Prospective evaluation and classification of endoscopic findings for ureteral calculi

Difficulty in performing ureteroscopic lithotripsy (URSL) depends on endoscopic findings surrounding calculi. In this multicentre prospective cohort study of 185 patients with a single ureteral stone who underwent ureteroscopic lithotripsy registered in the SMART study between January 2014 and February 2017, we established a classification of endoscopic findings and analysed risk factors for ureteral changes. We evaluated endoscopic findings (oedema, polyps, ureteral mucosa-stone adherence, and distal ureteric tightness) based on the SMART classification. Operative time and ureteral injuries were significantly correlated with endoscopic finding grades. Multivariate analyses revealed that mucosa-stone adherence (MSA) was strongly affected by hydronephrosis grade (odds ratio, 12.4; p = 0.022) and the interval before surgery (odds ratio, 1.10; p = 0.012). The cutoff value for MSA was 98 days, with a predictive accuracy of 0.78. Risk factors for distal ureteric tightness were age (odds ratio, 0.96; p = 0.004) and early intervention (odds ratio, 0.90; p = 0.023). The cutoff value was 34 days, with a predictive accuracy of 0.72. In conclusion, appropriate intervention around 34 days (limited to 98 days) after symptom onset is necessary for treating ureteral calculi. Even if intervention passed 98 days post-symptom onset, staged URSL, alternative procedures, and detailed informed consent should be planned in advance, assuming strong MSA.


Association between surgical outcomes and endoscopic findings.
Operative time was prolonged significantly as the grade level increased for oedema (p < 0.001), polyps (p = 0.005), and MSA (p < 0.001). With respect to SFR, there were no significant differences in terms of grade levels between these EFs. In contrast, patients with grade 3 DUT had significantly lower SFR because of the impossibility of UAS insertion. With regard to ureteral injuries, the incidence and degree of ureteral damage at the stone site increased significantly with increased EF grade, including oedema (p < 0.001), polyp (p < 0.001), and MSA (p < 0.001). Significant association between ureteral injuries due to UAS placement and EF grade was not detected. These results are also shown in Table 4.
Predictive factors for prolonged surgical time. We analysed the predictive factors for prolonged operative times of more than 90 min (  Table 6b. Large stone size (≤ 25 mm 2 vs. ≥ 50 mm 2 ; OR 9.50; p < 0.001) was an independent risk factor for the formation of oedema. Stone location (proximal vs. middle ureter: OR 3.24; p = 0.015) was an independent risk factor predicting polyp formation. Severe hydronephrosis (grade 0 or 1 vs. 2 or 3: OR 12.4; p = 0.022) and prolonged time interval until surgery (OR, 1.10; p = 0.012) were independent factors for the development of MSA to ureteral mucosa. Receiver operating characteristic (ROC) curve analysis revealed Table 3. Association between endoscopic findings and period until intervention.     (Fig. 1b). This cutoff had a sensitivity of 76.9% and a specificity of 66.9%).

