Percutaneous Nephrolithotomy Versus Ureteroscopic Lithotomy for Large (>15 mm) Impacted Upper Ureteral Stones in Different Locations: Is the Upper Border of the Fourth Lumbar Vertebra a Good Indication for Choice of Management Method?

Introduction

I mpacted upper ureteral stones can destroy kidney function and cause life-threatening sepsis. Complete removal of the stone is vital to prevent stone growth and associated infection and to preserve renal function. Minimally invasive surgical modalities have resulted in high success rates and low morbidity, making traditional open surgery unnecessary. ^1,2 Extracorporeal shock wave lithotomy (ESWL) is noninvasive, can be performed as an outpatient procedure, and is associated with low morbidity rates, and has been recommended as first-line treatment for upper ureteral stones by the European Association of Urology 2009 urolithiasis guidelines. ³ However, ESWL has a very limited role in managing large (>15 mm) impacted upper ureteral stones, due to its low success rate, high retreatment rate, and increased risk of pyelonepritis. ^{4 –7}

Percutaneous nephrolithotomy (PCNL) and ureteroscopic lithotomy (URSL) have become the first-line treatments for large upper ureteral stones. Although the stone-free rate of PCNL is somewhat higher, so are rates of complications, such as severe hemorrhage. Although these treatment modalities have been compared in patients with large (>15 mm) impacted upper ureteral stones, the results have been inconsistent. These studies, however, compared stones throughout the entire upper ureter.

The upper ureter extends along 2–3 lumbar vertebrae. Between the third and fourth vertebrae, the ureter is more flexible, with frequent tortuosity and angulation, hampering passage of an endoscope. Stones closer to the pelvis are more likely to be inaccessible to URSL. Moreover, stones are more likely to migrate upward, increasing the rate of salvage procedures and decreasing the initial stone-free rate. In contrast, stones far from the pelvis may be inaccessible to PCNL. These findings suggested that the outcomes of URSL and PCNL may be affected by stone location. Categorization of stones by upper ureteral position is therefore, necessary to more precisely evaluate the outcomes of PCNL and URSL. We therefore, prospectively evaluated the efficacy and safety of these two treatment modalities in patients with stones higher and lower than the upper border of the fourth lumbar vertebra.

Materials and Methods

We analyzed 174 patients presenting with a solitary, impacted, upper ureteral stone, >15 mm in diameter, and treated at our institution between January 2007 and January 2010. All patients had a single, upper ureteral stone >15 mm in diameter, located between the superior margin of the sacroiliac joint and the ureteropelvic junction; moreover, intravenous urography (IVU) confirmed that the stone was impacted, with no visible contrast media below the calculus on any IVU image. Patients were excluded if they had renal insufficiency, as shown by a serum creatinine concentration >133 μM; a history of previous irradiation or pelvic surgery; morbid obesity; pregnancy; or coagulopathy. The advantages and complications of URSL and PCNL were explained to each patient preoperatively, as was the possibility of conversion to another therapeutic modality, and all patients provided written informed consent before inclusion. The procedure for each patient was chosen by that patient.

All patients were evaluated before treatment with history taking, clinical examination, ultrasonography, and IVU. Radiolucent stones were diagnosed by CT. Laboratory investigations included urinalysis and urine culture, measurement of serum creatinine concentration, and coagulation profile. Grade of hydronephrosis was categorized as none, mild, moderate, or advanced, based on the appearance of the pelvis on ultrasonography and the presence of calices and/or parenchymal atrophy. ⁸

All PCNLs were performed by one surgeon with experience in >500 patients and all URSLs were performed by another surgeon with experience in >500 patients. PCNL was performed under general anesthesia with the patient in the prone position. The puncture point was in the eleventh intercostal space or, in some patients, below the twelfth rib, between the scapular and posterior axillary lines. Under color Doppler ultrasound guidance, while avoiding blood vessels, the target calyx of the kidney was punctured and a guidewire was introduced through the puncture needle. The access tract was dilated along the guidewire using a series of sheath dilators or serial coaxial metal dilators until a 26F Amplatz sheath could be introduced. A 24F rigid nephroscope (Wolf) was introduced through the Amplatz sheath along the guidewire. If the proximal ureteral stone became visible but could not be accessed with the nephroscope, an 8.5F semirigid ureteroscope (Wolf) was used. The stone was fragmented with the pneumatic and ultrasound LithoClast Master (EMS), and the stone fragments were removed. A double-J catheter (Bardia) was inserted and left in place for 1 week, and a nephrostomy tube was left in place for 2–3 days.

