- To evaluate whether body mass index (BMI) has an impact on the outcomes of tubeless percutaneous nephrolithotomy (PCNL).
Tubeless percutaneous nephrolithotomy (PCNL) has become a well adopted technique, owing to a multitude of benefits including decreased cost, less morbidity and earlier convalescence . While evidence has shown conventional PCNL to be a safe and effective procedure in obese patients, to date there has been limited data demonstrating similar results using a tubeless approach [2, 3]. Therefore, we retrospectively reviewed the outcomes of patients who underwent tubeless PCNL at our institution to determine if there was a relationship between body mass index (BMI) and stone-free or complication rates of tubeless PCNL.
After approval from the Duke University Medical Center Institutional Review Board, a database was compiled consisting of demographic and clinical information from all patients who underwent PCNL at our institution from 1 June 2006 until 30 June 2011. From this cohort, patients managed with a tubeless approach were selected for outcomes analysis, namely stone-free rates, complications and hospital length of stay (LOS). Patients without objective BMI data (obtained by clinical staff), and at least 3 months follow-up were excluded. All procedures were performed by one surgeon (G.M.P.). Percutaneous access was obtained by an experienced attending interventional radiologist in the operating room at the time of PCNL. The number and locations of tracts were at the discretion of the attending urologist, based on stone characteristics and anatomical variables. A high-pressure balloon was used for tract dilatation in all cases. Tubeless was defined as not leaving any type of percutaneous catheter at the conclusion of the operative case. However, patients who were left with an internal ureteric stent or an externalised ureteric catheter were also considered tubeless for the purpose of this study, in accordance with previously described terminology . Over the last few years, tubeless or tubeless, stentless PCNL has become our standard procedure for percutaneous stone removal except in those patients with a recognised collecting system injury, significant intraoperative bleeding, urinary diversion or planned staged procedures. Open-ended external ureteric stents were left overnight and removed before discharge on postoperative day 1. Indications for an internal ureteric stent included significant ureteric manipulation, nephrostomy tracts above the 11th rib and impacted stones at the PUJ.
Stone-free rates were determined using non-contrast CT or IVU at the 3-month follow-up, and required complete absence of any residual fragments. Complications were graded using the Clavien-Dindo system , and a comprehensive list is shown in Table 1. Urinoma was defined as any symptomatic (commonly persistent pain, fever of unknown origin, ileus) postoperative fluid collection consistent with urine identified on CT. The following clinical variables were collected: age, gender, BMI, American Society Anesthesiologists (ASA) physical classification score, stone characteristics (single, multiple, staghorn), stone burden, number of tracts (1 or ≥2), location of single punctures (upper, mid or lower pole), preoperative and postoperative haematocrit, type of ureteric stent (external, internal or completely stentless), and stone composition. Stone burden was calculated by multiplying the anterior–posterior and lateral–medial dimensions of the stone on axial images from the preoperative CT. In cases of multiple stones, the three largest stones were measured and the stone burden was considered the sum of the three.
|2||Retroperitoneal haematoma||2 (0.7)||Blood transfusion|
|Haematuria||4 (1.4)||Blood transfusion|
|Atrial fibrillation||2 (0.7)|
|3||Haemothorax||1 (0.4)||Pleural drain|
|Urinoma||8 (3.0)||Percutaneous drain or ureteric stent placement|
|Pneumothorax||1 (0.4)||Chest tube|
|Migrated ureteric stent||3 (1.1)||URS|
|Re-operation for residual stones||3 (1.1)||URS|
|Myocardial infarct||1 (0.4)|
The patients were divided into four groups based on BMI quartiles using a modified WHO classification scheme: <25 kg/m2 was considered ‘normal’; 25–29.9 kg/m2 was considered ‘overweight’, 30–34.9 kg/m2 was considered ‘obese’; ≥35 kg/m2 was considered ‘morbidly obese’ . Descriptive statistics were generated for the entire cohort, as well as stratified and compared by BMI group using non-parametric and Fisher's exact tests as appropriate. The outcomes (stone-free rates, rates of any, mild and severe complications, transfusion and urinoma rates as well as LOS of tubeless PCNL were compared between the groups using Fisher's exact tests. Multivariable logistic regression models were used to evaluate the independent contribution of BMI as a predictor of the outcomes. For these models, BMI was considered both as a continuous and as categorized variable. All models were adjusted for age, gender, ASA category and stone characteristics. In addition, the LOS model was also adjusted for complications using Clavien grade. All tests were two-tailed and a P < 0.05 was considered to indicate statistical significance. Data are presented as median (interquartile range [IQR]) or number (percentage) unless otherwise specified. The analyses were performed using the R v2.13 software (The R Foundation for Statistical Computing, Vienna, Austria).
