To assess the safety and outcome of paediatric percutaneous nephrolithotomy (PCNL) for atypical cases and compare the results with ‘standard’ unilateral paediatric PCNL.
To assess the safety and outcome of paediatric percutaneous nephrolithotomy (PCNL) for atypical cases and compare the results with ‘standard’ unilateral paediatric PCNL.
We retrospectively reviewed children who had had a PCNL between December 1997 and December 2004. Patients were grouped as follows: group 1, aged >5–16 years with normal anatomy and normal renal function undergoing unilateral PCNL or staged bilateral PCNL; group 2, ≤ 5 years with normal anatomy and renal function undergoing unilateral PCNL; group 3, undergoing bilateral simultaneous PCNL; group 4, impaired renal function in addition to renal stone disease; group 5, renal anatomical abnormality with calculi in the same kidney. Demographics, stone profile, procedure and outcome indicators were analysed for each group.
In all, 188 consecutive PCNLs in 169 children were included (mean age 3.3–10.3 years, mean stone burden 19.1–33.3 mm in the five groups). The mean duration of PCNL was 69–115 min. Stone clearance was satisfactory with single tract access in 90–100% of patients. Transient postoperative fever was the commonest complication (12.5–51%) followed by hyponatraemia and hypokalaemia. Blood transfusion was required in 0–7.7%. The mean stone clearance rates were 47–90% in the five groups; additional extracorporeal shockwave lithotripsy increased the cumulative clearance rates to 90–100%.
PCNL is safe for treating renal stones, with excellent results and minimal complications. Comparable results are achieved in the very young child, children with anatomically abnormal kidneys, children with impaired renal function and children with bilateral renal stones undergoing simultaneous bilateral PCNL. Hence none of these factors should be considered as relative contraindications.
The prevalence of stone disease in developing countries is 5–15%. Children commonly develop stones in multiple sites and recurrent stones are common. In this scenario, minimally invasive methods of treating stone disease become more important. Since the introduction of percutaneous nephrolithotomy (PCNL) in 1976, developing technology and techniques have made it possible to use this procedure in increasingly challenging situations. PCNL was used in children for the first time in 1985; from the cautious optimism in early reports , to several recent series citing good results with this procedure in children, paediatric PCNL has become a routine procedure. However, there are certain features that can potentially cause technical difficulties. Very young children, children with bilateral calculi, stones within functional or anatomical renal abnormalities at presentation are all groups that require special attention. We recently showed that PCNL was safe and effective in pre-school children ; we now add to that series and assess the groups described above, to compare complications and outcome with ‘standard’ unilateral PCNL.
We retrospectively reviewed all medical and radiographic records of children aged ≤ 16 years undergoing PCNL at our institution from December 1997 (when PCNL was first used at our centre) to December 2004. Demographic features, stone profile, presentation and history, family history, investigations, procedure details, ancillary procedures, complications, outcome and follow-up were documented on a standardized data-collection form.
The children were divided into five groups as follows: group 1, aged >5–16 years with normal anatomy and normal renal function, undergoing unilateral PCNL or staged bilateral PCNL. In the latter case, each renal unit was analysed as undergoing an individual operative procedure, while the demographics were included only once per patient in the data set. This group was taken as the ‘standard’ group against which the results of all other groups were compared; group 2, aged ≤ 5 years with normal anatomy and renal function undergoing unilateral PCNL; group 3, undergoing bilateral simultaneous PCNL; group 4, impaired renal function in addition to renal stone disease. A serum creatinine level of ≥ 1 mg/dL was arbitrarily taken as the threshold. As height was not recorded routinely in the patients it was not possible to calculate impaired renal function on the basis of GFR nomograms for each individual; group 5, renal anatomical abnormality (horseshoe kidney, bifid pelvis and duplex kidney) with calculi in the same kidney.
All children were evaluated and operated on by one surgeon. The preoperative evaluation, method used for calculating stone burden, criteria for selecting the patients for PCNL, operative technique and postoperative course were described previously . In those undergoing simultaneous bilateral PCNL a procedure similar to that for unilateral PCNL was adopted, except that initially a retrograde pyelogram was taken on both sides, in the lithotomy position. The child was then turned prone and PCNL access with stone clearance or de-bulking obtained sequentially for both sides. A percutaneous nephrostomy tube was left in situ after the procedure in all patients. Procedure and outcome indicators for individual renal units are cited in the results.
After PCNL, each child was reviewed by a dietician and given nutritional advice, as dictated by the stone analysis results. Complications and outcome in groups 2–5 were analysed by comparison with the ‘standard’, group 1, using the chi-squared test to estimate odds ratio and Yates’ corrected P values.