URSL was performed under epidural anesthesia with the patient in the lithotomy position, using a 7.5F or 8.5F semirigid ureteroscope (Wolf). A ureteral guidewire was introduced, and the ureteroscope was inserted along the guidewire. Generally, the ureteral orifice or ureter was not dilated unless it was too narrow for passage of the ureteroscope. If a semirigid ureteroscope could not be passed through a tortuous ureter, a 7.5F flexible ureteroscope (Karl Storz) was used. Stones were fragmented by Ho:YAG laser lithotripsy (Powersuite100W plus; Coherent, Inc.). A double-J catheter (Bardia) was inserted and left in place for 3–5 days. If excessive manipulation had occurred or intraluminal mucosa edema, ureteral polyps, ureteral strictures, perforation or injury had developed, a double-J stent was inserted for 2–4 weeks. Conversion to another treatment modality during URSL or PCNL was deemed treatment failure.

All patients underwent plain abdominal radiography 3 days after therapy to detect the presence of residual stones and D-J stents. Stone-free status was assessed 1 month later by plain abdominal X-ray film and/or IVU, with radiolucent stones assessed by CT. Hydronephrosis was evaluated by renal ultrasonography and IVU. Treatment success was defined as the absence of residual stones. All patients were followed-up regularly every 6 months for 18 months. The efficiency quotient (EQ) was calculated for the two groups, using the formula: EQ=[percentage stone-free/(100%+percentage requiring retreatment+percentage requiring an auxiliary procedure) ⁹ .

All results are reported as mean±standard deviation. The data were analyzed using SPSS 13.6. All parameters were analyzed statistically using unpaired Student's t tests and χ ² tests. A p value <0.05 was considered statistically significant.

Results

Of the 174 patients included in this study, 83 (47.7%) underwent PCNL and 91 (52.3%) underwent URSL as the primary treatment approach. Mean patient age, gender, stone size, and stone location were similar in the two groups (Table 1). Mean operation time (108.76±19.36 minutes vs. 63.56±16.38 minutes, p<0.05) and mean postoperative hospital stay (5.36±1.98 days vs. 2.49±1.23 days, p<0.05) were significantly longer in the PCNL than in the URSL group. There were three failures in the PCNL group (3.6%), two (2.4%) due to aborted punctures, and one (1.2%) due an inability to place the D-J stent in the antegrade position after stone fragmentation; all 3 patients underwent URSL as a salvage procedure and were therefore, excluded from subsequent evaluations. Six patients in the PCNL group (7.2%) required use of a percutaneous antegrade ureteroscope (PAU). Two patients in the URSL group (2.2%) were converted to the open ureterolithotomy, due to an inability to reach the stone resulting from the tortuosity and angulation of the ureter. In 24 patients (26.4%), the stones migrated upward during surgery; with stone migration significantly correlated with stone location. The retropulsion rates for stones higher and lower than the fourth lumbar vertebra were 47.5% (19/40) and 9.8% (5/51), respectively (p<0.05). Only 7 of 91 patients required secondary treatment with a flexible URS after failure to access the stone by semirigid URS, with stone retropulsion occurring in 4. Five of the 19 patients with upwardly migrating stones located higher than the fourth lumbar vertebra were converted to PCNL during the operation. Only one patient with a stone located lower than the fourth lumbar verterbra was converted to PCNL. In all six, the stones were >10 mm in diameter and could not be discharged spontaneously. Fourteen patients (15.4%) had residual stone fragments after 1 month and 13 patients underwent ESWL as the first salvage procedure, with 3 patients later undergoing URSL after failure of ESWL. The remaining patient again underwent URSL without EWSL. The 8 patients converted to other operative procedures, such as PCNL or open ureterolithotomy, were deemed to have failed URSL and were therefore, excluded from subsequent analyses. The overall stone-free rates at 1 month in the PCNL and URSL groups were 96.4% (80/83) and 75.8% (69/91), respectively.

We observed a statistically significant positive correlation between stone-free rate and stone position. URSL resulted in a significantly lower stone-free rate than PCNL for stones higher (57.5% vs. 97.8%, p<0.05), but not lower (90.2% vs. 94.7%, p=0.431) than the fourth lumbar vertebra. The total EQs for the PCNL and URSL groups were 0.93 and 0.59, respectively (p<0.05). EQs differed slightly in the PCNL group for high and low stone locations (0.96 vs. 0.90), but differed significantly in the URSL group for high and low stone locations (0.40 vs. 0.82, p<0.05).

