Percutaneous Nephrolithotomy in Horseshoe Kidneys: Results of a Multicentric Study

Abstract

Objective:

To report the outcomes of percutaneous nephrolithotomy (PCNL) in horseshoe kidneys (HSK) in 12 institutions worldwide and evaluate the impact of patient position during operation.

Methods:

We carried out a retrospective analysis of PCNL procedures performed between 2008 and 2018 in patients with HSK. Pre-, peri-, and postoperative data were collected, and a subgroup analysis was performed according to patient position. Success was defined as an absence of >4-mm fragments. Values of p < 0.05 were considered significant.

Results:

We analyzed 106 procedures. The transfusion, complication, and immediate success rates (ISRs) were 3.8%, 17.5%, and 54.7%, respectively. The final success rate (FSR) increased to 72.4% after a mean of 0.24 secondary procedures. Logistic regression showed that higher body mass index (BMI) and stone size were significantly associated with residual fragments ≥4 mm. Sixty-seven patients (63.2%) were treated in prone and 39 (36.8%) in supine position. The prone group had a significantly higher BMI than the supine group (30.1 vs 27.7, p = 0.024). The transfusion, complication, and ISRs between the prone and supine groups were 4.5% vs 2.6% (p = 0.99), 16.9% vs 18.4% (p = 0.99), and 52.5% vs 69.2% (p = 0.151), respectively. Surgical time was significantly longer in the prone group (126.5 vs 100 minutes, p = 0.04). Upper pole was the preferred access in 80.3% of the prone group and 43.6% of the supine group (p < 0.001). The prone group had significantly more Clavien 2 complications than the supine (p = 0.013). The FSR in the prone and supine groups increased to 66.1% and 82.1% after 0.26 and 0.21 secondary procedures, respectively. No complications higher than Clavien 3 occurred.

Conclusion:

PCNL in patients with HSK is safe and effective with a low complication rate. Higher BMI and stone size negatively impacted outcomes. Supine PCNL may be an option for treating kidney stones in patients with HSK.

Introduction

Horseshoe kidney (HSK) is the most common renal fusion anomaly, occurring in ∼1 in 600 births.¹ It is characterized by fusion of the kidneys at the lower poles, with consequent failure to ascend and malrotation of the kidneys. Usually, high insertion and lateralization of the ureter occur, causing a drainage deficit from the collector system and likely leading to the high incidence of kidney stones¹ among adults with HSK (35%).² Metabolic abnormalities may also contribute to stone formation in patients with HSK,³ and the abnormal anatomy makes stone treatment difficult. All available treatments for kidney stones can be used in HSK cases,⁴ but success rates are usually lower than in kidneys with regular anatomy, especially with extracorporeal shockwave lithotripsy (SWL).⁵

The guidelines of the American and European Urological Associations are often used to determine kidney stone treatment, including those with HSK. To treat stones larger than 2 cm, the first option is percutaneous nephrolithotomy (PCNL).^6,7 In patients with HSK, this is usually performed in the prone position, allowing entry through the upper pole and providing good access to the collecting system. However, in patients with normal kidney anatomy, the supine position is reliable and safe in most cases,⁸ but it is unknown whether the supine position is adequate in patients with HSK. Access through the upper pole is not usually preferred during supine PCNL,⁹ and some researchers fear it could compromise outcomes.

Several case series have investigated prone PCNL in patients with HSK,^10,11 but only one has addressed the supine position, to our knowledge, and it had no specific outcome analysis.¹² Another single case report described supine PCNL in a patient with HSK.¹³

This study aimed to report the outcomes of PCNL in patients with HSK, and to evaluate the impact of patient position on outcomes. Because this procedure is rare, we designed a multicentric, collaborative study to ensure enough cases.

Materials and Methods

We carried out a multicentric, retrospective analysis of procedures performed between 2008 and 2018 that investigated PCNL in patients with HSK and provided complete data. Through email, the main author (F.C.V.) invited collaboration from colleagues around the world, focusing on those routinely using supine PCNL and noncontrast CT scan as the preferred pre- and postoperative imaging method. The authors completed an Excel datasheet for each patient and sent it back to the main author for analysis. Patients with HSK ≥18 years of age who had undergone PCNL in the past 10 years were included. Patients with incomplete postoperative data were excluded. After obtaining institutional review board approval in each institution, we evaluated 106 PCNLs performed at 12 medical centers worldwide. Each procedure was evaluated as an independent case, even if the same patient had undergone multiple surgeries. In all patients, we recorded sex, age, body mass index (BMI), and previous interventions for stones. The researchers at each center evaluated the stones by size, density, and laterality, and retrospectively classified them according the Guy's Stone Score.¹⁴ The preoperative imaging modality was also recorded. The study was retrospective, so practices were not standardized across centers.

