Tubeless percutaneous nephrolithotomy: a prospective feasibility study and review of previous reports


Hemendra N. Shah, Chief Urologist, R. G. Stone Urological Research Centre, Mumbai, Maharashtra, India.



To evaluate the status of tubeless percutaneous nephrolithotomy (PCNL) in managing renal and upper ureteric calculi, from initial experience and a review of previous reports.


From September 2004 to December 2004, 46 patients were scheduled for tubeless PCNL in a prospective study. Patients with solitary kidney, or undergoing bilateral simultaneous PCNL or requiring a supracostal access were also enrolled. Patients needing more than three percutaneous access tracts, or with significant bleeding or a significant residual stone burden necessitating a staged second-look nephroscopy were excluded. At the end of the procedure, a JJ ureteric stent was placed antegradely and a nephrostomy tube avoided. The patients’ demographic data, the outcomes during and after surgery, complications, success rate, and stent-related morbidity were analysed. Previous reports were reviewed to evaluate the current status of tubeless PCNL.


Of the 46 patients initially considered only 40 (45 renal units) were assessed. The mean stone size in these patients was 33 mm and 23 patients had multiple stones. Three patients had a serum creatinine level of >2 mg/dL (>177 µmol/L). Five patients had successful bilateral simultaneous tubeless PCNL. In all, 51 tracts were required in 45 renal units, 30 of which were supracostal. The mean decrease in haemoglobin was 1.2 g/dL and two patients required a blood transfusion after PCNL. There was no urine leakage or formation of urinoma after surgery, and no major chest complications in patients requiring a supracostal access tract, except for one with hydrothorax, managed conservatively. The mean hospital stay was 26 h and analgesic requirement 40.6 mg of diclofenac. Stones were completely cleared in 87% of renal units and 9% had residual fragments of < 5 mm. Two patients required extracorporeal lithotripsy for residual calculi. In all, 30% of patients had bothersome stent-related symptoms and 60% needed analgesics and/or antispasmodics to treat them.


Tubeless PCNL was safe and effective even in patients with a solitary kidney, or with three renal access tracts or supracostal access, or with deranged renal values and in those requiring bilateral simultaneous PCNL. The literature review suggested a need for prospective, randomized studies to evaluate the role of fibrin sealant and/or cauterization of the nephrostomy tract in tubeless PCNL.


percutaneous nephrolithotomy


Percutaneous nephrolithotomy (PCNL) is currently the procedure of choice for removing large and complex renal calculi. Placing a nephrostomy tube is the last step after completing PCNL; it provides adequate renal drainage, allowing renal healing and avoiding urinary extravasation. It may also tamponade bleeding and allow the nephrostomy tract to mature and make second-look nephroscopy easier. However, in view of recent developments in the technique of percutaneous access, routine placement of a nephrostomy tube after percutaneous renal surgery was challenged by Bellman et al.[1,2]. In their opinion, placing a JJ stent alone may suffice in most uneventful PCNLs, allowing for a more rapid recovery with less discomfort.

Recently more publications have advocated tubeless PCNL [3–9], but it is still not widely adopted by endourologists, probably because of a relative lack of clinical experience with this technique. It is still considered by many as adventurous and risky. Hence, in a prospective study, we quantified the safety of tubeless PCNL. The series is smaller than that recently reported, but we use it here as a background to review previously reported cases and discuss future developments.


From September 2004 onwards, all patients with large renal and upper ureteric calculi, irrespective of the number and size of stones, amount of hydronephrosis and renal variables, were scheduled to undergo tubeless PCNL. Patients with a solitary kidney, those undergoing bilateral simultaneous PCNL or requiring supracostal access were also enrolled in the study. The exclusion criteria were patients needing more than three percutaneous accesses, presence of significant bleeding that persisted throughout surgery and was not adequately tamponaded by the Amplatz sheath, and the presence of a significant residual stone burden necessitating a staged second-look nephroscopy.

Informed consent was obtained from all patients before considering them for the study. Patients were told that a nephrostomy tube is generally left in place after a percutaneous renal procedure. They were told that in view of emerging reports on tubeless PCNL, a nephrostomy tube would only be placed at the end of procedure if any of the exclusion criteria were met. In patients who were candidates for bilateral simultaneous PCNL, the procedure was carried out first on the side that was symptomatic, obstructed or infected. Otherwise, the side that was relatively easier for PCNL was treated first. Patients were informed that the second side would be attempted only if the first side was accomplished safely and satisfactorily in a reasonable time.

