Introduction

In the United States, kidney cancer ranks as the sixth most prevalent cancer in men, and the ninth most prevalent in women1. Global cancer statistics indicate that the incidence of kidney cancer is the fourteenth-most common cancer type, with its mortality rate positioned sixteenth2. The early identification and screening of kidney cancer have been recognized as key focus areas in research3. If kidney cancer that is confined to the organ is identified at an early stage, surgical removal could be a potential curative option. At present, a large percentage (50–61%) of renal tumors are detected incidentally, which is a significant rise from 13% that was noted in the 1970 s4,5.

Small renal masses (SRMs) refer to kidney lesions that show enhancement with contrast and have a maximum size of 4 cm or smaller4. Computed tomography (CT) play an essential role in diagnosing kidney cancer. Contrast-enhanced CT scan is a frequently used for early detection of SRMs. Magnetic resonance imaging (MRI) is performed to evaluate renal masses more precisely. CT and MRI can offer valuable insights regarding the histological subtypes of small renal masses6. Despite advancements in imaging technologies, it has been reported that between 10% and 30% of SRMs, which are typically presumed to be RCC based on preoperative imaging assessments, are ultimately found to be benign following surgical excision7.

Several studies have reported differences in the interpretation agreement rates between CT and MRI in the evaluation of renal mass characteristics8,9,10. When discrepancies occur in the interpretation of renal masses on CT or MRI, the most conclusive method for assessing characteristics is the histopathologic examination through renal biopsies. Nonetheless, it is important to note that not all renal masses are amenable to biopsy, and certain masses may be situated in locations that pose significant risks for such procedures. Owing to these associated risks, there are numerous instances in which tissue biopsy is not feasible11,12.

While numerous studies have assessed imaging accuracy in renal mass characterization, few have focused on identifying patient- or tumor-related factors that are predictive of diagnostic discrepancy. Moreover, there is a paucity of data examining the concordance between CT/MRI and final pathology in real-world surgical cohorts. Our study addresses this gap by analyzing factors associated with radiologic-pathologic discordance in a contemporary patient population.

Therefore, the aim of this study was to evaluate clinical and radiological factors that are associated with discrepancies between imaging findings and final pathology in patients undergoing surgery for renal masses. These factors, if recognized preoperatively, may support radiologists and urologists in minimizing diagnostic uncertainty.

Methods

This study was approved by the Institutional Review Board (IRB) of Chung-Ang University Gwangmyeoung Hospital (IRB No.2503-229-047), which waived the requirement for informed consent due to the retrospective nature of the study. All study procedures were conducted in accordance with the principles outlined in the Declaration of Helsinki.

We retrospectively reviewed the medical records of 132 patients who underwent partial or radical nephrectomy at Chung-Ang University Gwangmyeoung Hospital between April 2022 and October 2024. Inclusion criteria required that patients had preoperative imaging, either CT or MRI. All patients initially underwent a CT scan, with MRI performed at the clinician’s discretion when additional information was deemed necessary. The mean size of the renal masses was measured using CT images.

Contrast-enhanced CT or contrast-enhanced MRI scans were performed to evaluate each renal mass. All imaging studies were interpreted by a single board-certified uroradiologist with over 10 years of experience. Imaging reports were classified into two main categories: benign and malignant. The benign category included diagnoses such as angiomyolipoma (AML), oncocytoma, hemorrhagic cyst, nodules, cystic lesions (Bosniak I, II, and IIF), and other benign findings. The malignant category included renal cell carcinoma (RCC), atypical RCC, Bosniak III, Bosniak IV, and cystic RCC. If the imaging report suggested malignancy, it was classified as malignant, even if benign conditions were also noted within the report.

Discrepancy between the imaging reports and pathology results was evaluated. If malignancy was suggested in either the CT or MRI report, a discrepancy was defined as the pathological report indicating benignity. When both CT and MRI were available, MRI findings were prioritized as the final imaging impression due to its superior soft-tissue contrast and higher reported accuracy in cystic lesion classification and small renal mass characterization9,17.

