Diagnostic Value of Rapid On‐Site Evaluation Combined With Prostate Biopsy in Prostate Cancer

ABSTRACT

Objective

To investigate the application value of Rapid On‐Site Evaluation (ROSE) in prostate biopsy.

Methods

All consecutive subjects who attended our clinic to underwent magnetic resonance imaging (MRI)‐ultrasound fusion biopsy due to highly suspicious findings on MRI for prostate cancer (PCa) and met the inclusion criteria were enrolled into our prospective study between October 2020 and January 2025. ROSE was performed concurrently in the same operating room with MRI‐ultrasound fusion biopsy. For each lesion with Prostate Imaging‐Reporting and Data System (PI‐RADS) 4‐5, one to two additional needle passes were taken for ROSE, in addition to the standard biopsy. The results of ROSE during the biopsy were recorded. The sensitivity, specificity, positive predictive value, and negative predictive value of ROSE were assessed using paraffin‐embedded histopathology of the biopsy specimens as the gold standard.

Results

A total of 313 lesions with PI‐RADS 4‐5 from 147 patients were ultimately included in this study. All biopsies were performed smoothly, with no severe complications occurring postoperatively. 192 lesions were pathologically diagnosed with PCa, yielding a positive detection rate of 61.3% (192/313). With paraffin‐embedded histopathology of the biopsy specimens serving as the gold standard, the sensitivity of ROSE for detecting PCa was 71.9% (138/192), specificity was 100% (121/121), accuracy was 82.7% (259/313), positive predictive value was 100% (138/138), and negative predictive value was 69.1% (121/175).

Conclusions

The application of ROSE technology in the diagnosis during prostate biopsy is accurate and reliable, with specificity and positive predictive value both reaching 100%.

1. Introduction

Prostate cancer (PCa) is one of the most prevalent malignancies in elderly males and a leading cause of cancer‐related death in males [1]. Prostate biopsy is crucial for the diagnosis, risk stratification, and treatment planning of PCa. Given the heterogeneity of PCa, discordance between biopsy histopathology and final post‐prostatectomy pathology is common [2, 3]. With the development of new imaging technologies, the accuracy of biopsy has improved significantly [4].

However, conventional magnetic resonance imaging (MRI) based on the Prostate Imaging Reporting and Data System (PI‐RADS) has been shown to result in the underdiagnosis of some patients with clinically significant PCa [4]. Although conventional molecular diagnostic methods such as Prostate‐Specific Antigen (PSA), are extensively used, they suffer from insufficient sensitivity and specificity, contributing to overdiagnosis. Meanwhile, various novel biomarkers lack sufficient validation [5, 6, 7]. Even the most advanced techniques such as Prostate Specific Membrane Antigen (PSMA) Positron Emission Tomography (PET)/Computed Tomography (CT) are not infallible, and false‐positives remaining a concern [8, 9]. The limitations of imaging assessment increase the number of biopsy cores, thereby augmenting the risk of complications such as hematuria, urinary retention, and infection [10]. Additionally, the need for secondary surgical intervention shortly after biopsy poses risks, as it is necessary to wait for the resolution of inflammatory adhesions or hematomas in the biopsied area before proceeding with surgery [11]. Improving the accuracy of prostate biopsy while minimizing unnecessary damage is a current clinical challenge. Although some studies have proposed radical prostatectomy (RP) without prior biopsy, it still needs cautious consideration [12].

Rapid On‐Site Evaluation (ROSE) is a diagnostic technique characterized by its rapid, real‐time, and on‐site nature. It involves immediate processing, staining, and assessment of specimens obtained during interventional biopsies. ROSE has been widely applied in biopsies of organs like the lung, thyroid, and pancreas, where it has been proven to reduce the number of biopsy passes, provide instant pathological feedback, and enhance diagnostic accuracy [13, 14]. Given these advantages, ROSE holds great potential in prostate biopsy, particularly for immediate pathological results and improved PCa detection. As no studies on ROSE in prostate biopsy have been reported yet, this study aims to evaluate its accuracy in prostate biopsy, with the expectation of facilitating its future application in this field.

