Urolithiasis (2014) 42:255–262 DOI 10.1007/s00240-014-0638-3
ORIGINAL PAPER
Tubeless percutaneous nephrolithotomy: yes but when? A multicentre retrospective cohort study Murat M. Rifaioglu • Kadir Onem • Ibrahim Buldu Tuna Karatag • Mustafa Okan Istanbulluoglu
•
Received: 3 October 2013 / Accepted: 8 January 2014 / Published online: 28 January 2014 Ó Springer-Verlag Berlin Heidelberg 2014
Abstract The aim of our study is to determine the predictive factors for placement of percutaneous nephrostomy tube (PNT) in percutaneous nephrolithotomy (PCNL) procedure and to evaluate the optimal cutoff points of the predictive factors. 229 patients, who had undergone percutaneous nephrolithotomy operation between February 2009 and February 2013 were reviewed retrospectively. Five patients were excluded from the study because of solitary kidney. All characteristics of 224 patients, stones and operative data were investigated. Patient and stonerelated factors, such as age, BMI, history of previous surgery or SWL, characteristics of the stone, renal parenchymal thickness (RPT), as well as procedural factors, such as percutaneous access number and location were analyzed by univariate and multivariate tests. The continuous variables were analyzed using Receiver operating characteristic curve analysis. There were no differences in sex, age, BMI and hemoglobin decrease between the groups. Previous operation status, RPT, stone size, multiplicity of the stone, stone localization, blood transfusion presence, access points, access number and operation time were found sta-
tistically different according to univariate analysis. Multivariate analysis showed that RPT, operation time and stone size were the independent factors that affected the PNT insertion. For RPT, operation time and stone size, the optimal cutoff points for insertion PNT were 13.75 mm, 75.5 min and 890 mm2, respectively. Tubeless PCNL should be chosen in patients with stone area less than 890 mm2, and parenchymal thickness thicker than 13.75 mm and procedure with operation period\75.5 min. Keywords Tubeless Percutaneous nephrolithotomy Renal parenchymal thickness Operation time Stone size Cutoff point Abbreviations PNT Percutaneous nephrostomy tube PCNL Percutaneous nephrolithotomy RPT Renal parenchymal thickness BMI Body mass index SWL Shock wave lithotripsy ROC Receiver operating characteristic
Introduction M. M. Rifaioglu (&) Urology Department, Medical Faculty, Mustafa Kemal University, 31005 Antakya, Serinyol, Hatay, Turkey e-mail:
[email protected] K. Onem Urology Department, Medical Faculty, Ondokuzmayıs University, Samsun, Turkey I. Buldu T. Karatag M. O. Istanbulluoglu Urology Department, Mevlana University Medical Faculty, Konya, Turkey
Placement of a percutaneous nephrostomy tube (PNT) at the end of the percutaneous nephrolithotomy (PCNL) is a standard procedure to tamponade bleeding from the access tract, maximize perioperative collecting system drainage, and maintain access for possible secondary procedures. Despite these potential advantages, multiple prospective comparative series have demonstrated that PNTs are a significant source of postoperative pain leading to longer hospital stays and increased analgesic requirements [1–4].
123
256
Initially, in 1984, Wickham and associates [5] argued that PNT placement could be omitted after selected PCNL cases. Main indications for tubeless PCNL (ureteral stent without PNT) included cases with a single access tract, minimal blood loss, no collecting system perforation, and no intention of second-look nephroscopy [1, 2, 6–9]. The indications of placement of the PNT are certain preoperatively, though, the time needed to decide insertion of a PNT during operation is not clear, yet. The purpose of our study is to determine the predictive factors for placement of PNT in PCNL procedure and to evaluate the optimal cutoff points of the predictive factors.