Discussion
In this multicentre prospective cohort study, we proposed the Skilled Endoscopic Management of Renal and Ureteral Stones (SMART) classification of ureteral changes based on EFs. To the best of our knowledge, this is the first study to classify EFs in this scenario and use the SMART classification to predict surgical difficulty. This classification worked well for evaluating EFs in terms of the association between grades of EFs , operative times, and intraoperative ureteral injuries at the stone site. Kuroda et al. reported that stone volume, HUs, operator experience, gender, preoperative stenting, and UAS diameter predicted operative times for fURSL 8 . However, that study did not include intraoperative EFs. We found that EFs but not stone volume, including oedema and MSA, were independent predictors for prolonged URSL (more than 90 min). If oedema and polyps prevent a clear view of the calculi, retrograde pressure irrigation and careful laser manipulation are needed to avoid miss-shooting. Embedded stones in the ureteral mucosa require time to be peeled off using URS. It is important to accurately predict operative time, because a longer surgical time represents a risk factor for severe complications 9,10 . Patients with grade 3 DUT had significantly lower SF rates due to failure of URS insertion. Ito et al. found that stone volume and cumulative stone diameter were independent predictors for SF status after single fURSL 11 .  www.nature.com/scientificreports/ However, this did not include intraoperative EFs. If a r-URS or UAS cannot be inserted into the stone site through the distal ureter, it is difficult to achieve SF status without complications 12 . URSL for impacted stones showed lower SFR and higher intraoperative complications than that for nonimpacted stones 5,13,14 . It is important to accurately predict impacted stones and ureteral condition at their surroundings to achieve SF status without complications. There are some reports analysing predictors for stone impaction. However, the results were not consistent because of varying definitions of impacted stones in practice 5,7,14,15 . Legemate et al. reported that female sex, ASA-score > 1, positive urine culture, prior stone treatment, and large stone burdens were predictors for impacted ureteral stones 5 . We previously reported that younger age, stones located at the middle ureter, and ureteral wall thickness (UWT) as measured at the stone site using preoperative non-contrast computed tomography 14 were independent predictors, and a UWT of 3.49 mm was the optimal predictor for cutoff value. We also reported that higher UWT was associated with unfavourable EFs, including oedema, polyps, and MSA 14 .
We found that large stone size was a factor influencing the formation of oedema. Ureteral oedema may arise from ischemia secondary to chronic pressure or may be an immunological reaction to the stone material 7,16 . Large stones might be caused by chronic ischemia because pressure on the ureteral wall leads to the formation of oedema. Because some oedemas might continue to grow, large stones should be treated as soon as possible.
Fibroepithelial polyps may form as a result of repeated inflammation of ureteral tissues 17 . Idiopathic ureteral polyps develop most frequently in the proximal ureter 18 . Nevertheless, we found that stones located at the middle ureter were risk factors for the formation of polyps. This result was also shown in our previous report 14 . With respect to polyps associated with urolithiasis, anatomic narrowing, with the ureteral crossing the iliac artery, may lead not only to disturbing spontaneous stone passage, but also chronic inflammation.
With respect to MSA, it is important to remove embedded stones completely, because residual stone fragments may provoke inflammation, leading to stricture formation 19 . In the present study, the independent risk factors were severe hydronephrosis and prolonged time interval until surgery, with a cutoff value of 98 days. The mechanism of the formation of MSA has not been elucidated clearly. However, a chronic ureteral inflammatory reaction caused by the immobility of stones and persistent irritation may induce the submucosal migration and epithelisation over the stones. We speculated that once the impacted stone in the ureter led to hydronephrosis and a tortuous ureter, the chronic inflammatory reaction around the stone might worsen with increased ureteral peristalsis. Finally, it would induce MSA, which might lead to more severe hydronephrosis with a longer time interval until surgery. Thus, in patients suspected of severe adherence with a more than three-month time interval after symptom onset or with severe hydronephrosis, retrograde URSL with nephrostomy or percutaneous antegrade URSL may be better alternatives 20,21 .
It is important to observe distal ureteral findings at the beginning of URSL to choose the size of UAS. Traxer et al. reported that 13.3% of patients had severe ureteral injuries due to 12/14Fr UAS placement 22 . In contrast, in this study, 1.7% of patients had severe ureteral injuries due to UAS placement because we chose the UAS size according to DUT grade. In another report, 16% of urologists expressed concerns regarding the failure in UAS placement 23 . If a UAS is inserted into a narrow-caliber ureter with excessive bulking force, it may induce injury, avulsion, or secondary stricture. If a DUT is classified as grade 2 or 3, URSL should be performed without a UAS, or it can be terminated with stent placement and a staged procedure can be performed later. Recently, preoperative alpha blockade has been reported to reduce the maximal UAS insertion force and the consequent risk of ureteral injury. In addition, pre-stenting was a predictor for successful UAS insertion 24 . This is why pre-stenting   25 . Presumably, progressive expansion in ureteral diameter with aging may be associated with these results 26 , in addition to the loss of surrounding muscle mass in older patients. With regard to renal function, canine studies suggest that renal impairment may occur after 2 weeks of high-grade obstruction; thus, ureteral calculi should be treated as soon as possible. However, immediate intervention within one month may result in the impossibility of r-URS or UAS insertion. This study had some limitations. We studied a relatively small cohort, limited to patients with a single ureteral stone without pre-stenting or pre-nephrostomy. Second, SMART classification was decided based on the operator's subjective judgement. Almost all cases were evaluated by three investigators to make an accurate judgement. Finally, this was a preliminary study; therefore, external validation with a multicentre large prospective cohort will be needed.
In conclusion, we first defined EFs during URSL and used the SMART classification to predict surgical difficulty worldwide. Retention time of a ureteral stone and the presence of hydronephrosis were risk factors for MSA. There is a relationship between early intervention and patient age with DUT. URSL should be performed as soon as possible after 34 days following the symptom onset. However, for suspected stone impaction cases with more than 98 days after symptom onset, alternative procedures and detailed informed consent should be planned in advance.