Descriptive statistics were used to summarize the frequencies and proportions of categorical variables. Continuous variables are presented as mean ± standard deviation. Comparisons of continuous variables were performed using the independent t-test, and categorical variables were compared using Fisher’s exact test. Predictive factors for discrepancies between imaging and pathology reports were assessed using a multivariate logistic regression model, with odds ratios (OR) and 95% confidence intervals (CI) calculated. Only significant factors identified through univariate analysis were included in the multivariate model. All statistical analyses were conducted using IBM SPSS® (version 27.0; SPSS Inc., Chicago, IL, USA). A p-value of < 0.05 was considered statistically significant.

Results

Among the 132 patients included in the study, the mean age was 59 years, with 81 male patients (61.4%) and 51 female patients (38.6%). The mean tumor size was 4.6 cm. Most patients (84, 63.6%) underwent robot-assisted partial nephrectomy (Table 1).

A total of 31 patients underwent MRI, and of these, 24 were diagnosed with RCC, with 18 of them having the clear cell subtype, the most common histological type. In contrast, 101 patients did not undergo MRI, and among them, 78 were diagnosed with RCC, with 68 patients (87.2%) having the clear cell subtype (Fig. 1).

Fig. 1
figure 1

Pathological report according to CT/MRI.

Full size image
Table 1 Baseline characteristics.
Full size table

In the group of 101 patients who only underwent preoperative CT, 88 patients (87.1%) were reported as RCC on the CT imaging report. Of these, 74 (84.1%) were confirmed as RCC on the pathological report. In 13 patients (12.9%) where the CT report indicated a benign condition, RCC was confirmed in 4 patients (30.8%) based on pathological report. Among them, six patients had cystic lesions, for which the Bosniak classification and corresponding pathological results are summarized in Supplementary Table 1. Among the 31 patients who underwent MRI, 30 patients (96.8%) were diagnosed with RCC on the MRI report, and 24 (80%) were confirmed to have RCC on the pathological report. One patient had a benign condition on both the MRI and pathological reports (Table 2).

Table 2 Pathological report according to CT and MRI imaging report.
Full size table

In the subset of 31 patients who underwent both CT and MRI, 29 patients (80.6%) had RCC reported on both the CT and MRI imaging reports. The median tumor size in this group was 3 cm. Of these, 6 patients (20.7%) had benign tumors, with a median tumor size of 2 cm. In two patients where the CT report suggested benign findings, one was confirmed as RCC by both MRI and pathology, while the other was confirmed as benign by both MRI and pathology report (Table 3).

Table 3 CT findings followed by MRI findings and pathologic results.
Full size table

The discrepancy between the CT imaging report and the pathological report was 20.6% (27/131), while the discrepancy between the MRI imaging report and the pathological report was 19.4% (6/31). Among the 27 patients with discrepancies between the CT imaging report and pathological findings, 7 patients underwent MRI, and in 1 case, the MRI report was consistent with the pathological report.

The overall discrepancy between imaging reports and pathological findings was 19.8% (26/131). Among the 26 discrepant cases confirmed as benign by pathology, the histologic diagnoses included AML (n = 21), oncocytoma (n = 2), simple cysts (n = 1), papillary renal neoplasm with reverse polarity (n = 1) and multilocular cystic renal neoplasm of low malignant potential (n = 1). The representative CT and MRI images showed a discrepancy between the imaging reports and the pathological findings, as shown in Fig. 2. Multivariable regression analysis revealed that being female (OR = 5.219, 95% CI = 1.891–14.405, p = 0.001) and having a tumor size of less than 2 cm (OR = 4.826, 95% CI = 1.785–13.051, p = 0.002) were significant factors influencing the discrepancy between imaging and pathological diagnoses (Table 4). Further detailed results regarding tumor size are provided in Supplementary Table 2.

Fig. 2
figure 2

Pathological report according to CT/MRI. The representative CT and MRI images show a discrepancy between the imaging reports and the pathological findings. The CT (A, C) report suggested RCC, and the MRI (B, D) report indicated atypical RCC, but the pathological report confirmed AML.

Full size image
Table 4 Factors associated with discrepancy between preoperative imaging and final pathology.
Full size table

Discussion

Our study provides important insights into the discrepancies between imaging and pathological findings in the renal masses, specifically focusing on predictive factors. The results revealed that smaller tumor size (< 2 cm) and female sex were significantly associated with discrepancies between preoperative imaging and pathological diagnoses. These findings emphasize the challenges in accurately predicting malignancy based on imaging alone, particularly for small renal masses9. Unlike previous studies, which primarily emphasized imaging accuracy, our research highlights patient-specific factors such as tumor size and sex contributing to misclassification13. By identifying these factors, we offer a novel perspective on preoperative imaging limitations and the need for personalized diagnostic approaches.