Complications of both procedures included blood loss requiring transfusion, transient postoperative fever, flank pain, hematuria, and urteral perforation. In the PCNL group, 9 patients experienced transient postoperative fever higher than 38.5°C, which was controlled with intravenous antibiotics appropriate to the results of middle stream urine culture sensitivity and supportive treatment. Hematuria was observed in 14 patients; of these, 3 had massive blood loss requiring transfusion. One of these had a persistent decrease in hemoglobin levels and underwent transarterial embolization of a traumatic aneurysm. In the URSL group, 8 patients had urinary tract perforation, including six who underwent URSL with semirigid and two with flexible URS, and were treated by D-J stent placement. Eight patients had transient fever, with two requiring a change in antibiotics based on middle stream urine culture sensitivity. Nine patients had gross hematuria, but recovered without special therapy after 1–3 days.

At 1 year follow-up, stenosis of the ureter was observed in 2 patients. One was due to an intraluminal stricture, resulting from perforation of the ureter during surgery. Six months later, the patient presented with hydronephrosis and underwent balloon ureteral dilation (Balloon ureteral dilator and sets; Cook Urological, Inc.). A D-J stent was inserted for 6 months, but the stricture and nephrosis still persisted. Stenosis of the ureter in the second patient was due to postoperative fibrosis around the ureter, with no intraoperative ureteral perforation. Both patients underwent secondary open surgery.

Details are shown in Table 2. Comparations of stone compositions and modalities were shown in Tables 3 and 4.

Discussion

Long-term impacted upper ureteral stones may interrupt urinary flow or cause hydroureteronephrosis or progressive back pressure on the kidney. Increased backflow resulting from intrapelvic pressure may ultimately lead to significant cortical atrophy and impairment of renal function. ¹⁰ The development of more advanced instruments and techniques for treating proximal ureteral stones has led to the gradual replacement of open surgery by minimally invasive procedures. ¹¹ ESWL is the method most frequently used for upper ureteral stones due to its advantages, including the outpatient nature of the procedure, the minimal requirements for anesthesia, and low complication rates. This procedure, however, has disadvantages in treating patients with large (>15 mm) impacted upper ureteral stones, including a low success rate, a high re-treatment rate, and an increased risk of obstructive pyelonephritis. For large stones (>10 mm), ESWL has been reported to result in stone-free rates of 50%–79%, ^4,5,7 with retreatment rates as high as 65%. ⁶ Due to improvements in fiberoptics and lithotripters, endoscopic lithotripsis, URSL and PCNL are now regarded as the primary management modalities in patients with large ureteral stones.

Over the past decade, URSL has achieved a high success rate in managing distal ureteral stones. ¹² In the treatment of upper ureteral stones, URSL has resulted in a 3-month overall stone-free rate of 82%. ¹³ When combined with different kinds of lithotripters, the stone-free rate has been reported to range from 35% to 87% for proximal ureteral stones >15 mm in diameter. ^{14 –16} We found that the overall stone-free rate was significantly lower for URSL than for PCNL (75.8% vs. 96.4%). Moreover, the stone-free rate was correlated with stone location, not with stone size. When we separated patients with stones higher and lower than the upper border of the fourth lumbar vertebra, we found that the 1 month stone free rates of patients treated with URSL and PCNL differed significantly for stones above (57.5% vs. 97.8%), but not below (90.2% vs. 94.7%). The lower stone-free rate of URSL for stones higher than the upper border of the fourth lumbar vertebra may be due to the retrograde movement and continuous high-pressure irrigation required for a clear operative visual field. A previous study reported that the total stone-retropulsion rate of upper urerteral stones was 25.8%, ¹⁰ similar to our result (26.4%). Most previous studies comparing URSL and PCNL did not consider stone position. The upper ureter can extend about 2–3 lumbar vertebrae, making stones nearer the pelvis more likely to migrate upward. Indeed, we found that the upward migration rates of stones higher and lower than the upper border of the fourth lumbar vertebra were 47.5% and 9.8%, respectively.

Stone retropulsion has been reported during the disintegration of upper ureteral stones using the holmium:yttrium-aluminum-garnet (Ho:YAG laser). ¹⁷ Although we thought that application of the Ho:YAG laser might reduce the stone migration rate, we found that use of this laser was associated with a lower rate of upward stone migration only for stones located below the upper border of the fourth lumbar vertebra.

We also found that the relationship between success rate and position was applicable to stones of different composition. Although the success rate of URSL differed for stones of different compositions, the differences in sample size prevented a determination of whether stone composition affected the success rate of URSL. New devices, such as the Stone Cone and N trap have been reported to prevent upward stone migration during lithotripsy. ¹⁸ However, these new devices are not yet available to most institutions in developing countries like China, and the serious morbidities associated with these devices, such as ureteral avulsion, have prevented their widespread use in our country. ¹⁹

The upper ureter is relatively flexible, especially between the third and fourth lumbar vertebrae, with tortuosity and angulation resulting from the weight of the stone. In these patients, the stone may not be accessible by URSL and may migrate upward; thus, requiring auxiliary procedures and retreatments. In most cases, ESWL was the first ancillary choice; however, for retropulsive stones >10 mm in diameter, we recommend a single session of PCNL.

