Predicting positive surgical margins in patients treated with robot-assisted partial nephrectomy: results from a prospectively maintained dataset of a single tertiary referral center

INTRODUCTION

Partial nephrectomy (PN) has been widely accepted as the preferred treatment for patients with clinical T1 renal cell carcinoma (RCC)^[1]. This method not only offers the advantage of preserving renal function but also has oncologic outcomes similar to radical nephrectomy^[2,3]. However, one challenge in PN is the potential for positive surgical margins (PSMs). PSM rates after PN have been reported to range from 0% to 10%^[4]. This risk appears to be higher in surgeries involving smaller, high-grade tumors or those considered imperative, regardless of the surgical approach used^[5,6].

The implications of PSMs on cancer recurrence and patient survival are still debated among experts. While some studies suggest that local tumor recurrence is more common in cases with PSMs, especially for inherently aggressive tumors, others find minimal impact of PSMs on cancer-specific survival^[7,8]. Given the lengthy progression of T1 RCC after surgery, the PSM could be used as a surrogate of oncologic outcome after surgery, and, as such, it is an essential measure for evaluating outcomes after PN. Evidence suggests that patients with PSMs might experience reduced overall survival; however, other factors related to patient characteristics, such as age and comorbidities, and tumor characteristics, such as tumor fat invasion and nucleolar grade, might play a role^[9]. In the current literature, there is a noticeable gap in models designed to predict the likelihood of PSMs after PN.

The present study aims to fill this gap by identifying predictors of PSMs after PN based on data from a prospectively maintained database of a tertiary referral center.

METHODS

RESULTS

From the 1,611 patients analyzed, Table 1 presented the demographic and clinical characteristics, with 1,120 (69.5%) males and 491 (30.5%) females. The median age was 62.6 years, with an IQR of 54.7 to 72.0 years. The BMI had a median of 26.9 kg/m² (IQR: 23.9-28.8), and 299 (18.6%) patients had an Eastern Cooperative Oncology Group (ECOG) score ≥ 1. American Society of Anesthesiologists physical status (ASA PS) score ≥ 3 was noted in 340 (21.1%) of the patients. The median CCI was 1 (IQR: 0-2), while the median age-adjusted CCI performance status (AA-CCI PS) score was 3 (IQR: 2-5). Regarding surgical indication, 1,458 (90,5%) were elective, 207 (12.8%) were relative, and 54 (3.4%) were imperative. For the tumor side, 811 (51.4%) were on the right and 800 (48.6%) were on the left.

The PADUA score median was 8.0 (IQR: 7.0-9.5). Baseline hemoglobin, creatinine, and estimated glomerular filtration rate (eGFR) were 13.9 mg/dL (IQR: 13.3-15.2), 0.9 mg/dL (IQR: 0.8-1.1), and 84.6 mL/min (IQR: 69.5-100.4), respectively.

In Table 2, surgical details and histopathology were reported. Pedicle clamping showed 217 (13.5%) patients were off-clamp and 1,394 (86.5%) were on-clamp. The median warm ischemia time was 16 min (IQR: 10-21). A retroperitoneal access was employed in 138 (8.5%) patients. Clear cell RCC was the predominant histotype, seen in 1,134 (70.4%) patients. SIB score distribution revealed that 1,129 (70.1%) patients had a score of 0-2, 468 (29%) had a score of 3-4, and 14 (0.9%) had a score of 5. The PSM was observed in 109 (6.7%) patients.

Table 3 displayed two models of the multivariable logistic regression analysis.

In the initial model, imperative surgical indication exhibited an OR of 6.06 (95%CI: 2.58-14.22, P < 0.001). SIB > 1 had an OR of 2.37 (95%CI: 1.43-3.92, P = 0.001), and off-clamp tumor resection demonstrated an OR of 3.00 (95%CI: 1.41-6.39, P = 0.004). The PADUA score resulted in an OR of 1.10 (95%CI: 1.03-1.17, P = 0.006). Age and the American Society of Anesthesiologists (ASA) score did not show statistical significance at multivariable analysis. In the second model, the PADUA score was replaced by collecting system tumor compression as a significant predictor (OR 1.34, 95%CI 1.12-1.84, P = 0.001), while tumor diameter did not reach significance at multivariable analysis.

DISCUSSION

In oncology surgical practices, the comprehensive excision of tumors significantly influences the surgical efficacy. In PN, notwithstanding the absence of consistent consensus in clinical studies, a heightened likelihood of local recurrence appears to be associated with patients presenting PSMs^[13]. It is pertinent to highlight that empirical evidence from various research studies indicates an elevated risk of recurrence for patients with PSMs, especially in the context of high-risk tumors characterized by determinants that intrinsically exhibit a pronounced relapse potential, such as escalated tumor dimensions, pT3a stage, and elevated grading^[14,15]. Conversely, in primary tumors of a lower grade, PSMs are postulated to exhibit reduced malignant potential owing to a diminished rate of cancer progression^[16]. Furthermore, the obliteration of malignant cells due to coagulation, mechanical perturbations, or induced ischemic damage during nephron-sparing surgery (NSS) might curtail the persistence and proliferation of neoplastic cells at the excision margin^[17].