2. Materials and Methods

2.1. Study Population and Design

All consecutive subjects who attended Department of Urology at Beijing Hospital from October 2020 to January 2025 for the evaluation of PCa using MRI‐ultrasound fusion biopsy and met the inclusion criteria were enrolled into our prospective study. The inclusion criteria were as follows: (1) age ≥ 18; (2) high clinical suspicion of PCa based on clinical features and MRI; (3) PI‐RADS 4–5; (4) absence of significant contraindications to biopsy. This study was approved by the Beijing Hospital Ethics Committee (Approval Letter No: 2023BJYYEC‐431‐02). Written informed consent was obtained from the 147 patients. Figure 1 presents a flowchart of the study design. The diagnostic efficiency of ROSE was analyzed using paraffin‐embedded histopathology results as the gold standard.

Figure 1.

Flowchart showing research design. PSA, prostate specific antigen. MRI, magnetic resonance imaging. PI‐RADS, Prostate Imaging‐Reporting and Data System. ROSE, Rapid On‐Site Evaluation.

2.2. Prostate Biopsy

The day before the procedure, patients were administered an oral polyethylene glycol electrolyte powder to cleanse the bowel, followed by a cleansing enema on the day of the procedure. The biopsy was performed in the operating room by physicians with a professional title of attending physician or above and extensive experience in prostate biopsies. Under general anesthesia and in the lithotomy position, the surgical area was routinely disinfected and draped. Antibiotics were administered prophylactically before the procedure. A urinary catheter was inserted, and the transrectal ultrasound probe was fixed to the biopsy frame. After the probe was inserted into the rectum, the biopsy template was secured to the frame and positioned tightly against the perineal skin. The internal urethral orifice was located using the water balloon of the urinary catheter. Based on the vertical distance between the center of the suspicious lesion and the internal urethral orifice on the coronal view of the MRI, the probe was retracted from the transverse section of the internal urethral orifice in 0.5 cm increments using a step‐by‐step device until it reached the same distance from the bladder neck to the transverse section where the lesion was located. At this level, the lesion was localized using ultrasound guidance, and a biopsy needle was inserted through the corresponding hole in the template to perform targeted biopsy of the lesion. The standard biopsy technique involved 1 to 2 needle passes per lesion, followed by systematic biopsy. Figure 2 illustrates some of the steps in the MRI‐ultrasound fusion biopsy procedure.

Figure 2.

MRI‐ultrasound fusion biopsy. [Color figure can be viewed at wileyonlinelibrary.com]

In addition to the standard biopsy, 1 to 2 extra needle passes were taken at each lesion for ROSE evaluation, with the same marking as that used for pathological specimen submission.

2.3. MRI

Prior to the biopsy, within 1 month, a 3.0 T MRI scanner equipped with a 32‐channel phased‐array surface coil was utilized. The scanning protocol included axial, coronal, and sagittal T2‐weighted imaging and T1‐weighted imaging, supplemented by diffusion‐weighted imaging and/or contrast‐enhanced scanning. All MRI images were evaluated by senior radiologists using the PI‐RADS (version 2.1, 2019). The location of suspicious PCa lesions was determined based on the lesion's position, diameter, and the distance from the lesion center to the bladder neck and apex.

2.4. ROSE

For each targeted biopsy specimen, one tissue squash slide was prepared. The Diff‐Quik staining was performed using the immersion method as follows: (1) The slide was immersed in Diff A solution for 5 to 10 s; (2) the slide was then washed in phosphate buffer saline for 5 s; (3) after gently shaking off the excess liquid, the slide was immersed in Diff B solution for 5 to 10 s; (4) the slide was rinsed in distilled water and then air‐dried to complete the staining process.

All stained slides were examined under an optical microscope by trained urologists, who rapidly assessed the cellular morphology. The presence of cancer cells was deemed a positive biopsy result, whereas the absence of cancer cells was considered a negative result. Figure 3 illustrates some of the cells under the optical microscope in the stained slides of ROSE.

Figure 3.

Cells in the stained slides of Rapid On‐Site Evaluation under the optical microscope. 1–2. The mitotic phase of cancer cells. 3–8. Enlarged nuclei of cancer cells, increased nuclear‐cytoplasmic ratio, prominent nucleoli, and monomorphic nuclei. 9. Inflammatory cells with nuclei not deeply stained. [Color figure can be viewed at wileyonlinelibrary.com]

2.5. Histopathology

All biopsy specimens obtained were fixed in 10% formalin, followed by dehydration and embedding in paraffin. The paraffin blocks were sectioned in three planes, and the sections were stained with hematoxylin and eosin. Unstained sections were reserved for immunohistochemical staining. The presence of PCa and its grading were assessed according to the criteria established by the International Society of Urological Pathology.