Operations were performed according to local preference. All cases from the Cleveland Clinic (28 cases) used the prone, split-leg position, whereas all those from Brigadeiro Hospital (7 cases), Ankara University School of Medicine (9 cases), Tel-Aviv Sourasky Medical Center (8 cases), Cottolengo Hospital (5 cases), Fundació Puigvert de la Universitat Autònoma de Barcelona (2 cases), and Ospedale S Croce e S. Carlo in Cuneo (2 cases) used the supine position. Marmara University School of Medicine (8 cases), Massachusetts General Hospital of Harvard Medical School (8 cases), UCSF School of Medicine (6 cases), and Hospital Regional de Alta Especialidad del Bajío (4 cases) used both positions, with prone being used initially, followed by supine after the surgeons had learnt the position.

All procedures started with retrograde pyelography using an open-ended ureteral catheter, and puncture was guided by fluoroscopy. Ultrasound was regularly used to check the puncture window. The calix for puncture and dilation technique were chosen by the surgeon and recorded. All institutions were teaching hospitals with residents, so all procedures were performed by senior staff in conjunction with a resident. Flexible instruments were used regularly and in accordance with visualization conditions at the end of the procedure. Only Cleveland Clinic regularly performed retrograde endoscopic to assist puncture.

Pre-, peri-, and postoperative data were collected using the datasheet, and a subgroup analysis according to patient position was performed. Hemoglobin variation was evaluated between the most recent preoperative period and the first postoperative day. Surgical time was measured from the beginning of cystoscopy until final kidney drainage.

Stone-free rate (SFR) was defined as the absence of any fragments, whereas immediate success rate (ISR) was defined as the absence of >4-mm fragments at the first control study, based on either noncontrast CT or ultrasound, depending on local protocols. Final success rate (FSR) was defined as the absence of >4-mm fragments at the control study after any auxiliary treatments (SWL, retrograde intrarenal surgery [RIRS], PCNL, etc.). Complications were graduated according the Clavien–Dindo scale during the 30-day postoperative period.

Continuous variables were described as mean and standard deviation or median and interquartile range. Categorical variables were described as absolute or relative frequencies. We compared parameters between the prone and supine positions using the Student's t-test or Mann–Whitney test for continuous variables, and the chi-squared or Fisher's tests for categorical variables.

We performed logistic regression to evaluate success factors. The effect of position on immediate or final relative success was evaluated using a logistic model. Models were adjusted for important variables identified in the initial analysis. Analyses were performed using R 3.5.1 software (R Core Team, 2018), with a 5% significance level.

Results

We analyzed 106 surgeries in 81 patients. Table 1 provides demographic data, stone characteristics, and surgical parameters. The transfusion rate, complication rate, and ISR were 3.8%, 17.5%, and 54.7%, respectively. The FSR increased to 72.4% after a mean of 0.24 secondary procedures (Table 2). The time of final analysis varied among institutions. CT was used in 55.6% of cases. No mini-percutaneous access (tract dilation ≤24F) was performed.

Table 1.

Patient Demographic Data and Stone Characteristics

	Total	Prone	Supine	p
N	106	67	39
Age, mean (SD)	50.5 (13.3)	51 (14.4)	49.5 (11.2)	0.580
Sex, male % (n)	80.2 (85)	76.1 (51)	87.2 (34)	0.211
ASA score, median (min–max)	2 (1–3)	2 (1–3)	2 (1–3)	0.093
Right side, %	51.9	52.2	51.3	0.99
BMI, mean (SD)	29.2 (5)	30.1 (5.5)	27.7 (3.5)	0.024
Previous SWL rate, %	36.8	41.8	28.2	0.085
Previous PCNL rate, %	32.4	33.3	30.8	0.99
Preoperative CT rate, %	91.8	88.3	97.4	0.20
Stone size, mm, median (min–max)	25 (19–35)	23 (18–35)	27 (21–35)	0.367
Stone density, HU, median (min–max)	1100 (836–1321)	1075 (710–1334)	1200 (1000–1300)	0.21
Guy's stone score, median (min–max)	3 (2–4)	3 (2–4)	3 (2–4)	0.062
1	0	0	0
2, %	34	41.8	20.5
3, %	60.4	52.2	74.4
4, %	6	6	5.1

Bold value indicates significant result.