CT and MRI are widely used for renal mass evaluation, each offering distinct advantages. Contrast-enhanced CT remains the primary imaging modality due to its accessibility and high sensitivity for detecting renal tumors. MRI, on the other hand, has demonstrated superiority in characterizing renal mass histology, cystic lesions, and tumor thrombus in vena cava14,15,16. Previous studies have shown that MRI outperforms CT in differentiating Bosniak IIF and III lesions and small RCC subtypes potentially reducing unnecessary surgeries, which could impact patient’s decline in renal function17,18. The discrepancy rate in our study was 19.8%, which aligns with previous literature reporting a 10–30% rate of benign diagnoses in renal masses initially suspected to be malignant based on imaging findings. Our findings highlight the limitations of current imaging modalities, including CT and MRI, in distinguishing benign such as lipid-poor AMLs from malignant lesions with absolute certainty. While MRI is generally considered superior to CT in differentiating renal mass characteristics and various additional analytical techniques have been introduced19,20,21, our study demonstrated that discrepancies persisted even when MRI was utilized, suggesting the need for improved diagnostic techniques22. Further, the development of reporting systems with reproducible and reliable algorithms, such as PI-RADS for prostate cancer and VI-RADS for bladder cancer, is also necessary23,24.

One of the key findings was the significant association between tumor size and diagnostic discrepancies. Small renal masses (< 2 cm) were 4.8 times more likely to be misclassified compared to larger tumors. This is likely due to the increased difficulty in accurately characterizing small lesions, as well as their higher probability of being benign entities such as oncocytomas or AML. These findings support the argument that active surveillance may be a viable option for select patients with small renal masses, potentially reducing unnecessary surgeries and associated morbidity. Additionally, female sex was found to be a strong predictor of imaging-pathology discrepancy, with women exhibiting a 5.2 times higher likelihood of misclassification. The reason for this gender disparity remains unclear but could be related to differences in tumor biology or referral patterns. Further research is needed to elucidate these potential contributing factors.

Despite the strengths of our study, including a well-defined patient cohort and the use of both CT and MRI for analysis, have several limitations. First, this was a retrospective study conducted at a single institution, which may limit the generalizability of our findings. Second, the relatively small sample size of patients who underwent MRI may have influenced statistical power, limiting our ability to draw definitive conclusions. Third, the selective use of MRI based on physician discretion may have introduced selection bias, as patients with more complex lesions or unclear CT findings were more likely to undergo MRI. Furthermore, our study did not incorporate advanced imaging techniques such as radiomics or AI-based analysis.

Future studies should explore the integration of radiomics and AI algorithms into renal mass imaging to improve diagnostic accuracy25. Several studies are reporting the usefulness of radiomics for the diagnosis of RCC26,27. Machine learning models trained on large datasets may provide enhanced risk stratification, minimizing discrepancies between imaging and pathology28,29. Furthermore, since imaging reproducibility is a critical prerequisite for the successful application of radiomics and AI, future studies should also consider data normalization across different imaging platforms, particularly in MRI. Variability in image acquisition protocols and scanner types may influence feature extraction and model generalizability, underscoring the need for standardized imaging pipelines. Prospective multicenter trials are also needed to validate our findings in broader populations. Additionally, investigations into sex-specific tumor characteristics may further elucidate the observed gender-related differences in imaging discrepancies. Lastly, assessing the impact of imaging-pathology discrepancies on clinical decision-making will be essential in refining management strategies for renal masses. Our study underscores the importance of considering tumor size and patient sex when interpreting preoperative imaging. Despite advances in imaging technology, a significant rate of discrepancy remains, necessitating caution in treatment decision-making. Further research incorporating advanced imaging modalities and AI-driven approaches will be crucial in optimizing diagnostic accuracy and patient management.

Conclusion

In this study, factors contributing to the discrepancy between the image reports and pathology reports for renal masses after undergoing CT and MRI were gender and the size of the renal mass. Specifically, the risk of discrepancy increased for renal masses less than 2 cm, suggesting that this factor as well as female sex should be considered when making clinical decisions.