The results of ROSE were compared with the histopathological findings from the paraffin‐embedded sections to evaluate the accuracy of ROSE detection.

2.6. Statistical Analysis

Data analysis was performed using SPSS software version 26.0. Continuous variables are presented as medians with interquartile ranges, while categorical variables are expressed as frequencies and percentages. Pearson's chi‐square test and Fisher's exact test were employed to analyze binary variables. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value of ROSE for PCa diagnosis were calculated using a 2×2 contingency table. The level of agreement between ROSE and pathological diagnosis was assessed using the Kappa statistic. A p‐value < 0.05 was considered to indicate a statistically significant difference.

3. Results

3.1. Baseline Characteristics of the Patients

Table 1 summarizes the clinical characteristics of the patients included in the study. A total of 147 patients were enrolled, with a median age of 70.5 years (66, 76) and a median PSA level of 9.44 ng/ml (5.69, 24.13). The final clinical pathological diagnoses were made based on histopathological findings from paraffin‐embedded sections and further surgical findings. The remaining 42 cases were diagnosed as benign lesions.

Table 1.

Baseline characteristics of the patients.

Variables		Median (IQR) or n%
Age (yr)		70.50 (66.00, 76.00)
PSA (ng/ml)		9.44 (5.69, 24.14)
fPSA		1.49 (0.85, 2.73)
PSAr		0.169 (0.12, 0.22)
With PSMA PET/CT or not	Yes	31 (21.1)
	No	116 (78.9)
Histopathology	Postive	105 (71.4)
	Negative	42 (28.6)

3.2. Baseline Characteristics of the Biopsy Lesions

Table 2 summarizes the baseline information of biopsy cores in patients. A total of 313 biopsy cores were performed in 147 patients with suspected PCa, including 205 cores with PI‐RADS 4 and 108 cores with PI‐RADS 5. Histopathology revealed 192 positive cores and 121 negative cores in 105 patients. The pathological results of the biopsy cores were scored as follows: Gleason 3 + 3 in 62 cores, Gleason 3 + 4 in 31 cores, Gleason 4 + 3 in 35 cores, and Gleason ≥ 8 in 60 cores.

Table 2.

Baseline characteristics of the biopsy lesions.

Variables		n%
PI‐RADS	4	205 (65.5)
	5	108 (34.5)
Result of PSMA PET/CT (Total = 65)	Consistent with MRI	59 (18.8)
	Inconsistent with MRI	6 (1.9)
Histopathology	Postive	192 (61.3)
	Negative	121 (38.7)
Histological Type	Acinar Adenocarcinoma	176 (56.2)
	Ductal Adenocarcinoma	3 (1.0)
	Intraductal Carcinoma	9 (3.0)
	Small Cell Carcinoma	2 (0.6)
	Urothelial Carcinoma	2 (0.6)
	Benign Lesions	121 (38.7)
Gleason score	3 + 3	62 (33.0)
	3 + 4	31 (16.5)
	4 + 3	35 (18.6)
	≥ 8	60 (31.9)

3.3. High Consistency between ROSE and Histopathology Diagnosis

Table 3 compares ROSE findings with paraffin histopathology results in 147 patients. With paraffin histopathology as the gold standard, ROSE shows a diagnostic sensitivity of 71.9% and specificity of 100%, with PPV of 100% and negative predictive value of 69.1%. Notably, none of the 138 cores diagnosed as malignant by ROSE were finally benign. In this study, ROSE has a diagnostic accuracy of 82.7%. Moreover, there is good consistency between ROSE and histopathology diagnosis (Kappa = 0.664, p < 0.001).

Table 3.

Comparison of prostate biopsy results between ROSE and histopathology.

ROSE	Histopathology		Total
	PCa	Benign lesion
Postive	138	0	138
Negative	54	121	175
Total	192	121	313

3.4. Comparison between the Groups

Table 4 illustrates the accuracy of ROSE across lesions with different Gleason scores. It can be observed that as the Gleason score of the prostate adenocarcinoma lesion increases, the positivity rate of ROSE also tends to rise. Specifically, the highest positivity rate of ROSE was found in lesions with a Gleason 4 + 3 (82.9%), followed by lesions with a Gleason ≥ 8 (73.3%). Additionally, in this study, both lesions of urothelial carcinoma were detected by ROSE, whereas neither of the two lesions of small cell carcinoma was identified.