In all, 509 patients were identified that underwent PCNL. Data for BMI calculations were missing for 41 patients who were excluded. Of the remaining PCNL patients, 268 (57%) were performed in a tubeless fashion and were included in the present study. The percentage of cases performed tubeless in 2006 was 20%, compared with 62% in 2011, thus clearly representing an evolution in this technique over time. Patient characteristics and their comparison between BMI groups are detailed in Table 2. The median (IQR) age was 51 (42–61) years, 47% were men and the median (IQR) BMI was 30.3 (25.9–35.7) kg/m2.
|Variable||Overall||BMI group, kg/m2||P|
|Patients, n (%)||268 (100)||55 (21)||74 (28)||67 (25)||72 (25)|
|Median (IQR) age, years||51 (42–61)||58 (41–67)||51 (36–62)||52 (45–61)||49 (41–57)||0.123|
|Gender: n (%)||0.470|
|male||125 (47)||22 (40)||37 (50)||35 (52)||31 (43)|
|female||143 (53)||33 (60)||37 (50)||32 (48)||41 (57)|
|Median (IQR) BMI, kg/m2||30.3 (25.9–35.7)||22.4 (20.4–24.0)||27.4 (26.3–28.8)||32.0 (31.2–33.3)||40.5 (37.4–45.9)||<0.001|
|ASA categories: n (%)||<0.001|
|1||30 (12)||14 (25.5)||16 (21.6)||0||0|
|2||125 (49)||20 (36.4)||39 (52.7)||38 (57.6)||28 (45.2)|
|3||102 (40)||21 (38.2)||19 (25.7)||28 (42.4)||34 (54.8)|
|Stone characteristics: n (%)||0.010|
|single||90 (34)||18 (32.7)||23 (31.1)||26 (38.8)||23 (31.9)|
|staghorn||51 (19)||11 (20.0)||24 (32.4)||10 (14.9)||6 (8.3)|
|multiple||127 (47)||26 (47.3)||27 (36.9)||31 (46.3)||43 (59.7)|
|Median (IQR) stone burden*, mL||2.58 (1.49–4.17)||2.09 (1.41–3.49)||1.98 (1.21–3.51)||2.80 (2.07–4.17)||3.00 (1.68–4.95)||0.485|
|Number of tracts: n (%)||0.664|
|1||239 (93)||50 (90.9)||67 (91.8)||62 (95.4)||60 (95.2)|
|≥2||17 (7)||5 (9.1)||6 (8.2)||3 (4.6)||3 (4.8)|
|Tract location: n (%)||0.116|
|upper pole||33 (14.3)||5 (10.4)||4 (6.2)||12 (19.3)||12 (21.0)|
|mid pole||104 (45.0)||22 (45.8)||27 (42.2)||28 (45.2)||27 (47.4)|
|lower pole||94 (40.7)||21 (43.8)||33 (51.6)||22 (35.5)||18 (31.6)|
|Stenting: n (%)||0.251|
|JJ||132 (49)||33 (60.0)||30 (40.5)||36 (53.7)||33 (45.8)|
|open ended||129 (48)||20 (36.4)||41 (55.4)||30 (44.8)||38 (52.8)|
|stentless||7 (3)||2 (3.6)||3 (4.1)||1 (1.5)||1 (1.4)|
|Stone composition: n (%)||<0.001|
|phosphate||72 (27)||20 (36.4)||20 (27.0)||19 (28.4)||13 (18.1)|
|monohydrate||79 (29)||12 (21.8)||21 (28.4)||20 (29.9)||26 (36.1)|
|uric acid||20 (7)||1 (1.8)||2 (2.7)||12 (17.9)||5 (6.9)|
|brushite||19 (7)||5 (9.1)||4 (5.4)||6 (9.0)||4 (5.6)|
|struvite||18 (7)||8 (14.5)||5 (6.6)||2 (3.0)||3 (4.2)|