From September 2004 to December 2004, 46 patients had PCNL on 52 renal units. Of the patients initially considered for the study, 40 (45 renal units) qualified for assessment; six met the exclusion criteria and were excluded. Two renal units needed a standard 28 F nephrostomy tube at the end of procedure because there was significant bleeding. In the remaining four renal units with staghorn or multiple renal calculi there was a suspicion of residual fragments at the end of procedure, and the patients were thought likely to benefit from a second-look nephroscopy. These patients had a standard 28 F nephrostomy tube placed at the end of procedure to permit the second-look nephroscopy.

All patients had general anaesthesia and a 6 F ureteric catheter placed transurethrally. Percutaneous access was created by one urologist (H.N.S.) in all cases under fluoroscopic guidance, with the patient prone, using an 18 G access needle into the selected calyx. A 0.89 mm J-tipped guidewire was placed into the collecting system. The access needle was removed and the skin and fascia incised. The nephrostomy tract was dilated with telescoping Alken metal dilators (Karl Storz, Tuttlington, Germany) and a 30 F Amplatz sheath (Microvasive, Natick, MA, USA) positioned into the renal collecting system. The stone was disintegrated using pneumatic lithotripsy. Additional tracts were made as necessary, with the aim of complete stone clearance. After complete clearance was confirmed fluoroscopically and endoscopically, a 6 F JJ stent (Microvasive) was placed in antegradely. In patients with a supracostal access tract, the integrity of the costophrenic angle was confirmed by chest fluoroscopy. On completion of the procedure the Amplatz sheath was removed and the wound strapped with a pressure dressing. The skin incision was left unsutured. In patients who were candidates for bilateral simultaneous PCNL, contralateral PCNL was carried out similarly. A 16 F Foley catheter was left in the bladder at the end of procedure.

Before and on the day after surgery, patients were assessed with a complete blood count to determine the decrease in haematocrit. Patients with supracostal access were given intensive chest physiotherapy and spirometry after surgery, and had a chest radiograph to detect any chest complications. Patients were given a visual analogue scale to analyse pain intensity, and diclofenac sodium when they complained of severe pain. The urethral Foley catheter was removed in all patients after 12–24 h, once the urine cleared. All patients were assessed with renal ultrasonography and radiography before discharge to confirm stone-free status and exclude the presence of urinoma. Non-contrast-enhanced CT was not used for economic reasons. During discharge patients were informed about the possibility of stent-related discomfort and instructed to take 50 mg of diclofenac sodium if the pain or discomfort was bothersome.

Both renal ultrasonography and radiography were repeated after 1–2 weeks during the follow-up before stent removal. The procedure was considered successful if the patient was either free of stones or had no infection or obstruction with residual fragments of < 5 mm in the largest dimension on renal ultrasonography and/or radiography. Patients with larger residual fragments were treated with ESWL before stent removal. After 1–2 weeks the stent was removed cystoscopically under local anaesthesia. During the follow-up patients were given a questionnaire to determine the stent-associated morbidity (Appendix).

The operative time, change in haematocrit, presence of urinary leak, transfusion requirement, analgesic requirement, hospital stay and return to normal activity were analysed, as were complications, success rate and stent-related morbidity.


All 45 tubeless PCNLs in 40 patients were successful, with no significant complications. The patients’ demographic data are shown in Table 1. In all, 51 tracts were required in 45 renal units; four patients needed two tracts and one three tracts in a renal unit during the procedure. Of the 51 tracts, 30 were supracostal. Only two patients required a blood transfusion after surgery; there was no urine leakage or urinoma, or significant chest complication in patients that required a supracostal access tract. The outcomes during and after surgery are also shown in Table 1; 39 renal units were completely stone-free and four had residual fragments of <5 mm. In two patients there were larger residual fragments and these were successfully treated with ESWL before stent removal. In all, 30% of patients had bothersome stent-related symptoms and 60% needed analgesics and/or antispasmodic medication to treat them.

Table 1.  The demographic and clinical data, and the outcomes during and after surgery, of 40 patients undergoing tubeless PCNL
Unilateral: bilateral35 : 5
Number of renal units45
Mean (range) age, years52 (20–78)
Male : female24 : 16
Renal units with multiple stones23
Mean (range) stone size, mm33 (14–70)
Serum creatinine level
 <2 mg/dL (177 µmol/L)37
 >2 mg/dL (177 µmol/L) 3
 Total tracts in all patients51
Number of tracts per renal unit
 Two 4
 Three 1
Outcome, mean (range)
Operative duration, min38 (25–124)
Decrease in haemoglobin, g/dL1.2 (0–6.2)
Analgesia, mg diclofenac40.6 (0–225)
Hospital stay, h26 (18–50)
Outcome, n (%)
Blood transfusion requirement 2 (5)
Overall success rate43 (96)
Stone-free status39 (87)
Residual fragments <5 mm 4 (9)
Re-treatment 2 (4)
Complications, n
Hydrothorax 1
Fever 2


In 1984, Wickham et al.[10] reported their experience with percutaneous stone extraction and stated that in selected cases it is not unreasonable to avoid leaving a nephrostomy tube. However, in 1986, Winfield et al.[11] reported prolonged hospitalization and pain in two patients after percutaneous stone removal when nephrostomy tubes were not left in place. Thereafter, placing a nephrostomy tube at the end of a percutaneous procedure became standard. Since then, with improvements in technique, the morbidity of percutaneous procedures has decreased dramatically [1].

