 7 cm.

Consistent with previous research, the diagnostic accuracy of CEUS-guided PTNB in this study was found to be 96.9%. Prior studies have reported improvements in diagnostic accuracy ranging from 10% to 15% when comparing CEUS-guided PTNB with US-guided PTNB (19). A recent study on CEUS for determining viable target areas in subpleural pulmonary lesions aligns with our research, as it highlights the potential of CEUS in improving diagnostic accuracy and biopsy strategy adjustment (20). However, in this study, the overall increase in diagnostic accuracy with CEUS-guided PTNB was only 2.7% (P < .05). After propensity modeling and one-to-one matching, the benefit of CEUS became insignificant, with only a 2.2% absolute difference (P > .05). This reduced advantage can be attributed to the improved diagnostic accuracy of the US group, which reached 94.2% in the overall cohort and 94.7% in the matched cohort. Previous studies have reported diagnostic accuracy ranging from 73.3% to 87.0% for US-guided PTNB (21).

When considering lesion size, differences in diagnostic accuracy between groups may become more apparent. Previous research has also highlighted the relationship between lesion size and biopsy accuracy (6,22). Our study indicates that CEUS is particularly advantageous in guiding PTNB for PPLs ranging from 2 to 7 cm but not for lesions larger than 7 cm. This finding diverges from previous studies that suggested a significant improvement in diagnostic accuracy with CEUS-guided PTNB regardless of lesion size (11,23). According to our findings, considering economic costs and the principle of “do no harm,” US remains the preferred method for guiding PTNB, especially in advanced medical centers with more experience in interventional procedures. These centers can achieve comparable success rates to those observed in our study. If initial PTNB attempts are unsuccessful, CEUS-guided PTNB may be considered for PPLs measuring between 2 and 7 cm. However, the utility of a repeat biopsy guided by CEUS is unlikely to be beneficial for larger lesions (>7 cm). Figure 6 illustrates the proposed workflow for US-guided biopsy of PPLs using stratified management. It is worth noting that lesions smaller than 2 cm were excluded from our study due to both necessity and safety concerns. Lesions measuring less than 2 cm and identifiable at conventional US are eligible for biopsy without necessitating further contrast-enhanced evaluation.

Figure 6:

Flowchart shows suggested proposal for the workflow of US-guided biopsy for peripheral pulmonary lesions (PPLs) by stratified management. CEUS = contrast-enhanced US, PTNB = percutaneous transthoracic needle biopsy.

To ensure the selection of an appropriate PTNB guidance modality, it is crucial to minimize the false-negative rate. False-negative biopsies can lead to delays in patient management and may necessitate repeat biopsies or invasive procedures such as secondary needle biopsy or excision biopsy, which can impose additional financial and psychologic burdens on patients. In the US group of our study, the overall false-negative rate was 5.8% (37 of 634 biopsies), while the use of CEUS reduced the false-negative rate to 3.3% (13 of 393 biopsies). Although CEUS was significantly better than US in the detection of necrosis in PPLs (61.6% vs 11.8%; P < .001), our study found that only 24% of the false-negative results (12 of 50) were due to extensive necrotic tissue. In the majority of cases (38 of 50), the false-negative results were caused by errors in selecting the target puncture area, which is defined as regions without malignant cells and necrotic components. This finding serves as a warning that enhancement areas observed at CEUS can mislead operators into targeting inflammatory or proliferative fibrous areas. We also found that in both the CEUS- and US-guided groups, there were five cases each where lung tissue in the puncture contributed to false-negative findings. The recommended criterion of enhancement time of less than 10 seconds, indicating a predominant supply from pulmonary arteries according to the European Federation of Societies for Ultrasound in Medicine and Biology guidelines (9), has been shown to be ineffective in distinguishing the blood supply of PPLs in a recent study (24). Therefore, currently, US remains a more viable option for guiding the initial PTNB procedure.

The finding that operators tend to select the central region as the target for small lesions and the peripheral region for larger ones aligns with common clinical practice. However, this study found no evidence of differences in tissue adequacy and diagnostic accuracy between the selected region, which is consistent with previous research (25). Binder et al (26) conducted a review of core needle biopsy for screening-detected lung cancer and concluded that intratumor heterogeneity does not hinder the accuracy of core needle biopsy for diagnosing malignancy and performing molecular testing.

We recognize certain limitations of our study. First, given the retrospective nature of the study, potential sources of bias remain, such as the selection for whether to perform CEUS assessment, which largely relies on the decisions of US specialists. We used PSM and a stratified analysis to minimize the potential differences (27). Second, although CEUS was used as a prebiopsy evaluation method, it was not directly used as a guidance tool. Operators performing CEUS participated in the biopsy procedure to provide guidance when necessary, aiming to improve the fusion and confirmation of images. Third, we refrained from conducting immediate on-site assessment to determine the adequacy of pathology specimens, which could potentially impact the diagnostic yield as well.

In conclusion, CEUS was effective in improving the accuracy of PTNB for PPLs ranging from 2 to 7 cm. However, skilled biopsy operators were able to achieve consistent diagnostic accuracy for PPLs larger than 7 cm, irrespective of their use of US or CEUS as the guidance tool. Our findings provide valuable insights for stratified management of PTNB. US guidance for PTNB is adequate, particularly in advanced medical centers. CEUS can be considered as a viable option for 2–7-cm PPLs in cases where initial attempts using other methods have failed. Future research could focus on optimizing the biopsy process for PPLs, using the guidance of CEUS or US to thoroughly explore the pathology outcomes.