In our study involving 1,611 patients, demographic and clinical characteristics were extensively analyzed. Notably, 109 (6.7%) of them had PSMs. Within the realm of surgical oncology, understanding the influence of various factors on outcomes, especially when dealing with PSMs, is crucial. Our multivariable logistic regression analysis to detect clinical predictive factors associated with a higher rate of PSMs revealed several pivotal findings.

The SIB score is an evaluative measure derived post-surgery that standardizes the thickness of healthy renal margins assessment around the tumor. In simple terms, it categorizes specimens into enucleation, enucleoresection, or resection based on the aggregate score of three distinct macro-areas. Our analysis highlighted a notable odds ratio (OR) of 2.37 for SIB > 2 (P = 0.001), highlighting its consequential influence on PSMs. The reason for such a result warrants an in-depth analytical discussion: indeed, the uncomplicated enucleation of the tumor along the pseudocapsule, in contrast to a conventional excision of the adjacent healthy parenchyma, might allow for better respect of the natural cleavage plane. This approach reduces the risk of inadvertently penetrating the neoplasm, especially in instances of anomalous configurations, and inadvertently leaving remnants of the SM.

Surgical indications allow for the classification of cases based on the necessity and background of the surgical procedure. They range from elective, for standard cases of unilateral RCC, to imperative, in more pressing situations such as bilateral tumors or when dealing with a solitary kidney^[18]. In the multivariable assessment, imperative surgical indications registered a substantial OR of 6.06 (P < 0.001), underscoring its indispensable role in shaping surgical outcomes.

The PADUA Classification, which offers a gradation system for nephrometric complexity of renal tumors, is an instrumental determinant for surgical planning^[19]. PADUA score was significantly associated with PSMs. This aspect has already been underscored in other multicentric cohorts^[20], highlighting the necessity for careful management due to the higher risk associated with complex masses. To better evaluate which nephrometric aspects could be associated with a higher risk of PSMs, an additional multivariable analysis was conducted, excluding the PADUA score and including tumor dimension and collecting system compression, which showed significant associations with PSMs in univariate analysis. These factors were consequently included in our multivariable analysis model, and it showed collecting system tumor compression as a significant predictor (OR 1.34, 95%CI 1.12-1.84, P = 0.001).

Among intraoperative surgical factors, off-clamp tumor resection held notable significance with an OR of 3.00 in the PSM prediction, reinforcing its critical role in the outcome^[21]. The choice to opt for off-clamp tumor resection in surgeries, while advantageous in some scenarios, presents its own set of challenges^[22]. The off-clamp approach may compromise the surgeon’s visual field, obscuring the clear delineation of structures and potentially decreasing the quality of the resection^[23]. Some experts advocate for optimal hilar clamping specifically to circumvent this limitation. However, it is imperative to note that the literature offers varying perspectives on the matter: in fact, some systematic reviews examining the potential impact of renal pedicle management techniques on partial nephrectomies have found that the specific technique employed - whether clamping or not - does not necessarily influence surgical or oncological outcomes^[24]. This suggests that while visualization and surgical field clarity are undoubtedly crucial, other factors, skills, and methodologies also play pivotal roles in determining the surgery’s success^[25].

In discussing these results, achieving tumor-free margins during PN is an essential oncological goal, with implications for surgical success. While literature suggests varying views, local recurrence appears more probable in patients with PSMs, particularly in the context of high-risk tumor characteristics^[14]. In contrast, low-grade primary tumors exhibit less aggressive PSM potential due to their slower cancer progression rate^[15]. Some believe the possibility of tumor cell destruction during NSS might mitigate malignant cell proliferation at resection sites^[16].

Another aspect that warrants attention is the influence of surgeon expertise and, more in general, the surgical caseload of the center on PSMs. The RECORD 2 study, in fact, underscored the significant contribution of the annual PN volume in influencing PSM outcomes, stressing the importance of centralization of this surgery in high-volume centers for this treatment^[26]. Notwithstanding our meticulous approach, inherent limitations such as potential biases due to the absence of information about surgeon expertise and variations in surgical techniques across surgeons must be acknowledged^[27,28].

Our study has inherent limitations. Firstly, although data collection was maintained prospectively in our internet-based platform, the nature of our retrospective analysis could introduce biases. Secondly, certain parameters such as clinical tumor size and deployment of intraoperative tools (e.g., ultrasound, near-infrared fluorescence) were unavailable in this database. Therefore, their potential influence on PSMs remains unexplored in this analysis. Novel technological advancements in robotic surgery and 3D modeling might further enhance surgical precision^[29,30]. Thirdly, the unicentric nature of this study could present potential variation in practices and techniques. Fourthly, our research did not involve a centralized pathological review, which could introduce variability in the reporting of PSMs based on individual practices within the institution. Fifthly, we did not detail the specific experience or learning curve of each surgeon involved in the procedures. Consequently, we did not consider the potential variance in outcomes due to surgical experience, relying instead on the overall center caseload as an approximate metric of experience.

It remains crucial for our model’s predictions concerning PSMs to be tested on other cohorts to ensure their applicability and relevance to a broader population undergoing PN.

In conclusion, the primary aim of PN, from an oncological perspective, is to secure negative margins. Based on our prospectively maintained dataset, tumor’s (higher PADUA score and, in particular, collecting system tumor compression) and surgical (imperative indication, SIB score > 2 and off-clamp approach) features represent the main drivers for PSMs. Further study remains necessary to assess the implication of PSMs on survival outcomes.