Tubeless PCNL was recently revisited; because of a desire to limit patient morbidity and shorten hospital stays. Bellman et al.[1] proposed that creating a nephrostomy tract is an example of controlled renal trauma, that, given proper drainage, should heal spontaneously. The various case series reported are shown in Tables 2 and 3; 437 patients were reported to have had tubeless PCNL. Most of the publications, including the largest reported series of 138 patients, originate from one institute [1,2,12–15] (Table 3). Patients requiring more than two percutaneous tracts, or having significant obstruction, bleeding, perforation or residual stone burden, were excluded by most authors. Similarly, the average stone burden was smaller in most of the published series. Our hospital is a tertiary referral centre for endourological management of urinary calculi and has a team experienced in percutaneous renal surgery. Hence we expanded our indications for tubeless PCNL and showed that it was safe in patients with larger stones, those requiring three access tracts and in those with significant obstruction. Few authors have avoided tubeless PCNL in patients with a supracostal access tract [6,8] or upper calyceal puncture [7]. In our experience, avoiding a nephrostomy tube is feasible in patients with upper calyceal or supracostal access, without increasing pulmonary morbidity. Four of the present patients had simultaneous bilateral tubeless PCNL, as also shown by others in one case each [16,17].

Table 2.  A summary of previously published case series of tubeless PCNL, with no aids for haemostatic measures (299 procedures)
  1. RCT, randomized controlled trial; RS, retrospective study; PS, prospective study; CS, case series; UC, ureteric catheter.

Year published19841997199819992001200420042004
Number of procedures10036331037301043
Mean stone size, mm1830.6<30250 mm2
Mean operating time, min604575
Postoperative drainage, stentnone32 none; 4 JJJJ6-UC; 4 JJUC × 48 hnoneJJnone
Blood transfusion, %600000
Haematocrit decrease, %1.74.2
Analgesia, mgminimal38.6 meperidine30 pentazocine88 diclofenac9.8 morphine
Hospital stay, days2.821.52.33.631.53.41.6
Success rate, %947994809290
Complicationsbleeding 22%; infection 10%010minor bleeding in 2 patientsinfection in 2 patients
Table 3.  Summary of previously published case series of tubeless PCNL with no aids for haemostatic measures, from one centre (maximum patients reported, 138)
  1. Abbreviations as Table 2; *efficacy assessed in obese patients; †30 council catheter, 20 no nephrostomy tube; ‡discharged with urethral catheter in situ; ¶1.56 days for antegrade endopyelotomy with two patients having gross haematuria; PA, pseudoaneurysm.

Year published19972001200220032004
Type of studyRSRCTCSCSRS*
Number of procedures5081164138
Mean stone size, cm4.4 cm23.3>3.0
Mean operating time, min158130
Postoperative drainage, stentJJJJJJtailed uretericJJ
Blood transfusion, %00606.8
Analgesia, mg morphine11.65.25
Hospital stay, days0.61.91.25<11.82
Success rate, %9394.5
Complications001 PA0

Most authors used a JJ stent or a ureteric catheter for renal drainage [4–6,8], but stent-related discomfort is ignored in all published series of tubeless PCNL. To obtain a true assessment of tubeless PCNL, the morbidity of the nephrostomy tube for 48 h should be compared with that from a JJ stent. In the present study, 30% patients had bothersome stent-related symptoms and 60% needed analgesics and/or antispasmodics to relieve them, but this stent-related morbidity was not compared directly with that of a nephrostomy tube in a standard PCNL. To avoid possible discomfort related to an external ureteric catheter or stent-related morbidity, authors from Iran omitted them in their series of totally tubeless PCNL [8,9]. They hypothesized that the ureter can be used as a drain in selected patients. The procedure was safe and effective in both series, as in other series in which a ureteric stent was placed. The procedures were associated with less discomfort, a shorter hospital stay, and a lower analgesic requirement.