|dihydrate||16 (6)||1 (1.8)||9 (12.2)||4 (6.0)||2 (2.8)|
|calcium oxalate||3 (1)||0||0||1 (1.5)||2 (2.8)|
|unknown||41 (15)||8 (14.5)||13 (17.6)||3 (4.5)||17 (23.6)|
Of note, a higher BMI was associated with a higher ASA score (P < 0.001). We also noted differences in stone characteristics across BMI groups (Table 2). In fact, while multiple stones were more common in higher BMI groups, there were more staghorn stones in patients with lower BMIs (P = 0.010). Additionally, there was a trend toward a higher stone burden in heavier patients, but this was not statistically significant (P = 0.485). Overall, stone composition varied significantly across BMI categories (P < 0.001, Table 2), but the most common was calcium oxalate (95, 35%), and almost all uric acid stones (85%) were found in patients with a BMI of ≥30 kg/m2. Despite these differences, the number and location of tracts used as well as the methods of ureteric stenting were comparable across the groups.
The overall stone-free rate was 52.5%, and univariable analyses found no association with BMI (P = 0.864, Table 3). IVU with tomograms was the most common postoperative imaging method used, independently of BMI group. In addition, IVU yielded the highest stone-free rate at 60% compared with 34% and 33% for CT and plain abdominal radiograph of the kidneys, ureters and bladder (KUB), respectively (P = 0.001). In all, 51 patients (19.0%) had complications with similar rates across BMI groups (P = 0.890). Minor (Clavien 1–2) and major (Clavien 3–4) complications occurred in 10.4% and 8.6%, respectively (Table 3). The rates of minor and major events, including urinomas were comparable across BMI groups (Table 3). However, there was significantly more blood transfusions in patients with a normal BMI (P = 0.005, Fig. 1), despite the fact there was no difference in the preoperative haematocrit between BMI groups (P = 0.381). Overall, there was a decrease in haematocrit of 6% after tubeless PCNL, compared with a 19% decrease in patients that received a blood transfusion, indicating that transfusion correlates with postoperative bleeding. In addition to having more blood transfusions, patients with a normal BMI were significantly more likely to have a prolonged LOS (≥2 days), when compared with an overweight BMI (Fig. 2, odds ratio [OR] 0.039–0.046, P = 0.032). Additional multivariable analyses, adjusting for demographic and clinical characteristics, did not reveal an association between BMI and stone-free rates, overall or severe complications (Table 4).