BMI = body mass index; HU = Hounsfield units; PCNL = percutaneous nephrolithotomy; SD = standard deviation; SWL = extracorporeal shockwave lithotripsy.

Table 2.

Estimated Coefficients for Immediate Success Rate

Variables	Odds ratio (95% CI)	p
Supine position	2.312 (0.765–7.611)	0.148
BMI	0.878 (0.789–0.966)	0.011
Stone size	0.964 (0.928–0.995)	0.038
Sex, male	0.689 (0.180–2.344)	0.563
Surgical time	0.996 (0.988–1.005)	0.408

Bold value indicates significant result.

CI = confidence interval.

Logistic regression showed that BMI (odds ratio [OR] = 0.878, 95% confidence interval [CI] = 0.789, 0.966, p = 0.011) and stone size (OR = 0.964, 95% CI = 0.928, 0.995, p = 0.038) were significantly associated with residual stones ≥4 mm (Table 2) and FSR (Table 3).

Table 3.

Estimated Coefficients of Relative Final Success Rate

Variables	Odds ratio (95% CI)	p
Supine position	3.212 (0.907–13.692)	0.086
BMI	0.899 (0.810–0.991)	0.036
Sex, male	0.669 (0.132–2.603)	0.586
Stone size	0.961 (0.923–0.993)	0.030
Surgical time	1.000 (0.991–1.009)	0.949

Bold value indicates significant result.

When analyzing procedures according to patient position, 67 patients (63.2%) were placed in prone and 39 (36.8%) in supine. Prone patients had a significantly higher BMI than supine patients (30.1 vs 27.7, p = 0.024). Table 4 summarizes the operative parameters and outcomes. The upper pole was the preferred access in 80.3% of the prone group and 43.6% of the supine group (p < 0.001). The transfusion rates, complication rates, and ISRs in the prone and supine groups were 4.5% vs 2.6% (p = 0.99), 16.9% vs 18.4% (p = 0.99), and 52.5% vs 69.2% (p = 0.151), respectively. Surgical time was significantly longer in the prone group than in the supine group (126.5 minutes [85–178.8 minutes] vs 100 minutes [75–139 minutes], p = 0.04). The prone group had significantly more Clavien 2 complications than the supine group (p = 0.013); no complications higher than Clavien 3 occurred in this series (Table 4). The FSRs in the prone and supine groups increased to 66.1% and 82.1% after 0.26 and 0.21 secondary procedures, respectively (Table 4).

Table 4.

Operative Parameters and Outcomes

	Total	Prone	Supine	p
N	106	67	39
Surgical time, minutes, median (min–max)	120 (80–167)	126.5 (85–178)	100 (75–139)	0.040
Fluoroscopy time, minutes, median (min–max)	3 (1.2–6)	4 (3–6)	1.9 (1–10.2)	0.120
Number of punctures, median (min–max)	1 (1–2)	1 (1–2)	1 (1–2)	0.141
1, %	92.5	95.5	87.2
2, %	7.5	4.5	12.8
Calix of puncture, %				<0.001
Upper calix	66.7	80.3	43.6
Middle calix	28.6	18.2	46.2
Lower calyx	4.8	1.5	10.3
Dilation technique, %				0.002
Serial	52.8	40.3	74.4
Balloon	47.2	59.7	25.6
Nephrostomy rate, yes, %	59.8	62.1	55.6	0.664
Double-J stent rate, yes, %	65	60.3	73	0.756
Lithotripter, %				0.001
Ultrasonic	71.8	84.4	51.3
Ballistic/laser	28.2	15.6	48.7
Hb variation, mg/dL, mean (SD)	1.5 (0.9)	1.5 (1)	1.6 (0.8)	0.575
Transfusion rate, %	3.8	4.5	2.6	0.999
Complication rate, %	17.5	16.9	18.4	0.999
Clavien 1		16.7	85.7	0.013
Clavien 2		75	14.3
Clavien 3		8.3	0
Clavien 4/5	0	0	0
CT as last image modality, %	55.6	55.9	66.7	0.520
Large residual stone POD, mm, median (min–max)	0 (0–35)	3 (0–35)	0 (0–25)	0.071
Length of hospital stay, days, median (min–max)	2 (1–4)	2 (1–13)	2 (1–13)	0.932
Immediate stone-free rate (0 fragment), %	50.0	45.8	59	0.283
Immediate success rate, ≤4 mm, %	59.2	52.5	69.2	0.148
Auxiliary procedures, %	24.5	26.6	21.1	0.637
SWL, %/RIRS, %	28/20	29.4/11.8	25/37.5	0.576
PCNL, %/pyelolithotomy, %	48/4	52.9/5.9	37.5/0
Final success rate, ≤4 mm, %	72.4	66.1	82.1	0.086