Methods
Study population. This prospective study was performed at three academic centres in the SMART study group. The study was conducted in accordance with the Helsinki Declaration. This study was also approved by the institutional review boards at Kansai Medical University (T24-5) and Toyota Kosei Hospital (24ST-04) before the start of the study, and selected patients who consented to participation. The study flow chart is shown in Fig. 2. Among 832 patients who underwent URSL for kidney and ureteral stones registered in the SMART Study Group between January 2014 and February 2017, we analysed 185 patients who underwent URSL for single ureteral stones. We excluded all patients undergoing planned staged procedures; those with multiple calculi, a medical history of URS or URSL, or an unclear date of when the stone passage stopped; and those who underwent preoperative ureteral stent insertion (56 patients) or preoperative nephrostomy insertion (21 patients). All patients provided written informed consent prior to inclusion. The surgeons recommended patients to be treated as soon as possible. However, the treatment timing was decided according to the patient's desire, hospital capacity, and efficacy of medical expulsive therapy. As there were no reports analysing the association between the impacted stones and the time interval until surgery, we calculated the sample size based on our preliminary data. Our preliminary data showed that the ratio of grade 2 MSA of patients who were treated within 30 days and could not be treated within 60 days were 6.0% and 23.7% respectively. Calculated based on these data, 61 www.nature.com/scientificreports/ patients were required in each group with a significance level of 0.05% and a power of 80%. Totally, 183 patients were required, and in the present study we analysed the data from 185 patients.
Data collection. Table 1 shows the collected data. Hydronephrosis (grades 0/1/2/3) was defined according to the Ellenbogen classification 27 . The duration until intervention was calculated as days from the identification of clinical signs related to stones to the day of surgery. Preoperative evaluations included clinical examination and standard imaging (abdominal plain X-ray [KUB], ultrasound, and computed tomography) to determine the location and size of stones. Stone volume was estimated using the formula described by Tiselius and Andersson (length × width × 3.14 × 0.25) 28 . We usually perform preprocedural urine cultures, and we would change the antibiotics based on the results of urine cultures. Postoperative fever was defined as > 38.5 degrees Celsius. Ureteral injury was graded according to Traxer's classification 22 . SF was defined as not only endoscopic clearance at the end of the operation, but also as absence of stones or fragments > 2 mm on KUB and ultrasound one month after surgery. Surgical procedure. All URSL procedures were performed using a 6.4/7.8Fr (Olympus, Tokyo, Japan), 6.5/8.5Fr (Richard Wolf, Knittlingen, Germany), or 7.0/8.0Fr (Karl Storz, Tuttlingen, Germany) r-URS and fURS, including 7.5-F Flex-X2 (Karl Storz), 7.95-F URF-P6 (Olympus), or 8.5-F UFV-V2 (Olympus) under spinal or general anaesthesia.
At the beginning of the URSL, distal ureteral findings were observed using an r-URS. If the r-URS could not be inserted, the surgery was discontinued after inserting a ureteral stent. A UAS was placed in almost all patients except for those with distal ureteral stones. A ureteral access sheath of 9.5/11.5Fr, 10/12Fr, 11/13F, or 12/14 F sizes was placed in 113 patients (Table 3).
A holmium laser was used to fragment the target stone while securing a clear view of the endoscopic field using retrograde irrigation. Controlled irrigation with the single action pumping system (Boston Scientific, Natick, MA) was used for each case. Stone fragments were extracted until they completely disappeared using basket forceps and the one-surgeon basketing technique, as previously reported 29 . Selection of the type and size of device was based on the surgeon's discretion under the surveillance of members of the SMART study group (SH, SO or TI). At the end of each procedure, the UAS was removed, and ureteral injuries were visually assessed. A ureteral stent was inserted at the end of each surgery and left in situ, generally for 1-4 weeks.

EFs (SMART classifications).
We evaluated the EFs at the stone site (oedema, polyps, MSA) and DUT prospectively based on the SMART classification that was independently created for this study (Fig. 3). Grade 1 was defined as minor oedema that does not prevent seeing calculi with natural irrigation of about 100 cm saline. If retrograde pressure irrigation using a single action pumping system was necessary to clarify the endoscopic view and to fragment the calculi due to severe oedema, the oedema was graded as 2. Polyp grades were determined based on the presence or absence of polyps (grade 0 or 1). Grade 0 MSA classified stones that never stuck to the ureteral mucosa and easily moved up to the upper ureter with retrograde irrigation. G1 was defined as cases where stones easily separated from the ureteral mucosa; G2 cases were defined as those requiring removal using a URS or a laser fibre. Regarding DUT, if mild resistance to r-URS insertion was noted during observation of the lower ureter, the narrowing was graded as 1. G2 cases were defined as those involving strong resistance. www.nature.com/scientificreports/ In cases of difficulty in inserting r-URS, double guidewires were used to overcome these situations; however, a ureteral balloon catheter was not used in the present study. If r-URS could not be inserted due to severe DUT, the DUT was graded as 3. EFs were evaluated by three investigators (SH, SO or TI).
endpoints. The primary endpoint was incidence of EFs according to the periods until interventions. The secondary endpoint was to analyse an association between EFs and surgical outcomes including operative time, stone free rate, and ureteral injury. The additional endpoint was to analyse the predictive factors for prolonged surgical time and preoperative factors influencing EFs.
Statistical analysis. Normally distributed variables are expressed as means ± standard deviations. Categorical variables are expressed as numbers (percentages). Clinical variables were compared using the Mann-Whitney U-test, one-way analysis of variance, and the Chi-square test. Univariate and multivariate analyses were performed using a logistic regression model. The duration cutoff values predicting stone adhesion or DUT and their predictive accuracies were determined using ROC curve analysis. Differences were considered statistically significant when p < 0.05. All statistical analyses were performed using EZR for R 30 by two authors (SH and RA), then validated by the others.