In all, 169 children undergoing PCNL in 188 renal units at our institute during the study period were included in this series. Using the defined criteria, patients were grouped as shown in Table 1. Two children with horseshoe kidneys were also aged <5 years and were included in both groups 2 and 5. Similarly, two children with bilateral stones in horseshoe kidneys had PCNL in both renal units, one simultaneous and the other staged; they were included in both groups. Ninety children undergoing unilateral PCNL and seven undergoing bilateral staged PCNL on different admissions were included in group 1, thus bringing the number of renal units in this group to 104. In group 5, six children had horseshoe kidneys, of whom two had renal calculi in both renal moieties. One child had a complete duplex system and another had a bifid pelvis.
|Characteristic or variable||Group|
|Number of patients||97||43||12||13||8|
|Number of renal units||104||43||24||13||10|
|mean (median)||10 (10.6)||3.3 (3.3)||7.3 (7.2)||10.3 (10.2)||9.8 (13.6)|
|Documented UTI, n (%)||20 (21)||13 (30)||3||7||2|
|History of stone disease, n||16||3||1||2||0|
|Synchronous stone(s) at other sites, n||24||16||0||6||2|
|Family history of stone disease, n||34||7||0||2||0|
|Mean stone burden, mm||31.4||20.5||19.1||26.3||33.3|
|Number of calculi|
|Mean duration, min||93||80||57||69||115|
|Mean fluid, L||13.5||6.6||6.5||13.1||7.5|
|Single tract access, n (%)||99 (95)||42 (98)||24 (100)||13||9|
|Access site, n (%)*|
|lower pole||71 (68)||32 (75)||19 (79)||11||0|
|upper pole||23 (22)||10 (23)||5 (21)||1||9|
|Scope used, n (%)|
|17 F||59 (57)||42 (98)||21 (87)||6||5|
|26 F||45 (43)||1 (2)||3 (13)||7||5|
|Complications, n (%)|
|Fever||53 (51)||17 (40)||6||6||1|
|Transfusion||3 (3)||2 (5)||1||1||0|
|Na+≤ 130 mmol/L||6 (6)||3 (7)||0||1||0|
|K+≤ 3.0 mmol/L||3 (3)||4 (9)||1||1||0|
|Total||65 (67)||26 (60)||8||9||1|
|PCNL monotherapy, n (%)||49 (47)||29 (67)||19 (79)||9||9|
|Additional ESWL, n (%)||47 (45)||13 (30)||3 (13)||4||0|
|Cumulative clearance, n (%)||96 (92)||42 (97)||22 (92)||13||9|
A significant proportion (26%) of the children (Table 1) either had a history of UTI confirmed on culture, or had a positive culture at the time of admission for PCNL. Antibiotic therapy, as dictated by the sensitivity pattern, was initiated. In all, 22 children (13%) had a history of stone disease; they had either had surgery for stone disease or had spontaneously passed stones urethrally. Synchronous stone in one or more sites were found in 47 (28%) children; synchronous stones of <1 cm in the longest axis, not impacted at the PUJ and not causing gross hydronephrosis were treated with ESWL. Such synchronous stones were found in the contralateral kidney in 26 children (15%). There were calculi of varying sizes in the bladder (10, 6%), unilateral or bilateral ureters (23, 14%), and urethra (one, 0.5%). The stone profile for each group is given in Table 1; cumulatively, 40% of children had one calculus while 60% had more than one renal stone.
The duration of PCNL was calculated from the time of induction to the time of recovery from general anaesthesia. Thus, it includes the time taken for a retrograde pyelogram, repositioning and PCNL per se. Those instances where additional procedures (e.g. circumcision, ureterolithotomy) were performed under the same anaesthesia were excluded from this calculation. In group 3, the mean duration of surgery was halved, to give the mean duration per renal unit. The mean duration in the five groups was 57–115 min (Table 1).
In 90–100% of patients one access tract was sufficient to achieve adequate stone clearance (Table 1). Lower pole access was used most often but where necessary upper pole access was used with no undue problems. In group 2, a 17 F nephroscope was used almost exclusively (in 42 children). In older children a 26 F nephroscope was used relatively often.