Table 4.

ROSE results of 188 prostate adenocarcinoma lesions.

Glaeson score	ROSE
	Postive	Negative
3 + 3	43 (69.4%)	19
3 + 4	20 (64.5%)	11
4 + 3	29 (82.9%)	6
≥ 8	44 (73.3%)	16

4. Discussion

Currently regarded as the gold standard preoperative diagnosis for PCa, prostate biopsy is of vital importance for the diagnosis and treatment of PCa. Existing diagnostic guidelines recommend prostate biopsy for patients with elevated PSA levels or abnormal findings on digital rectal examination. For patients with suspicious findings on MRI, a combination of systematic and targeted biopsies is advised [1]. However, as an invasive procedure, prostate biopsy can cause unnecessary harm to many patients. A large‐scale, nationwide population‐based study from the United Kingdom revealed that prostate biopsy is associated with infectious complications such as sepsis, urinary retention, and hematuria [10]. Moreover, due to the local inflammatory response in prostate tissue after biopsy, undergoing surgery in the short term increases the rate of complications [11].

Some studies have begun to explore whether prostate biopsy can be avoided before RP in patients with highly suspicious PCa on imaging. While MRI has been proven to have sufficient sensitivity and specificity in detecting csPCa, it still not only misses about 20% of csPCa cases but also has a PPV of only 50%‐70% [15, 16]. To compensate for the limitations of MRI, various molecular imaging techniques, such as, such asPSMA PET/CT, have been proposed. Although PSMA PET/CT offers higher diagnostic efficacy, it still cannot guarantee absolute diagnostic accuracy [17]. A prospective multi‐center study shows that even when MRI is combined with PSMA PET/CT, although the sensitivity increases to 97%, the PPV remains at 67% [18]. Meissner VH et al. reported on 25 patients with PCa who underwent RP without prior biopsy, using MRI and PET‐CT as the main diagnostic tools preoperatively. However, the PPV for the 52 lesions tested was only 87%–90%, indicating potential unnecessary resection of benign lesions or normal tissue [12].

Therefore, a more reliable rapid diagnostic method is needed to ensure the effectiveness and timeliness of prostate biopsy diagnosis.

ROSE is a rapid, real‐time, on‐site diagnostic technique for interventional biopsies. It has shown promising accuracy in diagnosing various adenocarcinomas, such as lung adenocarcinoma, thyroid cancer, and pancreatic cancer. This might be due to the cytological features of adenocarcinomas, like enlarged nuclei, increased nuclear‐cytoplasmic ratio, and prominent nucleoli, which make the cancer cells easier to detect under a microscope after rapid staining [14, 19, 20]. Typical PCa cells often exhibit similar nuclear morphology, with relatively monomorphic nuclei and prominent nucleoli [21].

This study firstly applies ROSE to prostate biopsy. By analyzing the results of 147 patients with suspected prostate cancer who underwent concurrent ROSE testing during prostate biopsy at Beijing Hospital, it explores the application value of ROSE in prostate biopsy. Through analyzing specificity, sensitivity, positive predictive value, and negative predictive value, this study evaluates the diagnostic efficacy of ROSE for PCa. Our results show that with paraffin histopathology as the gold standard, ROSE has 82.7% accuracy and good consistency with histopathology (Kappa = 0.664, p < 0.001). Most importantly, it has a 100% positive predictive value. This proves ROSE can correctly diagnose prostate cancer by identifying cancer cell morphology. In this study, the average time from obtaining biopsy samples intraoperatively to obtaining ROSE results was about 3 min. ROSE demonstrated consistently high specificity and timeliness in diagnosing PCa, as seen in other fields [19, 22, 23]. If confirmed by large‐scale sample studies, ROSE could allow immediate surgery based on its results, offering benefits like reduced waiting times and less biopsy burden for patients. Also, as a rapid pathology diagnosis method, ROSE might detect positive surgical margins during prostate cancer surgery, replacing intraoperative frozen sections and further shortening waiting times. Given ROSE with 100% positive predictive value, immediate PCa surgery post‐ROSE poses no risk of unnecessary resection.