PCNL is an alternative, minimally invasive modality used in the treatment of upper ureteral stones when ESWL is not indicated or has failed, and when the upper urinary tract is not amenable to retrograde URSL, and has been recommended as first-line treatment by the EAU2012 urolithiasis guideline. ²⁰ The combination of ESM pneumatics and ultrasound lithotripsy with the holmium laser has resulted in more effective stone disintegration and a higher stone free rate than URSL. In treating large (>15 mm) stones, PCNL has been reported to result in stone-free rates of 86% to 98.5%, ^1,10,19 similar to our finding of 96.4%. However, we found, that the success rate of PCNL was not related to stone size, position, or compositions. Since stones impacted in the ureteral mucosa interrupt urinary flow, most patients developed hydroureteronephosis, allowing easier and safer puncture, with more space for the nephroscope. The nephroscope can easily access stones higher than the upper border of the fourth lumbar vertebra and can also access the fifth lumbar vertebra in patients with a high degree of hydroureteronephosis. Access to stones located below the upper border of the fourth lumbar vertebra is more difficult, especially in patients with slight ureterodilation. For these patients, percutaneous antegrade ureteroscopy (PAU) through the nephroscope sheath may be a good alternative modality, since the semirigid ureteroscope can reach the entire upper ureter. ²¹ PAU has been reported to have the same stone free rate as PCNL. ^21,22 Moreover, routine use of the ureteroscope for percutaneous access may reduce bleeding. ^23,24 The nephroscope is safe and effective in most patients. ²² Thus, routine use of an ureteroscope is not necessary for most patients. Of our patients, only six required a ureteroscope, with most other stones successfully removed with a nephroscope with slight bleeding.

We observed comparable overall complication rates in our URSL and PCNL groups. Moreover, complications were not related to stone location. In the URSL group, the most frequent common complications were fever and hematuria, although the rates were low. In contrast, ureteral avulsion, a more serious complication, was not observed. Ureteral perforation rates have been reported to be 2%–25%, ^25,26 consistent with our results. Ureteral perforation mainly resulted from injuries caused by the Ho:YAG laser. Although a flexible URS could pass more easily through tortuous and angulated ureters, use of flexible URS did not reduce perforation rates. We found that the perforation rate using a semirigid URS was 6.8% (5/76), whereas that using a flexible URS was 28.5% (2/7). Due to the small number of patients in the latter group, the difference was not statistically significant. As placement of a D-J stent was routine in our study, all perforations recovered spontaneously. The rate of strictures, a serious long-term complication of URSL, has been reported to be 0%–3%. ²⁷ Perforation and avulsion of the ureteral mucosa are important factors associated with intraluminal stricture. We found that one patient with a ureteral perforation developed a stricture. We found that only one patient experienced an extraluminal stricture, 1 year after the operation. This stricture was discovered when the patient underwent an open operation for ureteral stenosis, at which time fibrosis was observed around the ureter. Fibrosis may be caused by extravasation of water resulting from high pressure irrigation.

The most important complication of PCNL is hemorrhage. Most patients had some postoperative bleeding, which was controlled by stoma occlusion and usage of hemostatics. Previous studies have reported serious bleeding rates of 0.5% to 3%, ²⁸ and rates of heavy bleeding requiring transfusion of about 2%–5%. ²⁹ We found that PCNL was safe in avoiding bleeding in patients with stones above the upper border of the fourth lumbar vertebra and heavy hydroureteronephosis. For stones far from the pelvis or in patients with no or mild hydroureteronephosis, the puncture position should be chosen carefully to avoid bleeding. If the angle between the calyx and ureter is <90°, a nephroscopic sheath via the calyx is more likely to result in hemorrhage secondary to the rupture of the calyceal neck. Choosing a puncture position such that the angle between the calyx and ureter is >90° is important in avoiding hemorrhage.

Prophylactic administration of appropriate intravenous antibiotics has been reported to result in a post-PCNL fever rate of 2.8%–32.1%, with most patients successfully managed conservatively. ^{30 –33} Although fever was more serious in the PCNL than in the URSL group, few patients progressed to sepsis.

EQ, a measure of the need for retreatment and auxiliary procedures, can be used to compare the results of different treatment modalities. An ideal treatment has an EQ of 1, whereas insufficient treatments have an EQ <0.5. We found that the overall EQ was higher for PCNL than for URSL, due to a lower requirement of the former for another treatment or auxiliary procedure. EQ was also associated with stone position in the URSL, being 0.4 and 0.82 for stones higher and lower than the upper border of the fourth lumbar vertebra, respectively.

Our study had several limitations, including the small number of patients and the relatively short follow-up time.