The mean hospital stay was 4.5 days, which included admission to hospital one day before PCNL. Patients were followed for a mean (range) of 30.2 (4.6–82.7) months, and complications and outcomes during this time were evaluated. Complications were compared among the groups (Table 1); fever was the commonest in all groups, but blood and urine cultures were negative and fever settled spontaneously in 24–48 h. Several children requiring a preoperative blood transfusion for nutritional anaemia were not included in Table 1; only those requiring a transfusion subsequent to blood loss during PCNL were listed. In all, blood was transfused in seven children undergoing PCNL in nine renal units, within 24 h of surgery. The endoscope used (17 F in six and 26 F in three renal units) and the mean duration of the procedure (85 min) in this group were comparable with the overall results. In one initial PCNL, reverse-osmosis water was used and the child developed symptomatic hyponatraemia at a serum sodium level of 122 mmol/L; he responded well with conservative management. In all subsequent cases, normal saline was used, and serum electrolytes returned to normal within 48–72 h of surgery.
In addition to the complications listed in Table 1 a child in group 1 developed ureteric obstruction secondary to a residual fragment, for which he had ureterorenoscopy and fragmentation. UTIs confirmed by culture were documented in four children from group 1, and one each in groups 2, 4 and 5. There was bleeding from the nephrostomy tract in two children in group 1 and one in group 4. There were no significant differences in the overall complication rates of groups 2–4 when compared with group 1; group 5 had significantly fewer complications than group 1 (P < 0.05).
There was complete stone clearance on monotherapy with PCNL in 47–90% of renal units (Table 1). With additional ESWL, the cumulative clearance rates were 90–100%. There were fine fragments of <2 mm in nine renal units (six in group 1 and one each in groups 2, 3 and 5). Residual fragments in two renal units (one each in groups 1 and 3) were found and scheduled for ESWL, but the patients failed to appear for the procedure and were thus documented as being lost to follow-up. One child in group 1 continues to have a 6-mm fragment despite multiple sessions of ESWL after PCNL; he remains under close surveillance. The odds ratio and P values are shown in Table 1; all P values were significant (<0.05) for all groups except group 4. However, when combining ESWL to assess cumulative clearance rates, the P values for all groups were not significant. Recurrent stones developed in seven children during the follow-up period.
Stone analysis reports were available for 155 of 169 children. Stones were most commonly composed either entirely of calcium oxalate (66, 43%), or with additional urate and/or phosphate (77, 50%). The remaining stones comprised variable combinations of calcium, oxalate, ammonium, urate and phosphate. There was only one cystine stone.
In group 4 (13 children) nine had acute renal insufficiency secondary to stone disease, with serum creatinine levels returning to normal once the obstructive uropathy was relieved. In two children there was a decrease in serum creatinine level after the procedure, although not to normal levels, and mild renal impairment persisted. There was no change in one child, whereas one had a transient increase in serum creatinine level, this returning to baseline within a week of surgery.
The prevalence of renal stone disease is alarmingly high in Pakistan , as in other countries belonging to the ‘stone belt’. The contributing factors that have been postulated are suboptimal fluid intake, dietary imbalances, metabolic factors and chronic diarrhoea . Recent reports indicate changing patterns, with renal stones becoming more common than bladder stones . Paediatric stone disease is even more worrying; young children frequently present late with renal impairment that significantly affects their life-expectancy and quality of life. The recurrence rate for stone disease increases over time , and children are accordingly at the greatest risk. Therefore, individuals forming stones at when very young require the closest attention, with the emphasis on the best possible clearance rates with the least invasive treatment. A search for, and management of, predisposing factors is an important part of treatment [4,5].
The introduction of minimally invasive methods of treating stone disease has caused a dramatic change in treatment patterns over the past three decades. As compared to open surgical procedures, ESWL and PCNL are less invasive, have lower complication rates and a significantly shorter convalescent phase [6,7,8]. In addition there is a significant reduction in the cost of treatment . Economics are an important consideration in developing countries with a high burden of stone disease. ESWL has the advantage of being less invasive than PCNL, although in children this procedure requires general anaesthesia. However, certain stone and patient factors limit the success rates of ESWL; larger stones, complex or multiple stones, large lower pole calculi, cystine stones and kidneys with abnormal anatomy are difficult to treat with ESWL alone [9,10]. In a series reported by Lim et al. ESWL was unsuccessful for multiple calculi, or single calculi of >10 mm in children. Where the stone burden is larger and multiple sessions of ESWL are predicted, the possible risk of renal damage must be considered , in addition to the time required and economic repercussions . A large proportion of our patients come from distant areas, therefore multiple sessions of ESWL are often not a feasible option. Moreover, in large stones stenting is required, which itself precludes and delays stone clearance.