Although ROSE showed excellent specificity (100%) in this study, its sensitivity (71.9%) was less than ideal, missing a significant number of csPCa cases. On the one hand, PCa lesions are multifocal, making it hard to ensure every biopsy core from the target area contains enough cancer tissue. In this study, for lesions with ≤ 50% cancer tissue in paraffin pathology, ROSE positive rate was significantly lower (67.3% vs 78.8%). On the other hand, as a cytological test, ROSE diagnostic result depends on shed cancer cells. Gleason 5 lesions, packed with tumor nests, yield fewer shed cells. Consequently, ROSE positive rate for Gleason ≥ 8 lesions (73.3%) is significantly lower than for Gleason 4 + 3 lesions (82.9%).

Conventional ROSE requiring pathologist assessment imposes a significant time burden, further limited by the impracticality of continuous on‐site pathologist availability. Additionally, because ROSE reimbursement is substantially lower than that of traditional surgical pathology, pathologists receive inadequate compensation [24]. Therefore, we designed our study with urologists as ROSE operators and pathologists in a training role. All participating urologists completed a 1‐week dedicated ROSE training program. In this study, only trained urologists interpreted ROSE results, which might be slightly less accurate than professional cytopathologists. However, our prior research showed no significant accuracy difference between cytopathologist involved and urologist only ROSE ninterpretatios for prostate cases (82.7% vs. 84.5%, p = 0.664). A meta analysis also indicated no significant accuracy difference in ROSE interpretation between cytopathologists and non‐cytopathologists [25]. However, challenges persist, including the substantial case volume and time investment required to overcome the learning curve, as well as the need for standardized training protocols. Therefore, our future work will focus on AI‐assisted cytopathological analysis for ROSE to further enhance diagnostic accuracy, minimize operator‐dependent variability, and reduce the burden of additional training. Previous studies have demonstrated that AI performance is comparable to pathologists in cytopathological diagnosis, although initial data annotation remains pathologist‐dependent [26].

Compared with standard paraffin pathology, ROSE significantly reduces time to diagnosis and associated costs. Although specific data were not prospectively collected in this study, evidence from other centers supports this advantage. At a US‐based center, ROSE performed by non‐pathologists required a mean of 4.02 min from specimen acquisition to result generation, substantially shorter than the 5–7 days required for processing, fixation, sectioning, and staining in traditional pathology, with costs approximating only 10% of standard pathology [27]. These findings are consistent with observations from our center. Nevertheless, given that paraffin pathology provides additional immunohistochemical information and tumor classification, the cost differential is relatively modest; rather, the extremely rapid diagnostic turnaround represents the potential to transform prostate cancer care pathways. Prospective studies incorporating formal cost‐effectiveness analyses are planned to quantify the clinical benefit of ROSE for patients with prostate cancer.

It should be emphasized that this study merely assesses the potential value of ROSE in prostate cancer diagnosis. However, without prospective, large‐scale, multi‐center studies, ROSE can not yet alter standard diagnostic and treatment pathways.

Our study initially reveals the potential of ROSE in PCa diagnosis, highlighting its high specificity and timeliness. Nevertheless, certain limitations persist. First, this single‐center study had a relatively small sample size, predisposing it to inherent bias. Second, both ROSE diagnostic efficiency and that of conventional prostate biopsy are surgeon dependent, with experienced surgeons significantly enhancing cancer detection rates [28].

5. Conclusions

The application of ROSE technology in the diagnosis during prostate biopsy is accurate and reliable, with specificity and PPV both reaching 100%.

Author Contributions

Hong Ma designed the study and analyzed the data; Anzi Cao wrote the paper; Huiming Hou and Pengjie Wu contributed to the data analysis; Ben Wan contributed to the methodology and interpretation of results; Ming Liu provided critical revisions and supervised the project.

Ethics Statement

This study was performed in line with the principles of the Declaration of Helsinki. Approval was granted by the Beijing Hospital Ethics Committee (Approval Letter No: 2023BJYYEC‐431‐02). Written informed consent was obtained from the all participants.

Conflicts of Interest

The authors declare no conflicts of interest.

Ma H., Cao A., Hou H., Wu P., Wan B., and Liu M., “Diagnostic Value of Rapid On‐Site Evaluation Combined With Prostate Biopsy in Prostate Cancer,” The Prostate 86 (2025): 542‐549, 10.1002/pros.70113.

Data Availability Statement

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