The present challenge is to make a percutaneous renal operation an outpatient procedure, which would be more palatable to patients and more cost-effective, without compromising safety. In the present series, the mean hospital stay was 26 h, with a blood transfusion required in 5% and none with urinary extravasation or urinary leakage after PCNL. These results are comparable with those of Bellman's team [13,15]. To minimize or eliminate the risk of haemorrhage and urinary extravasation after PCNL, newly developed haemostatic agents have been used as an adjuvant to tubeless PCNL [18–22] (Table 4).

Table 4.  Summary of previously published case series of tubeless PCNL with aids for haemostatic measures
  1. Abbreviations as Table 2; *compared with tubeless procedures with no aid for haemostatic measures; †included in addendum.

Year published20032004200420042004
Type of studyRS*CSCSRSRR
Number of procedures202 + 5108451
Mean stone size85.3 cm3Multiple3.37 cm2953 mm22.7 cm
Haemostatic measureTissealFlosealHemoseelCauterization of tractCauterization of tract
Mean operating time, min103127
Postoperative drainage, stentJJJJJJJJ + PenroseJJ + Penrose
Blood transfusion, %0001.20
Haematocrit decrease6.8%2.2 g2.8%
Analgesia, mg morphine13.219
Hospital stay, days1.14<481.13.62.2
Success rateOne stone-free80%83%80%
Complications, %Fever, 10; seroma, 50Fever, 8.4; hydrothorax, 3Fever, 5; prolonged urine leak, 1; delayed bleeding, 1

Mikhail et al.[18] used TisseelTM vapour-heated sealant (Baxter Medical, Fremont, CA) in 20 consecutive patients undergoing tubeless PCNL. They took care to ensure that fibrin glue was not instilled into the collecting system by verifying that the sheath was out of the collecting system. Use of fibrin glue was associated with a significant decrease in the mean hospital stay, but there was no significant difference in the decrease in haematocrit or analgesic use. Similar results were reported by Noller et al.[20] using Hemaseal APRTM (Haemacare Corp., Sarasota, Florida) in 10 renal units.

Lee et al.[19], in a laboratory study, found that gelatine matrix haemostatic sealant formed a fine suspension of particles on contact with urine, whereas fibrin glue created a thicker mucoid material which failed to dissolve 5 days later. Hence they used a gelatine matrix haemostatic sealant (FlosealTM, Baxter Medical) in seven cases of tubeless PCNL. They blocked the point of entry of the nephrostomy tract into the collecting system using a 7 F 11.5 mm occlusion balloon catheter passed retrogradely through the urethra over the ‘through-and-through’ guidewire. The sealant was then injected into the nephrostomy tract with simultaneous withdrawal of the nephrostomy sheath. Kim et al.[23] later reported in an animal study that both Floseal and Tisseel, if injected into the collecting system, can result in obstruction which does not resolve even over 5-days. They recommended using an occlusion balloon to prevent inadvertent injection of these materials into the collecting system.

Jou et al.[21,22] electrocauterized the bleeding points after stone extraction in 249 PCNLs; in 84 patients (34%) it resulted in a bloodless tract, enabling them to avoid a nephrostomy tube at the end of the procedure. They found it was safe and effective in making more patients suitable for tubeless PCNL. In their experience, bleeding frequently originated from the junction of the collecting system and the renal parenchyma.

A prospective, randomized study comparing tubeless PCNL with or without fibrin sealant and/or cauterization is warranted to evaluate further the role and efficacy of fibrin sealant in percutaneous renal surgery. The present report adds information on the safety of tubeless PCNL. In our experience it is safe in patients with a solitary kidney, in those with three renal access tracts or supracostal access, in those with deranged renal values and in those requiring bilateral simultaneous PCNL. This review suggests that the use of fibrin sealant and/or cauterization of bleeding points might help to extend the benefits of tubeless PCNL to more patients undergoing percutaneous procedures.


None declared.



The questionnaire used to assess symptoms, using a modified visual analogue scale.

Dear patient,

We are currently assessing the morbidity of the ureteric stent placed in your ureter after tubeless PCNL. Kindly fill out the enclosed questionnaire for the same. In first three questions we request you to grade the severity of urinary symptoms experienced by you on a scale of 0–5.

1. How much increased frequency of urination did you experience since the time of discharge. (0 being no increase and 5 being a very major increase)

0  1  2  3  4  5

2. How much increased urgency of urination did you experience since the time of discharge. (0 being no increase and 5 being a very major increase)

0  1  2  3  4  5

3. How much pain did you experience (during urination, flank or abdominal) since the time of discharge. (0 being no pain, 1 being minor niggle and 5 being distressing pain)

0  1  2  3  4  5

4. Did you need oral medications (bladder relaxants or analgesics) for the above problem


Total score = 0–15. (Grading of severity of symptoms- 1–5 = mild; 6–10 = moderate; 11–15 = severe).