PCNL has reported clearance rates of ≥ 90% even with complex calculi in children [13–15]. Some authors have reported clearance rates of 50% with staghorn stones and with anatomical abnormalities, but in our experience and that of others [10,16,17], good results can be obtained with increasing expertise and in combination with ESWL. In the present series, there was complete clearance in 47–90% in all groups with PCNL monotherapy. With additional ESWL, clearance rates increased to 90–100%. The results for group 1 were lowest using PCNL alone. This may reflect late presentation and advanced stone burden when only one kidney is affected. Different methods for calculating stone burden have been used by different authors . Due to this variability in the method of calculation, results are often not truly comparable. The sum of the longest axis of each stone as seen on a plain abdominal X-ray gave an approximate stone burden in the present patients [1,10].
Importantly, many children in the present series had synchronous stones at other sites in the urinary tract, or had stones previously that had either passed spontaneously or required intervention. A history of documented UTI was given by 25% of the children and a quarter of all children gave a history of stone disease in first-degree relatives. Seven children developed recurrent stones after initial complete clearance, even in the short mean follow-up of 30 months. Higher recurrence rates were reported in the presence of pre-existing anatomical or metabolic abnormalities [5,11]. We have come to rely on PCNL as the mainstay of treatment, as there is a high possibility that affected children may undergo many procedures during their lifetime. Indeed, since introducing this procedure for children at our institution in 1997, we have not needed to use open pyelolithotomy. At the same time, we follow an aggressive policy of aiming to achieve complete stone clearance and do not accept residual fragments of any size. Even a tiny residual fragment can act as a nidus for stone growth, which places the individual at higher risk of recurrent stone formation [4,5].
When there is an underlying abnormality or the child has had previous open surgery on the same kidney, the less-invasive ESWL and PCNL make subsequent operations less hazardous . Reports in adults have shown good clearance rates in previously operated kidneys . In the present series, six children had previously had open pyelolithotomy in the same renal unit that was treated by PCNL. We encountered no undue technical difficulty or complications in these children.
In the earlier stages of PCNL in children there were some reservations about risks to renal reserve and technical difficulties in young children . However, a recent report  from our centre showed that PCNL is safe and effective in children aged ≤ 5 years. The success of PCNL depends on meticulous technique and experience. Inevitably, experience and time are required, with improving results over time [8,14].
PCNL is unlikely to give rise to scarring [18,19]; in the present series of 51 children with DMSA scans taken 2–3 months after PCNL (isotope scans were not taken before PCNL in most children), scarring was possibly attributable to PCNL in only three cases (unpublished data). Adult-sized nephroscopes have been used in children, with acceptable complication rates and little or no evidence of scarring or sequelae [15,17,20,21]. In contrast, Gunes et al. reported a significant increase in complications in children aged <7 years when using adult-sized equipment. Regardless, it is reasonable to try to limit the size of nephroscope to the smallest possible that does not compromise the outcome. In recent years we have increasingly used a 17 F nephroscope; especially in children aged ≤ 5 years this nephroscope was used in all but one; the single use of a 26 F nephroscope was necessitated by a technical fault with the 17 F instrument noted at the time of the procedure.
In the initial period patients requiring bilateral PCNL had a staged procedure on two different admissions several weeks apart. With increasing experience we now routinely perform bilateral simultaneous PCNL, even with larger stone burdens. The results are equally good, with high patient satisfaction. A comparable series of children undergoing bilateral simultaneous PCNL was recently published by Salah et al.; they reported similar satisfactory outcomes, with the added benefit of decreased healthcare costs.
PCNL has been used successfully in patients with renal failure secondary to stone disease  with little deleterious effect on the renal function. Indeed, a significant improvement was documented for impaired renal function with the removal of stones and clearance of infection [19,24,25], regardless of the treatment adopted.
PCNL has been used successfully in horseshoe kidneys  with certain caveats; upper or middle pole puncture is preferred, and a more posterior puncture is warranted. We used upper pole access in nine and middle pole access in one renal unit with abnormal anatomy, with good results.
UTIs were documented in 21–54% in the present groups; in our experience it is not always possible to eradicate the infection completely before PCNL. With good antibiotic cover, the risk of complications was no greater. A high incidence (29%) of postoperative pyrexia of >38.0 °C was reported previously . If the increase in temperature is transient, with no clinical evidence of bacteraemia, no bacteriological evaluation is required .
More than in adults, PCNL in children requires a good understanding of renal anatomy and meticulous technique to minimize blood loss and transfusion [28,29]. As a result of painstaking attention to detail, we have not needed to convert one PCNL to an open procedure since introducing this technique at our institution.
In conclusion, PCNL is safe for treating renal stones in children, with excellent results and minimal complications. Results are comparable in the very young child, children with anatomically abnormal kidneys, those with impaired renal function and those with bilateral renal stones undergoing simultaneous bilateral PCNL. Hence none of these factors should be considered as relative contraindications.