Bold values indicate significant result.

POD = postoperative day; RIRS = retrograde intrarenal surgery.

Discussion

PCNL is commonly used to treat kidney stones >2 cm in diameter, or when SWL or RIRS fail or are contraindicated,⁶ even in special situations such as HSK or pelvic kidneys. PCNL is usually performed in prone, so descriptions of supine PCNL in patients with HSK are scarce,^12,13 and no studies have described outcomes in such cases, to our knowledge. To investigate PCNL and the impact of position on outcomes in a large enough number of patients with HSK, we performed a multicentric study of high-volume centers worldwide. Ultimately, 106 cases were analyzed, showing that this is an infrequent procedure. The PCNL outcomes in patients with HSK were similar to those in patients with normal kidney anatomy, and the supine position has been used by urologists with experience of this technique to perform PCNL in patients with HSK.

Our ISR was 54.7%, increasing to 72.4% after a mean of 0.24 secondary procedures, and our median Guy's stone score was 3, comparable with a non-HSK Guy's 3 case series.⁹ Skolarikos and colleagues¹⁵ reported a success rate of 65.5% for PCNL in patients with HSK, whereas Satav and colleagues¹⁰ reported 87.5% success rate; both studies used plain X-ray of the kidneys, ureters, and bladder, and cases were not classified using Guy's stone score. Conversely, noncontrast CT was used in 55.7% of our cases, leading to higher residual stone detection rate and possibly an inferior recorded success rate.

In this study, the transfusion and complication rates were 3.8% and 17.5% respectively, comparable with other series.^10,16 No complications higher than Clavien 3 occurred, despite the complexity of the cases.

In our multivariate analysis, BMI and stone size were inversely proportional to the success rates. Large stones are associated with lower success rates, as seen in other PCNL series.¹⁵ However, BMI is not usually associated with worse outcomes.^17,18 A recent meta-analysis of 7 studies and >4500 patients revealed no differences in success or complications among normal weight, obese, and super obese individuals.¹⁸ In the case of HSK, this has not been studied, but we suppose that the kidney is located more medially in the abdomen in these patients, those with higher BMI may have a longer tract from skin to stone. As a result, more torque and longer instruments would be necessary, and it would be more difficult to manipulate the nephroscope. Skolarikos and colleagues¹⁵ used logistic regression analysis to evaluate prognostic factors of PCNL in patients with HSK, revealing that staghorn calculi were the only significant predictor of SFR.

Most urologists believe that PCNL should be performed in the prone position in patients with HSK, and that access should be gained through the upper pole, which is usually subcostal and offers a straight pathway to most calices. The outcomes are good and the technique is well established. However, supine position for PCNL is gaining popularity worldwide, even in the United States, where it used to be rare.¹⁹ Using this position, turning the patient prone is unnecessary; therefore, the total operative time may be reduced.⁸ This position has even been used for complex cases, showing similar efficacy.²⁰ One group used to performing PCNL in the supine position, when faced with HSK, probably tended toward using that position. However, until now, little information was available supporting this choice. Thus, this study is pioneer, showing that supine PCNL is performed in several high-volume centers worldwide, and that its outcomes are similar to prone PCNL.