|Outcome||Overall||BMI group, kg/m2||P|
|Stone-free||115 (52.5)||25 (53.2)||27 (50.0)||33 (56.9)||30 (50.0)||0.864|
|Postoperative imaging method:||0.947|
|CT||44 (20.3)||9 (19.1)||13 (24.5)||10 (17.6)||12 (20.0)|
|IVU||149 (68.6)||31 (66.0)||35 (66.0)||41 (71.9)||42 (70.0)|
|KUB||24 (11.1)||7 (14.9)||5 (9.5)||6 (10.5)||6 (10.0)|
|Overall||51 (19.0)||10 (18.2)||16 (21.6)||13 (19.7)||12 (16.7)||0.890|
|1||20 (7.5)||3 (5.5)||7 (9.5)||5 (7.5)||5 (6.9)|
|2||8 (3.0)||4 (7.3)||0||2 (3.0)||2 (2.8)|
|3||19 (7.1)||3 (5.5)||8 (10.8)||4 (6.0)||4 (6.0)|
|4||4 (1.5)||1 (1.8)||0||1 (1.5)||2 (2.8)|
|Clavien grade 1–2||28 (10.4)||7 (12.7)||7 (9.5)||7 (10.4)||7 (9.7)||0.937|
|Clavien grade 3–4||23 (8.6)||4 (7.3)||8 (10.8)||5 (7.5)||6 (8.3)||0.905|
|Urinoma||15 (5.6)||3 (5.4)||5 (6.7)||5 (7.5)||2 (2.8)||0.638|
|Transfusion||9 (3.4)||6 (10.9)||0||1 (1.5)||2 (2.8)||0.005|
|Median (IQR) preoperative haematocrit, %||41 (37–43)||39 (36–43)||41 (38–44)||40 (37–44)||41 (38–44)||0.381|
|N (%) LOS, days||0.229|
|0–1||179 (67.5)||30 (54.5)||54 (73.0)||47 (71.2)||48 (68.6)|
|2–3||58 (21.9)||17 (30.9)||15 (20.3)||10 (15.2)||16 (22.9)|
|≥4||28 (10.6)||8 (14.5)||5 (6.8)||9 (13.6)||6 (8.6)|
|Outcome||BMI groups, kg/m2||OR (95% CI)||P|
|Stone-free rate||continuous||1.00 (0.96–1.03)||0.960|
|Complications, any||continuous||0.98 (0.94–1.02)||0.509|
|Complications, severe||continuous||0.98 (0.93–1.03)||0.509|
Obesity is a growing health epidemic with a prevalence of 32.2% and 35.5% among men and women, respectively . As obesity has consistently been shown to be an independent risk factor for stone formation , it comes as no surprise that the incidence of stone disease is on the rise as well . Given these trends over the last few decades there have been considerable efforts in identifying the best therapeutic strategies for treating obese patients with complex renal and ureteric calculi.
Shockwave lithotripsy (SWL) is a recommended first-line treatment of small (<2 cm) renal and ureteric stones . However, there have been several reported limitations for its use in obese patients, namely poor stone focusing and table weight restrictions. A large skin-to-stone distance, as seen in obese patient, is also thought to decrease the effectiveness of SWL. Pareek et al. [10, 11] reported increased skin-to-stone distance as well as higher BMI to be a significant predictor of worse stone-free outcomes for lower pole renal stones. Given these limitations, SWL is not considered an ideal treatment method for managing stones in obese patients. Consequently, this has led to an increase in surgical methods, such as PCNL. In fact, using data from the Healthcare Cost and Utilization Project, Morris et al.  reported that the PCNL rate had doubled (1.4–2.5) between 1988 and 2002. In addition to the increasing incidence of obesity and stone complexity, many factors are probably contributing to this trend, e.g. improved technology, less invasive surgical techniques, and superior stone-free outcomes.
Although obese patients are at increased risk for medical co-morbidities associated with adverse surgical outcomes, obesity has not been shown to increase operative morbidity or mortality compared with non-obese patients . This finding appears to hold true for tubeless PCNL as well. The presen results suggest that obese patients have higher ASA scores, without additional complications. With appropriate precautions and perioperative considerations from a multi-organ standpoint, surgery can be safely and effectively be performed in obese patients .
As the optics and stone fragmenting technologies have improved, the application of ureterorenoscopy (URS) in obese patient has increased over the last decade. A recent systematic review by Aboumarzouk et al.  reported stone-free and complication rates of 87.5% and 11.4%, respectively using URS in patients with a mean BMI of 42.2 kg/m2. Stone burden remains the biggest limitation with URS, as stones >2 cm treated with URS are more likely to require an additional procedure, and have longer operating times . As such, PCNL, remains the ‘gold standard’ for the treatment of large (>2 cm), complex renal and proximal ureteric stones [17, 18].
Several unique challenges of PCNL in the obese population have led to widespread reporting of outcomese. When looking at PCNL in 5803 patients stratified by BMI, Fuller et al.  reported no difference in the complication rates, but inferior stone-free rates in the obese cohort. Conversely, both El-Assmy et al. . and Tomaszwski et al.  found that stone-free rates were independent of BMI, in addition to complication rates and LOS. Despite well documented difficulties with positioning, anaesthetic administration and fluoroscopic imaging, evidence suggests that PCNL is safe and effective in obese patients [2, 3, 19-22].