Patients and methods Patients This is a retrospective review of two centers undergoing PCNL, comparing the results of two cohorts: Patients who had a PNT placed and those who did not. Between February 2009 and 2013, we performed PCNLs on 229 patients. Five patients were excluded from the present study because of solitary kidney. All the 224 patients were evaluated preoperatively using abdominal plain radiography (KUB), intravenous urography, ultrasonography, and/ or computed tomography. One hundred and seventeen (52.2 %) patients had a PNT placed at the end of the case, and 107 (47.8 %) did not. One of two experienced (case load at least 50 [10]) urologists (MMR or OI) acquired percutaneous access and performed the PCNL in every case as previously reported [11, 12]. Patient demographic characteristics, including age, sex, body mass index (BMI) and history of shock wave lithotripsy (SWL) were recorded. Preoperative laboratory tests included serum creatinine, hemoglobin measurements, platelet counts, coagulation screen tests, and urine cultures. All patients had either sterile urine cultures or they were treated according to the antibiotic sensitivity tests. Method Renal parenchymal thickness (RPT) [11] was measured at the planned calix of puncture (reported as the average in case of multiple tracts) on ultrasonography and IVU, respectively [13]. Complete blood count (done 24 h before and after the procedure) and the number of blood units transfused (between the preoperative and the final blood count) determined the perioperative hemoglobin decrease [13]. Stone size was assessed as the surface area calculated according to European Association of Urology guidelines [14]. Stone multiplicities were classified as either multiple
123
Urolithiasis (2014) 42:255–262
(partial or complete staghorn calculi or multiple calyceal stones) or single including pelvic stones, and isolated calyceal stones. Stone locations were classified as either renal or both renal and ureter. Access points were defined as supracostal or infracostal. The operation time was defined as the time from the puncture for an access tract to the final placement of a PNT. Treatment success was defined as no residual stone fragment greater than 5 mm on postoperative computerized tomography or plain abdominal radiography, and no need any further intervention [15]. Repeat PCNL, ureteroscopy, and SWL were considered as accessory treatment alternatives when indicated. Clavien-Dindo grading system was used for the classification of surgical complications [16]. Length of hospital stay was defined as the time interval between day of the surgery and discharge from hospital. The decision to place a PNT at the end of the procedure was depended on surgeon’s discretion. Patients in whom a PNT was placed had a malecot re-entry (range 16–18 Fr) PNT. General indications for tubeless PCNL were minimal bleeding during the case, no evidence of collecting system perforation, and low index of suspicion for residual stone fragments. In the PNT group, the urethral catheter was removed the day after surgery. The tube was removed on the second postoperative day if there was no significant bleeding or retained stone fragments. Patients were discharged home later that day if pain was controlled. In the tubeless group, the urethral catheter was removed the morning after surgery if there was no urine extravasation from the access site. Patients were then discharged home once their pain was controlled with oral agents. The primary end point of the study was the incidence of perioperative complications. Secondary end points included estimated blood loss, length of hospital stay, and treatment success. Statistical analysis In the univariate analyses, Mann–Whitney U test for continuous data, Chi-square test, and Fisher exact test for binomial data were performed to compare the two cohorts. Patient and stone-related factors, such as age, BMI, history of previous open renal surgery or SWL, characteristics of the stone, RPT, as well as procedural factors, such as percutaneous access number and location, were analyzed by univariate and multivariate tests using SPSS 16 Windows software (SPSS, Inc., Chicago, IL). The continuous variables were analyzed using Receiver operating characteristic (ROC) curve analysis. When a significant cutoff value was observed, the sensitivity, specificity, positive and negative predictive values were presented. A multivariate
Urolithiasis (2014) 42:255–262
binary logistic regression analysis was performed for further investigation if any parameter was found to be significant with univariate analyses. The Hosmer–Lemeshow goodness-of-fit test was adopted to evaluate the significance of the logistic regression model. In all analyses, twosided hypothesis testing was carried out, and the significance level for P value was assumed to be \0.05.