The supine position can be used in simple cases and complex cases. One recent meta-analysis by Li and colleagues⁸ studied 15 randomized controlled trials comparing safety and efficacy between the supine and prone positions for PCNL in patients with renal or upper ureteral calculi. No differences occurred between the groups in terms of success, hospital stay, or complication rate. However, the supine group had a significantly shorter operation and a lower rate of fever. Vicentini and colleagues²⁰ reported that complex cases treated using supine PCNL had fewer infectious complications (Guy's 3 and 4) than those treated using prone PCNL (7.8% vs 2.1%; p = 0.042), perhaps because the Amplatz sheath is in the horizontal or downward direction in the supine position, so irrigation outflow is under low pressure, lowering intrarenal pressure during the procedure, preventing pressure peaks, and reducing the chance of pyelovenous backflow of contaminated liquid. However, this mechanism has not been investigated. Moreover, the lower operation time during supine PCNL may lead to fewer complications. In this study, complication subtypes were not specifically addressed, but operative time was significantly lower in supine than in prone cases, which stands to reason, because patients in the supine position need not be turned to the prone split-leg position. Had all operative times been collected, the time from puncture until the end of the procedure may have been similar, but we do not have these data.

Differences occurred regarding dilation technique and energy for stone fragmentation (Table 4). Balloon dilation with ultrasonic lithotripter tends to be faster than sequential dilation with ballistic lithotripter, but this difference did not impact outcomes in this study, perhaps because we investigated too few cases to register an effect.

Regarding the punctured calix findings in this study, 20% of prone cases were accessed through the middle or lower calix, although current recommendations advocate accessing the kidneys through the upper pole. The kidneys may have been too high in HSK cases, increasing the risk of thoracic lesions. Furthermore, most supine cases were performed using middle calix (46.2%) or lower pole access (10.3%), without any negative effect on outcomes. The ideal and most common (80%) puncture site for supine PCNL is the inferior calix, because it allows upper calix access in most cases.²¹ However, in this study, only 10% of the supine punctures were performed on the lower pole, indicating that this is not the best calix for supine PCNL in HSK cases, probably because it has a medial and anterior position. The upper pole, which is only used rarely in patients with normal kidney anatomy, was the entry point in >40% of supine cases in this study, whereas the middle calix was the ideal entry site in most cases, perhaps because it conferred a shorter tract to the stones in the supine position.

These results show that urologists use their own judgment, based on the stone location, to determine the kidney entry point during supine PCNL in patients with HSK.

This study had several limitations. It was retrospective and therefore lacked randomization, limiting its power and evidence level and preventing firm conclusions. However, some positive points must be noted. Postoperative evaluation methods varied among institutions, but CT was the most common imaging technique for pre- and postoperative control. We analyzed the outcomes of high-volume centers worldwide using prospective electronic databases, so our results reflect real-world practice. During our search for supine cases, we contacted centers with experience using this position, which may have limited the number of prone cases.

This study included a significant number of cases, allowing us to evaluate PCNL outcomes in patients with HSK. Moreover, it was the first comparative study of patient position, and it will edify urologists who already perform supine PCNL in patients with HSK, but who were unsure whether their practice was correct as it goes against formal recommendations. Until a multicentric, randomized study is performed, our article offers some evidence that supine PCNL is safe and effective to treat patients with large stones and HSK.

Conclusions

In conclusion, PCNL in patients with HSK is a rare but effective treatment, with a low complication rate, similar to PCNL for complex cases in regular kidneys. Higher BMI and stone size seem to worsen outcomes of PCNL in patients with HSK. Supine PCNL may be a suitable option for patients with HSK and kidney stones.

Footnotes

Acknowledgment

The authors thank Lucas Petri Damiani for statistical analysis.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This article had no funding support.

Abbreviations Used

References

Weizer

, Silverstein

, Auge

, et al. Determining the incidence of horseshoe kidney from radiographic data at a single institution. J Urol, 2003; 170:1722–1726.

Pawar

, Thongprayoon

, Cheungpasitporn

, Sakhuja

, Mao

, Erickson

. Incidence and characteristics of kidney stones in patients with horseshoe kidney: A systematic review and meta-analysis. Urol Ann, 2018; 10:87–93.

