Percutaneous renal puncture was first described in 1955 by Goodwin and Casey, who placed a trocar directly into the collecting system [2]. Later, the Seldinger method of nephrostomy placement was adopted, in which a fine guide wire is placed into the collecting system through the core of the needle that was used to perform the initial renal puncture. A coaxial catheter can then be placed over this initial guide wire and the renal pelvis drained even if it was not dilated. The addition of a preformed pigtail to these nephrostomy catheters ensured that they could not be easily displaced from the pelvis. In 1976 Fernstrom and Johansson described a method of dilating such an antegrade nephrostomy by use of graded plastic dilators introduced coaxially down the track [3]. After a number of days, the tract was used for intrarenal manipulation utilizing Dormia baskets and other grasping tools. Currently, the percutaneous tract can be dilated by acute dilation with a balloon catheter. Percutaneous nephrostomy (PCN) is, and will continue to be, the cornerstone of every percutaneous procedure of the upper urinary tract.

The topographical position of the kidney depends on its embryological development. Classically, the pelvis lies opposite the lower border of the first lumbar vertebra on the right and slightly higher on the left.

Among numerous factors, enveloping fascia, vascular connections, and intra-abdominal pressure are probably the most important factors holding the kidney in position. Within the renal fascia, the surrounding fat allows a considerable amount of renal movement despite its apparent density, although the kidney is tethered by the short vessels rigidly anchored to their midline connections. Abdominal tone provided by the anterior abdominal wall may be the most important factor for renal stability. The position of the liver limits the cranial movement of the kidney on the right side. The proximity of the pancreas to the anterior aspect of the left kidney is said to be especially important in limiting the movement of the kidney. The suprarenal attachments and ligaments to the liver and duodenum probably do not play an important role in maintaining renal position [2].

Movement of the diaphragm in respiration causes the kidney to move downward in inspiration and upward in expiration. The amplitude of movement is very variable, but it is usually within 3-5 cm. Such movement is more pronounced in women than in men and in the right than in the left kidney. When the patient is in the prone position with bolsters under the chest and upper abdomen, the kidneys are further displaced in a cephalad direction.

The main renal artery divides into two main branches, the anterior and the posterior. The anterior division further subdivides into the four anterior segmental arteries, which supply the anterior and polar areas of the kidney. The posterior segmental artery supplies the rest of the posterior area of the kidney. In more than 50% of kidneys, the posterior segmental artery is located in the middle or upper half of the posterior renal surface, and it may be damaged with an excessively medial needle puncture of an upper calyx. The segmental arteries divide into the interlobar arteries after crossing the renal sinus and become the arcuate arteries at the corticomedullary junction. The interlobular arteries branch off the arcuate arteries at right angles. The Brodel line delineates an avascular plane between the anterior and the posterior blood supplies. By taking a posterolateral transparenchymal path, the needle traverses the area of the Brödel line, and damage to major blood vessels can be avoided. A direct posterior puncture that is too medial risks injury to the posterior segmental artery, which is the artery most commonly injured in endourologic procedures. A needle directed end-on to a posterior calyx passes transparenchymally, and the chance of significant bleeding is minimized.

Although Rupel and Brown first removed a renal calculus through an operatively established nephrostomy tract in 1941 [4], the first PNL via a nephrostomy tract created for the sole purpose of stone removal was performed in 1976 by Fernstrom and Johansson [2]. The introduction of this technique was further refined over the years. As operative technique and endoscopic equipment improved, PNL was performed with increasing efficacy and decreasing complications [5]. PNL has replaced open techniques in removing complex urinary calculi in most institutions. In Korea, the first PNL was performed in 1984 by Koh et al [6]. This operation has now spread to several other institutes in Korea [7-9].

The practice of PNL has changed over time and is continuing to evolve. Differing aspects of the procedure, such as the ideal dilating method, the type of nephrostomy tube used, and the technique of treating calyceal diverticula, have been debated. Even the need for a nephrostomy tube has been questioned.

