Discussion
The current European Association of Urology (EAU) and American Association of Urology (AUA) guidelines recommend PCNL for the treatment of renal stones sized >2 cm.19,20However, due to the potential advantages of fURS (e.g., not causing renal parenchymal damage and severe bleeding, applicability in patients with bleeding diathesis or those receiving anticoagulant therapy, short length of hospitalization, and daily work routine not being restricted) and its ability to access almost all calyceal stones as a result of improvements in deflection, fURS has become a preferred method for the treatment of both proximal ureteral and renal stones.2,21 In a recent meta-analysis, the final SFR was reported to be 89.4% in an average of 1.4 procedures performed in 2-3 cm stones, and this rate was stated to be comparable to PCNL.22 Although complication rates of up to 16% have been reported in previous studies, most were classified as minor. In addition to the development in fURS technology, increasing surgical experience has reduced the rate of major complication from 5.01% between 1990 and 2011 to 1.48% between 2011 and 2016 and increased the success of treatment.23 While semi-rigid URS is sufficient in most cases in the treatment of ureteral stones, performing a procedure without fURS in the treatment of middle and upper ureteral stones creates problems in terms of medicolegal aspects. In a study published by the Clinical Research Office of the Endourological Society ureteroscopy study group in 2014, it was reported that middle and proximal stones were both larger and difficult to reach compared to distal stones. In the same study, it was emphasized that the risk of perforation due to impaction was high in middle ureteral stones and low-caliber URS should be used.24 fURS, which has a wide range of treatment options in terms of both localization and stone burden and is even preferred in much larger stones to avoid PCNL-related complications, requires further investigation in terms of how to predict its outcomes. Almost all the scoring systems used in fURS have been developed for renal stones, and STONE, which is an option for ureteral stones in which fURS will be performed11,12,14,15, is based on low stone burden and use of old devices. Hori et al. identified this gap in the literature and defined a scoring system, which they named T.O.HO., including the stone diameter, localization and density parameters and covering patients with larger stones that are planned to undergo fURS.16 The authors reported that T.O.HO. had better predictive value than STONE (AUC = ​​0.833 and 0.633, respectively).
In this study, the external validation of the T.O.HO. and STONE scores was performed, and SFS-related parameters after fURS were evaluated. The overall SFR was found to be 79.9%, and SFR in renal and ureteral stones was 77.6% and 90.5%, respectively. Our SFR was similar to the rate reported in the original study. In addition, in our study, the SFS prediction accuracy (AUC value) was calculated as 0.758 for T.O.HO. and 0.634 for STONE. T.O.HO. The cut-off value of T.O.HO. was determined as 8, at which it had 71.4% sensitivity and 68.8% specificity in predicting SFS. STONE was able to predict SFS with 57.5% sensitivity and 63.2% specificity at a cut-off value of 11. When compared to the original T.O.HO. study, the modified T.O.HO. score had lower specificity but similar sensitivity (AUC = 0.758). This difference may be due to the large stone sizes in our study and the lower number of patients with ureteral stones compared to the original TOHO study. The modified T.O.HO. scoring system was observed to have better predictive value than the original version (AUC = 0.821 and 0.758, respectively). At a cut-off value of 7, the modified T.O.HO. scoring system was able to predict SFS with 71.2% sensitivity and 80.8% specificity.
Many studies in the literature have shown that stone burden is the most important parameter affecting SFS after fURS.11,12,14,15 In the T.O.HO. scoring system, stone burden was reported to be the most important predictive value.16 We also determined that stone burden was associated with SFS (p < 0.001). The effect size of stone burden was clearly demonstrated by the constructed nomogram (Figure 1). Stone diameter is widely used in clinical practice since it is simple and easy to obtain in the assessment of stone burden.25 The EAU and AUA guidelines also include stone diameter in their recommendations concerning decision-making with regard to the treatment of urinary system stones.19,20Hori et al. used stone diameter while evaluating stone burden and categorized it over 5 points based on the effect size obtained from the nomogram. They reported that treatment success decreased by <30% in patients with 5 points.16 However, since stone diameter does not reflect the width and depth of the stone, it will naturally have certain limitations in predicting the results of the operation compared to stone volume. Ito et al., evaluating patients who underwent fURS, emphasized that stone diameter was able to accurately predict stone volume in <2 cm stones but it was necessary to directly calculate stone volume in stones larger than 2 cm.25 Considering that stone volume increases exponentially as stone diameter increases, this result is expected. Today, with the developments in technology and increase in experience, it is possible to apply fURS treatment to larger stones; therefore, it would not be realistic to expect stone diameter alone to predict success. Supporting this, in our study, the rate of treatment success was 42% in the patients with 5 points in stone diameter (≥30 mm) according to the original T.O.HO. score while it was only 21% for those with 5 points (>480 mm2) according to the modified T.O.HO. score, in which stone area rather than diameter was evaluated. Hori et al. also stated that the STONE scoring system, which has different cut-off values, does not have predictive value for stone size classification.16 Consistently, we found that the patients scoring 3 points (>10 mm) in the stone diameter of the STONE scoring system had a treatment success rate of 79%. This indicated that the stone size classification of the STONE scoring system was far from differentiating SFS.