Raj

, Auge

, Assimos

, Preminger

. Metabolic abnormalities associated with renal calculi in patients with horseshoe kidneys. J Endourol, 2004; 18:157–161.

Kartal

, Cakici

, Selmi

, Sari

, Ozdemir

, Ersoy

. Retrograde intrarenal surgery and percutaneous nephrolithotomy for the treatment of stones in horseshoe kidney; what are the advantages and disadvantages compared to each other?. Cent European J Urol, 2019; 72:156–162.

Gokce

, Tokatli

, Suer

, Hajiyev

, Akinci

, Esen

. Comparison of shock wave lithotripsy (SWL) and retrograde intrarenal surgery (RIRS) for treatment of stone disease in horseshoe kidney patients. Int Braz J Urol, 2016; 42:96–100.

Türk

, Petřík

, Sarica

, et al. EAU guidelines on diagnosis and conservative management of urolithiasis. Eur Urol, 2016; 69:468–474.

Assimos

, Krambeck

, Miller

, et al. Surgical management of stones: American Urological Association/Endourological Society Guideline, Part I. J Urol, 2016; 196:1153–1160.

, Gao

, Li

, Zhang

, Jiang

. Supine versus prone position for percutaneous nephrolithotripsy: A meta-analysis of randomized controlled trials. Int J Surg, 2019; 66:62–71.

de Souza Melo

, Vicentini

, Beraldi

, Hisano

, Murta

, de Almeida Claro

. Outcomes of more than 1 000 percutaneous nephrolithotomies and validation of Guy's stone score. BJU Int, 2018; 121:640–646.

10.

Satav

, Sabale

, Pramanik

, Kanklia

, Mhaske

. Percutaneous nephrolithotomy of horseshoe kidney: Our institutional experience. Urol Ann, 2018; 10:258–262.

11.

Etemadian

, Maghsoudi

, Abdollahpour

, Amjadi

. Percutaneous nephrolithotomy in horseshoe kidney: Our 5-year experience. Urol J, 2013; 10:856–860.

12.

Osther

, Razvi

, Liatsikos

, et al. Percutaneous nephrolithotomy among patients with renal anomalies: Patient characteristics and outcomes; a subgroup analysis of the Clinical Research Office of the Endourological Society global percutaneous nephrolithotomy study. J Endourol, 2011; 25:1627–1632.

13.

Sohail

N AA

, Abdelrahman

, Bhatti

. Supine percutaneous nephrolithotomy in horseshoe kidney. J Taibah Univ Med Sci, 2017; 12:261–264.

14.

Vicentini

, Marchini

, Mazzucchi

, Claro

, Srougi

. Utility of the Guy's stone score based on computed tomographic scan findings for predicting percutaneous nephrolithotomy outcomes. Urology, 2014; 83:1248–1253.

15.

Skolarikos

, Binbay

, Bisas

, et al. Percutaneous nephrolithotomy in horseshoe kidneys: Factors affecting stone-free rate. J Urol, 2011; 186:1894–1898.

16.

Shokeir

, El-Nahas

, Shoma

, et al. Percutaneous nephrolithotomy in treatment of large stones within horseshoe kidneys. Urology, 2004; 64:426–429.

17.

Kuntz

, Neisius

, Astroza

, et al. Does body mass index impact the outcomes of tubeless percutaneous nephrolithotomy?. BJU Int, 2014; 114:404–411.

18.

Zhou

, Sun

, Chen

, et al. Effect of obesity on outcomes of percutaneous nephrolithotomy in renal stone management: A systematic review and meta-analysis. Urol Int, 2017; 98:382–390.

19.

Tzou

, Metzler

, Usawachintachit

, Stoller

, Chi

. Ultrasound-guided access and dilation for percutaneous nephrolithotomy (PCNL) in the supine position: A step-by-step approach. Urology, 2019; 133:245–246.

20.

Vicentini

, Perrella

, Souza

VMG

, Hisano

, Murta

, Claro

JFA

. Impact of patient position on the outcomes of percutaneous neprolithotomy for complex kidney stones. Int Braz J Urol, 2018; 44:965–971.

21.

Sofer

, Giusti

, Proietti

, et al. Upper calyx approachability through a lower calyx access for prone versus supine percutaneous nephrolithotomy. J Urol, 2016; 195:377–382.