Results Patient demographics, access characteristics, and perioperative outcomes were summarized in Table 1. There were no differences between the groups according to sex, age, BMI and hemoglobin decrease. There were higher median preoperative creatinine levels and RPT in tubeless group than nephrostomy group (P \ 0.001) (Table 1). Both renal and ureter stone localizations were obtained in 14 (13.1 %) of the tubeless patients, and none of the nephrostomy patients. In nephrostomy group, patients had higher multiple (partial or complete staghorn calculi or multiple calyceal stones) calyceal stones and higher median stone volume than tubeless group (64.1 vs. 49.5 %, P = 0.028, 700 vs. 420, P \ 0.001, respectively). More patients in the nephrostomy group (19.7 %) had undergone a previous surgery (PCNL or open) than in the tubeless cohort (6.5 %) (P = 0.004). Two access tracts were used in 20 (17.1 %) nephrostomy patients and 6 (5.6 %) tubeless patients. The remainder of the patients underwent only one tract dilation. Supracostal access was used in 16 (15 %) tubeless patients and 4 (3.4 %) nephrostomy patients. Overall success was achieved in 93.5 and 67.5 % of tubeless and nephrostomy patients after one session of PCNL, respectively. The success was confirmed in 220 (98 %) and 4 (2 %) patients by computerized tomography and plain radiography, respectively. Median hospitalization time was significantly higher in nephrostomy group compared with tubeless group (P \ 0.001). Blood transfusion rates and Clavien score results were lower in tubeless group in comparison to nephrostomy group. For creatinine level, the optimal cutoff point on the ROC curve for insertion PNT was 1.7 mg/dL with a sensitivity of 43 % and specificity of 97.2 % (area under the curve 0.635, P \ 0.0001). For operation time, the optimal cutoff point on the ROC curve for insertion PNT was 75.5 min with a sensitivity of 76.9 % and specificity of 85 % (area under the curve 0.852, P \ 0.0001, Fig. 1a). For RPT, the optimal cutoff point on the ROC curve for insertion PNT was 13.75 mm with a sensitivity of 65.7 % and specificity of 87.7 % (area under the curve 0.805, P \ 0.0001, Fig. 1b). For stone size, the optimal cutoff point on the ROC curve for insertion PNT was 890 mm2 with a sensitivity of 46.2 % and specificity of 88.8 % (area under the curve 0.673,
257
P \ 0.0001, Fig. 1c). Patients were categorized into two groups according to cutoff point of RPT, stone size and operation time (cutoff = 13.75 mm; group 1: B 13.75 mm; group group 2: [ 13.75 mm, cutoff = 890 mm2; 2 2 1: B 890 mm ; group 2: [ 890 mm , cutoff = 75.5 min; group 1: B 75.5 min; group 2: 75.5 min, respectively). These categorized parameters according to cutoff points were summarized in Table 2. Multivariate analyze results were summarized in Table 3. Multivariate binary logistic regression analysis showed that RPT (OR 4.781, 95 % CI 1.640–13.940, P = 0.004) operation time (OR 8.970, 95 % CI 3.066–26.237, P \ 0.001) and stone size (OR 3.749, 95 % CI 1.044–13.461, P = 0.043) were the independent factors that affected the PNT insertion. The Hosmer–Lemeshow goodness-of-fit test was performed and obtained a P value of 0.258, which indicated a good fit of the data.
Discussion It is well known that tubeless PCNL is a safe and effective procedure and has advantages of reducing the hospital stay, analgesic requirement, and the time to return to normal activity without increasing complications in comparison to standard PCNL [17]. In a meta-analysis comparing the tubeless vs. standard PCNL, it was concluded that tubeless PCNL was a good option in non-complicated cases, with the advantages of reduced hospital stay and little need for postoperative analgesia [18]. In current study, our aim was to evaluate the definition of ‘‘non-complicated cases’’. Previous studies on tubeless PCNL worldwide had been published, there seems to be agreement that tubeless PCNL should be performed in selected patients [19–23]. The authors propose a few conditions for tubeless PCNL: (1) Operation duration should be \1 h, (2) stone size should be \3 cm, (3) there should be no noticeable hemorrhage during the operation, (4) there should be no significant pelvic or ureteric perforations, (5) there should be no requirement for a secondary percutaneous procedure for residual stones and (6) there should be no distal stone obstruction, ureteric strictures or severe prostate hyperplasia [18]. In accordance with literature, in present study, there was a selection bias such as single access tract, minimal bleeding during the case, no evidence of collecting system perforation, and low index of suspicion for residual stone fragments in order not to hazard the patient. This selection could be reduced by the complication rate in the tubeless group than the nephrostomy group (4.7 vs. 27.4 % P \ 0.001) and this made the tubeless group such as ‘‘non-complicated’’ group. Some experts consider that tubeless PCNL is unsuitable for cases whose stone size is above 3 cm [24], but others
123
258
Urolithiasis (2014) 42:255–262
Table 1 Comparison of characteristics of patients, stones and operative data Parameters
Tubeless group (107 patients) n (%)
Nephrostomy tube group (117 patients) n (%)
69 (64.5)
81 (69.2)
Sex Male Female
P value
0.451 38 (34.5)
36 (30.8)
Age median (min–max)
51 (8–76)
48.5 (14–73)
0.07
BMI median (min–max) (kg/m2)
27.75 (16.65–46.56)
27.75 (16.89–35.84)
0.613
Creatinine median (min–max) (mg/dl)
0.9 (0.42–2.6)
0.8 (0.46–1.9)
