Developing a Scoring System for Preoperative Factors that Affect Stone-free Rate in X-ray free Ultrasound-guided Percutaneous Nephrolithotomy: P.O.N.C.O. Prognostic Score

Background

Current guidelines have recognized Percutaneous Nephrolithotomy (PCNL) as the gold standard of treatment for large and complex renal calculi.^1,2 The surgical technique has evolved thoroughly, improving the effectiveness and safety of surgery. As opposed to conventional fluoroscopy-guided PCNL, X-ray-free ultrasound-guided (XFUS) PCNL has been endorsed as a safe and feasible alternative that reduces radiation exposure to both patients and operators.¹ However, the approach is novel, and limited literature is available on elaborating and evaluating the possible factors that may affect surgical outcomes.^1,2 An assessment to predict the outcome and evaluate the characteristics affecting the stone-free rate is needed.

Notable scoring systems such as the S.T.O.N.E. (stone size, tract length, obstruction, number of involved calices, and essence) score, Guy’s stone score, and The Clinical Research Office of the Endourological Society (CROES) nomogram have been introduced to assess and predict the stone-free rate after PCNL in the fluoroscopy-guided approach.^3–5 This scoring system was developed and validated specifically for X-ray-guided PCNL; thus, its compatibility with XFUS PCNL is uncertain. However, several preoperative factors have been shown to be clinically and statically significant in affecting the stone-free rate, such as stone size, obstruction, number of stones, contour, obesity, and number of involved calyces.^3–5 However, the weight of each factor influencing the stone-free rate in XFUS PCNL has yet to be determined.

Having a standardized scoring method specifically for XFUS PCNL in SFR prediction will be beneficial for clinical practice, as patients could be counseled accurately about the possibility of a stone-free status early on. It may also help the operator advocate technical modification or consider a more conventional approach for the specific condition.⁶ Therefore, this study aimed to determine the possible factors and characteristics that specifically influence the XFUS PCNL SFR and develop the data into a simple and user-friendly tool that is accessible for daily practice.

Methods

This study developed a scoring system to predict residual stones, specifically post XFUS PCNL. A consecutive prospective cohort study was conducted in a tertiary referral hospital in Indonesia between November 2019 and November 2023 in patients who had undergone XFUS PCNL. This study was approved by the Ethics Committee of the Faculty of Medicine, Universitas Indonesia (Institutional Review Board), under ethical approval number KET-79/UN2.F1/ETIK/PPM.00.02/2024, on January 12, 2024. All procedures were performed in accordance with institutional guidelines and adhered to the ethical principles outlined in the Declaration of Helsinki. Written informed consent was obtained from all participants prior to enrollment in the study. As part of the hospital’s standard patient registration form, participants consented to the anonymous use of their medical data for academic and research purposes. No identifiable personal information (such as full name, email address, or phone number) was collected or stored, ensuring patient confidentiality. The ethics committee approved the data collection and consent procedure.

All patients with caliceal, pelvic, and upper ureteral stones with a stone burden ≥20 mm were included. Patients with uncorrected coagulopathy, congenital kidney anomalies, or intraoperative conversion to fluoroscopy were excluded from the study. Demographic data, preoperative characteristics, and operative data were collected prospectively. Preoperative evaluation of the stones was performed using non-contrast abdominal computed tomography (CT-scan. The proportion (stone burden) was calculated using the cumulative size of the largest axis of each stone in cases of multiple stones. Staghorn stones were defined as stones occupying the renal pelvis and a minimum of one infundibulum. Hydronephrosis was graded using the SFU grading of hydronephrosis using CT-Scan. Stone-free status was evaluated based on two criteria: 1) direct visualization using the same rigid nephroscope used in PCNL and USG evaluation of the kidney immediately after stone extraction; 2) postoperative kidney-ureter-bladder (KUB) photo for patients with radio-opaque stones and Kidney USG for patients with radio-lucent stones taken 24 hours post procedure; and 3) Repeat Kidney USG 7 days post procedure. Patients with significant residual fragments (>4 mm) from KUB/USG or doubtful radiolucent stones were examined using repeat non-contrast CT-scan. Multiple residual stone fragments ≤4 mm in size were measured for cumulative stone burden and deemed significant if they were >4 mm.

Results

Model discrimination

The scoring model showed excellent discrimination in random sampling with an AUC of 0.845 (95% CI 0.71-0.98; P = 0.002), as assessed by the ROC curves depicted in Figure 1. The model also showed good calibration, with a non-significant Hosmer–Lemenhow Chi-square of 1.025 (P = 0.96). Comparison between the poor- and good-risk groups from the P.O.N.C.O. prognostic score using the cut-off was statistically significant (P=0.006), as depicted in Figure 2.

