Comment
Recent technical improvement including miniaturization, improved deflection and enhanced optical quality have increased the use of flexible ureterorenoscopes in renal stones. Although f-URS is not recommended as a first line treatment modality for renal stones <2cm at current guidelines, a recent meta-analysis resulted an average of 91% stone-free rate with acceptable complication rates and with an average of 1.45 procedures per patient (range 1.11-1.82) [6,8]. As being more widely used, we need some scoring systems that can be used to predict the success and potential complications after f-URS for counseling of the patients preoperatively.
The variables that were used in these four studies were not standard. Although abnormal renal anatomy is one of the parameters of RUSS, patients with musculoskeletal and renal abnormality directly excluded in R.I.R.S. Also, renal anomaly was not studied in modified S-ReSC. IPA was an independent predictive factor in RUSS and R.I.R.S., however it was not studied in modified S-ReSC and nomogram by Ito et al. So, if a variable was not evaluated in a study we cannot know if it is an independent predictive factor or not.
The success was defined as stone-free status in all scoring systems. However, no standard imaging modality was used and also the timing of imaging was different between studies. And thus, there can be a significant bias between studies. In R.I.R.S stone-free status was evaluated with kidney-ureter-bladder (KUB) plane graphy at 1st month and >2mm fragments were accepted as not-stone free, computed tomography (CT) was used whenever necessary. In the nomogram of Ito et al. stone-free status was evaluated with non-contrast CT at 3rd month. In RUSS, treatment success was defined as stone-free or residual fragment ≤1mm at 1st month non-contrast CT imaging. In modified S-ReSC stone free was defined as no evidence of residual stone at 1st month NCCT.
The success of a procedure can be defined as to reach a targeted end-point without any complication or with acceptable complication rates. Thus, an ideal f-URS scoring system scoring system should foresee complications besides stone-free. None of the scoring systems examined the relation of complexity scores and complications. Only in R.I.R.S. authors indicated that the score is correlated with operation time and thus it can be used to predict to leave the operation for a second session and complications can be avoided. All four scoring systems were able to predict complications in the present study, but only nomogram by Ito et al. could predict Clavien ≥2 complications.
Stone burden is one of the most important predictors of stone-free after f-URS as we know from previous studies [8,9]. As expected, stone burden was an independent predictor of success in all scoring systems except modified S-ReSC. This is unsurprising for modified S-ReSC, because the mean stone burden was significantly lower compared to other studies. The authors indicated that patients with high stone burden were underwent percutaneous nephrolithotomy and excluded from the study. Although modified S-ReSC can be predicted stone free status even in our study, it can be no doubtfully said that success of f-URS for 1cm stone cannot be equivalent to 3 cm stone in the same location that both was given same score in S-ReSC.
There are two studies comparing the current nomograms. Erbin et al. compared the RUSS and modified S-ReSC and found out that RUSS, musculoskeletal deformity and stone size were independent factors in logistic regression analysis [10]. And the AUC values were not satisfactory in terms of nomogram’s predictive accuracy (0.655 and 0.596 for RUSS and modified S-ReSC, respectively). In another recent study by Richard et al. RUSS, modified S-ReSC, S.T.O.N.E and Ito’s nomogram were compared [11]. The major limitation of that study was evaluation of the stone-free status that they made it via visual inspection using ureterorenoscope and fluoroscopy at the end of the procedure in the absence of clinical complications. They concluded that all scoring systems were predicted the stone-free status but none of them were predictive of complications after RIRS. Different from the former study all nomograms could predict the stone-free status and complications in our study as being the Ito’s nomogram most sensitive.
In the present study all four nomograms were underwent external validation in a significantly high number of patient population. Although all nomograms were predictive of stone-free status, the AUC values were lower than that in original study and were not satisfactory. The AUC values in original study and in our study were 0.806 vs 0.657, 0.87 vs. 0,697 and 0.904 vs.0.690 for modified S-ReSC, Ito’s nomogram and R.I.R.S., respectively. The AUC value was not calculated in RUSS.
Strengths of our study include high number of study group, standardized post-operative imaging at a determined time period and various cases including different anatomical abnormalities. Limitation of our study is its retrospective design.
We think that none of these four scoring systems is ideal. Because the success definition is not standard and studied variables are limited. To constitute an ideal nomogram a multi-center study examining the all known potential variables with a strict success definition in a defined time period can be planned with high number of patients. The ideal scoring system could also be used in daily clinical practice. In clinical practice it is hard to calculate stone burden by measuring the diameter of all calculi (especially in multiple stones) or to measure the IPA. In our opinion a stone score should be given by radiologists in imaging reports whenever a widely used ideal nomogram can be developed.