In 1997, experimentation with various PCNL modifications led to the first formal series of tubeless PCNL reported by Bellman et al. . However, the concept itself significantly predates this. In one of the earliest reports of PCNL in 1984, Wickham et al.  comments that he preferred nephrostomy tube removal postoperatively; however, the percentage of patients managed this way was not reported. The presence of the nephrostomy tube postoperatively was historically thought to provide haemostasis, maintain renal drainage, prevent infection, and maintain access for ‘second-look’ procedures. However, routine use in all patients was hypothesised to be a major contributor to patient discomfort and increased LOS. Several studies have since shown the tubeless approach is not only safe and effective, but leads to decreased postoperative pain, analgesia use, LOS and cost [1, 23, 25-31]. While there have been several series reporting tubeless PCNL to be safe and effective in the elderly (age >60 years) , children (age <16 years)  and in patients with renal anomalies , to our knowledge there has been only one study evaluating tubeless PCNL in obese patients.
In 2004, Yang and Bellman  prospectively reviewed their series of 104 tubeless PCNL patients, stratified by BMI. While excluding patients with a surgical time of >2 h, significant bleeding, significant residual stone burden and staged ‘second-look’ nephroscopy they reported an overall transfusion and stone-free rate of 9% and 94%, respectively. Although 75% of their cohort was not obese (BMI >30 kg/m2), they found equivalent outcomes among BMI groups for stone-free rate, blood transfusions and LOS .
The present review provides a larger, less selective, and more contemporary analysis of this growing cohort of obese patients requiring percutaneous stone treatment. Additionally, for the first time we are reporting results that suggests obesity may provide a ‘protective’ role when performing tubeless PCNL. Theoretically, the abundance of retroperitoneal fat in obese patients could provide added external compression and support after removal of the percutaneous catheter, thus effectively decreasing the potential for haemorrhage and urine extravasation. Conversely, patients with a normal BMI may be at increased risk for complications, e.g. haematoma and urinoma, secondary to the paucity of retroperitoneal fat.
We retrospectively reviewed the present cohort of 268 patients that underwent a tubeless PCNL, and stratified them by BMI to assess the impact on outcomes, including stone-free rates, complications and LOS. Overall, there was no significant correlation between BMI and stone-free rates or overall complications. However, on multivariable analysis we found LOS was significantly longer in normal weight individuals despite lower ASA classification and equivalent stone burden. This could be related postoperative bleeding, as we found a statistically significant increase in the transfusion rates in normal weight patients (P = 0.005). When individually assessed, the indication for transfusion could not be obtained from the medical record for one patient. However, the remaining eight received a blood transfusion for symptomatic anaemia, with a haematocrit of ≤22% and a median decrease in haematocrit of 19% in all cases. As these patients were not anaemic preoperatively compared with the overall cohort (median haematocrit 39% vs 41% overall), this fact suggests that blood transfusion correlates well with postoperative bleeding. One potential cofounder for this finding could be stone characteristics, as there were significantly more staghorn calculi in normal and overweight patients. However, multivariable models adjusted for this, and found no association with BMI and stone-free rates or complications. Also, the highest number of staghorn stones (33%) was in the overweight group, which had no transfusions, and only one transfusion occurred in a patient with a complete staghorn stone. Contrary to the present study, Yang and Bellman  found no statistical difference in the transfusion rates between BMI groups, but interestingly, all nine blood transfusions were in patients with a BMI of <30 kg/m2.