Another component of the T.O.HO. scoring system is stone localization. Studies have shown that stone localization is an independent marker in the treatment of ureteral and renal stones, and especially lower pole stones are associated with fURS treatment failure.11,12,14,15,19,26 For practical use, T.O.HO. classified renal stone localizations as upper, middle and lower pole and ureteral stones as proximal, middle, and distal. In our study, it was observed that the rate of SFS was 71.2% in lower pole stones and 66.7% in multi-calyceal stones, while it was 89.5%, 91.2% and 87.9% for middle ureteral, proximal ureteral and pelvic stones, respectively. However, the authors that developed T.O.HO. did not specify how multi-calyceal were graded in this scoring system. In order to continue the validation process, we scored multi-calyceal stones containing those with middle and upper pole localizations and similar SFR as the upper and middle pole group, and multi-calyceal stones with low SFR located in the lower calyx as the lower pole group. According to the T.O.HO. score based on stone localization, the worst SFR was in the lower pole, and this was at a statistically significant level (p < 0.001). In the multivariate analysis, it was determined that the middle ureter and middle pole stones provided an increase of 76.2% and 50.4% in the operation success, respectively, compared to the lower pole stones. In the original T.O.HO. study, lower SFR (51.6%) was reported in the upper pole stones than in the lower calyceal group, whereas in our study, higher SFR (82.4%) was found in the upper pole stones similar to the middle pole stones. This difference was attributed to the small number of patients with upper pole stones in both studies and presumably different stone sizes.
As an important parameter in the treatment of urinary system stones, stone density is also a component of the T.O.HO. scoring system. The relationship of stone density with SFS has been shown in many studies.12,27 Hussain et al.28 used the cut-off value of T.O.HO. stone density as 1100 HU and graded the cases over 3 points. In our study, it was observed that stone density was an independent marker for SFS in the multivariate analysis, and the cut-off value was calculated as 1125 HU. A 100 HU increase in stone density increased treatment failure by 1.1 times. The STONE scoring system, which has a different cut-off value for stone density, was also found to have similar predictive value for SFR (AUC = 0.570 for STONE and T.O.HO for and 0.581). In our study, according to the stone density score, SFR was determined as 83.7%, 84.0%, and 72.2% for 1, 2 and 3 points, respectively. There was no difference between 1 and 2 points in terms of SFS (p > 0.05). We consider that the HU value can be reduced to 2 points for a practical scoring system. However, in the current study, we left the HU prediction values ​​as in the original system since it would not further increase the predictive value of the modified system.
In addition to the three main parameters explained above, many other parameters have been defined in the literature to be associated with SFS after fURS, such as the number of stones, preoperative stenting, presence of hydronephrosis, and operator experience.11,12,14,15 The STONE scoring system uses preoperative stent application, number of stones, and presence of hydronephrosis as predictive factors.12 In the original T.O.HO. study, Hori et al. reported both parameters to be associated with SFS but found no independent marker in the multivariate analysis.16 Similarly, in our study, the presence of multiple stones was statistically significantly associated with SFS, but it was observed that there was no independent marker in the multivariate analysis. Since stone burden is directly related to the number of stones, the latter loses its importance. We did not determine preoperative stenting to be associated with SFS. There are publications in the literature stating that preoperative stenting increases the success of ureteral access sheath and is not associated with SFS.29,30 Another parameter included in the STONE scoring system is the presence of hydronephrosis. Hori et al. did not evaluate the presence of hydronephrosis. In our study, although the presence of hydronephrosis was high in patients with residual stones, it was not found to be a statistically significant parameter. In the nomogram developed by Ito et al., the presence of hydronephrosis had very low power but it was not used as a marker in other scoring systems.15 The same authors also used operator experience as a marker in their nomogram. Since all surgical procedures in our study were performed by experienced endourologists, similar to the original T.O.HO. study, this marker is not discussed further.
Our study has certain limitations. The main limitations are retrospective design, relatively small number of patients, and lack of data on second-session attempts in patients with residual stones and final success rates. Another important limitation is the lack of stone composition that may affect SFS.