Figure 1. Area under the receiver operating characteristic (ROC) curve in validation cohort.

Figure 2. Comparison between Poor risk and good risk group from P.O.N.C.O. prognostic score for occurrence of residual stone. (P-value measured by chi-square).

Discussion

This study presented a novel prognostic scoring tool for residual stones in XFUS PCNL. We developed a multivariable tool with good discrimination, calibration, and clinical utility in cohorts of patients, using five demographic and preoperative characteristics that are easily accessible to the surgeon.

Among the five independent factors included in the study, all have been claimed in previous studies to be potential risk factors for residual stones after PCNL. The stone-free rate has been suggested to be highly dependent on stone burden in the literature.^7,8 Preminger et al. showed that the higher the stone burden, the less likely it is that PCNL would be able to serve as a monotherapy.³ The findings in our study supported the claim that a cut-off of 30 mm seemed to increase the odds of residual stone post-surgery. However, in contrast to other scoring systems, we found that stones > 40 mm were weighted significantly lower than those sized 30-40. Using multivariate analysis, it seems that stone burden is closely related to other predictors (i.e., contour and number of stones); when we remove the confounding factors, it also influences the weight of the effect. Contour is a variable that was also known as a risk factor for residual stone. Mishra et al. elaborated in their staghorn morphometry study that staghorn stone is independent of stone burden in influencing residual stone, despite being very closely related.⁹ Our study also supports this claim by showing similar results in the multivariate logistic regression analysis. The obstruction or hydronephrosis status also plays an important role in predicting residual stones. Our study showed that patients with hydronephrosis had higher odds of developing residual stones after surgery.¹⁰ BMI and the number of stones have also been suggested by other studies as predictors of residual stones after surgery.^4,11,12 Similarly, they were also independent predictors in multivariate analyses.

Multiple studies have suggested the comparability of X-ray-free ultrasound-guided PCNL in terms of its efficacy and safety in fluoroscopy-guided treatment of renal stones.^13–15 However, a notable difference between the two is the radiation exposure. Chen et al. raised concerns about radiation exposure hazards during the treatment of nephrolithiasis.¹⁶ Although low-dose fluoroscopy is readily available, the stochastic effects that are independent of radiation dose but dependent on the amount of exposure to radiation increase the risk of malignancy.¹⁶ This not only affects the patient but also the surgeons who are more exposed to radiation in their daily X-ray assisted surgeries. As this problem persists, X-ray-free ultrasound-guided PCNL could be a modality of choice. However, as fluoroscopy-guided PCNL has long been the standard approach, a tool that may support familiarization and increase the learning curve for X-ray-free ultrasound-guided PCNL would benefit current practice. Stratifying the difficulty of the patient by creating a low-risk residual stone group will benefit surgeons who are trying to learn X-ray-free ultrasound-guided PCNL by patient selection during surgery.¹⁷ Additionally, this practical tool should help educate the patient regarding the outcome of the surgery and provide proper consideration in planning the surgery. Higher risk patients should be notified of a high possibility for a second surgical procedure due to the residual stone and should be considered for additional fluoroscopy support.

To the best of our knowledge, this is the first prognostic score for residual stones in the treatment of renal calculi with X-ray free USG-guided PCNL. Previously known residual stone scorings were made for PCNL using a fluoroscopy approach, such as the S.T.O.N.E. (stone size, tract length, obstruction, number of involved calices, and essence) score,³ Guy’s stone score,⁴ and The Clinical Research Office of the Endourological Society (CROES) nomogram.⁵ Although some of the variables included in the model are similar, a direct comparison is not straightforward. There are also other limitations to our study, such as operator bias within the study (since we only included two operators, and different operators might have different preferences (i.e., position and puncture approach]). In addition, the learning curve and surgical experience of both operators were different at the start of the study. Further investigation in future studies should be conducted on a larger scale involving more operators and preferably multicenter studies to remove bias. There should also be a subsequent study to validate the scoring system to elaborate on the reproducibility and consistency of this scoring system.

Conclusion

This study presents the first prognostic scoring tool for residual stones in X-ray-free ultrasound-guided PCNL with good discrimination, calibration, and clinical utility in cohorts of patients, using five predictive factors that are easily accessible for the surgeon to collect in daily practice.