The reported overall complication rate for PCNL in obese and non-obese patients is between 2.3% and 22.1% [3, 19, 20, 36, 37]. The first reported series of tubeless PCNL found equal complication rates compared with PCNL in 112 patients . Since that time, many studies have corroborated this finding . In the present study, we had an overall complication rate of 19.0%. Our most common complications were bleeding (requiring transfusion) and urinoma at rates of 3.4% and 5.6%, respectively. In other series of tubeless PCNL transfusion rates have been between 0% and 11.9%, while urinomas occur in 0–4.1% of patients . Assuming many urinomas are asymptomatic and go undetected, this finding is likely an underestimation of the true rate, as we did not routinely image all patients in the immediate postoperative period. Nevertheless, of the 15 urinomas encountered in the present series only eight required a drain or ureteric stent placement.
The reported stone-free rates for PCNL vary widely owing to discrepancies in the definition of ‘stone-free’, the type of imaging used as well as the timing for postoperative imaging. Early studies from the 1980s reported outstanding (>90%) stone-free rates [38, 39]. However, many early studies lacked a clear definition for stone-free criteria and used earlier and less sensitive imaging techniques. In a contemporary series of 987 patients, where the strictest stone- free definition was applied (no residual fragments), the stone-free rate was 76% overall, and only 47% for staghorn stones . Nonetheless, PCNL success rates appeared to be unchanged in the obese population in most cases, with exception of one series, which reported a 65.6% stone-free rate in super-obese patients (BMI >40 kg/m2), compared with 77.5% in normal weight patients (P < 0.001) . Additionally, tubeless PCNL has been shown to have equivalent stone-free results in almost all accounts .
In the present study, we have reported a 52.5% overall stone-free rate that was also not impacted by BMI. The relatively low stone-free rate in the present report may reflect several factors, such as our use of a strict definition of no residual fragments, the complexity of the stones treated (66% multiple or staghorn stones) and that we did not routinely perform secondary procedures to treat residual fragments. In addition, we did not exclude patients based on surgery duration or stone burden. Yang and Bellman  reported stone-free rates of between 80% and 94%, but with more rigorous exclusion criteria including: surgical time >2 h, >2 percutaneous tracts, significant residual stone burden, and staged ‘second-look’ nephroscopy.
Limitations of the present study include the retrospective design from a single institution. Overall, we think that BMI did not influence the choice of other treatment methods, as we perform very few SWL (<10 annually) and prefer PCNL in any stone >2 cm, regardless of BMI (assuming they can tolerate prone positioning). However, while more recently we have adopted the tubeless approach for most routine PCNLs at our institution, the indications for leaving a nephrostomy tube, although defined above, are still subjective and left to the discretion of the surgeon. This fact may introduce a selection bias. Additionally, we used CT for postoperative imaging in >20% of cases. CT is known to have increased sensitivity compared with either IVU or KUB. This may partly account for the overall lower stone-free rates. However, in our clinical practice the selection of imaging studies is often based on clinical scenario. Patients are often selected to undergo CT when there is a strong suspicion of residual fragments, which were not clearly visualised on intraoperative fluoroscopy or nephroscopy. This is particularly true in patients we may select for a staged procedure. This selection bias may limit the significance of the reported stone-free rate by imaging method, but should not affect the validity of study overall.
In conclusion, we have reported a large series examining tubeless PCNL in the obese population. We have shown equivalent safety and efficacy outcomes between BMI groups, with exception of LOS and transfusion rates, which were significantly higher in patients with a normal BMI. The present study further supports the use of tubeless PCNL in obese individuals. However, caution should be taken with thin individuals who may be at increased risk of bleeding and longer LOS. Further studies are needed to support this finding and better determine the mechanism underlying this observation.
A.N. reports grants from Ferdinand Eisenberger of the Deutsche Gesellschaft für Urologie (German Society of Urology), ID NeA1/FE-11, outside the submitted work.
M.N.F reports fees from Intuitive Surgical, outside the submitted work.
G.M.A reports grants from Endourological Society and Cook Urological, outside the submitted work.
M.E.L. has received personal fees from Boston Scientific Corp., outside the submitted work.
G.M.P. has received personal fees from Boston Scientific Corp. and Mission Pharmacal, outside the submitted work.
No other conflicts disclosed.
American Society Anesthesiologists
body mass index
plain abdominal radiograph of the kidneys, ureters and bladder
hospital length of stay
(tubeless) percutaneous nephrolithotomy