Endourologic management of ureteropelvic junction (UPJ) obstruction was introduced by Whitfield and Wickham in 1983 as a 'percutaneous pyelolysis' and was popularized shortly thereafter by Smith, who coined the term 'endopyelotomy' [29]. Despite various nuances in the name of the procedure and in the technique performed, the basic concept is constant and involves a full-thickness incision through the obstructing proximal ureter from the ureteral lumen out to the peripelvic and periureteral fat. The incision is stented and left to heal based on the early work of Davis, who used an 'intubated ureterotomy' in the course of an open operative procedure for UPJ obstruction [30]. Smith introduced endopyelotomy with a live surgery workshop to Korean doctors in 1987. After that, endopyelotomy started to be reported in papers from Korea [31,32].

Contraindications to a percutaneous endopyelotomy are similar to the contraindications to any endourologic approach and include a long segment (>2 cm) of obstruction, active infection, or untreated coagulopathy. The impact of crossing vessels is controversial [33-36].

Compared with the retrograde techniques of endopyelotomy (incision with a cold knife, Acucise catheter, Greenwald electrode, or laser), the antegrade technique offers the advantage of an incision under direct vision. The incision must be extended into the perirenal fat and into the healthy ureter. Although several clinicians suggest that the incision should always be made laterally, in fact, the ureter may be inserted into the renal pelvis on the anterior or posterior wall. In such cases, the incision should instead marsupialize the proximal ureter into the renal pelvis such that an anterior or posterior incision may be required [30]. Percutaneous endopyeloplasty, horizontal percutaneous suturing of a conventional longitudinal endopyelotomy incision, was recently developed with good clinical results. The technical simplicity and shorter operative time are advantages compared with laparoscopic pyeloplasty [37,38].

Once the incision is complete, stenting is accomplished. A 14/7 Fr endopyelotomy stent can be used and is passed in an antegrade fashion with the large diameter end of the stent positioned across the UPJ. There was a trend for better results with the use of a 14/7 Fr stent in patients with secondary stricture, although the difference in success rates between the 6 Fr and the 14/7 Fr stent was not significant statistically [39].

The immediate and long-term results of percutaneous endopyelotomy are well established. Clearly, percutaneous endopyelotomy compares favorably with open operative pyeloplasty in terms of postoperative pain, the length of hospital stay, and the return to prehospitalization activities [30,40]. Currently, success rates approaching 85% to 90% are reported at experienced centers (Table 1). Little difference in outcome is noted between primary and secondary UPJ obstruction and no difference in methods of incision.

Laparoscopic pyeloplasty has recently been reported, with success rates in excess of 95% [43]. Moreover, laparoscopy and robot-assisted surgery can be applied in patients with severe hydronephrosis requiring pelvic reduction and in patients with crossing vessels that may require ureteral-vascular transposition. However, the steep learning curve in laparoscopic surgery or the economical condition may limit laparoscopic or robot-assisted pyeloplasty to select centers proficient in reconstructive laparoscopy or robotic surgery.

A calyceal diverticulum is a smooth-walled, nonsecretory cavity in the renal parenchyma that is lined with transitional cell epithelium. It receives urine by passive retrograde filling from the adjacent collecting system, usually through a narrow forniceal channel or infundibulum. Calyceal diverticula are believed to be congenital in origin, likely from failed degeneration of small ureteral buds. They are typically less than 1 cm in diameter with no predilection for sex or kidney side. Uncomplicated, asymptomatic calyceal diverticula may be managed conservatively without routine follow-up imaging. However, because of their cystic, urine-containing nature, they are frequently associated with stone formation and infection and become symptomatic in up to one third of patients [44].

Treatment of a calyceal diverticulum containing a stone has evolved from open surgical excision to SWL, to percutaneous and ureteroscopic ablative techniques. The preferred technique continues to be debated. Percutaneous management of the calyceal diverticulum is challenging because the cavity is often small, making localization for direct access difficult. The first Korean paper was reported in 1994 and the percutaneous process was mentioned [45]. Cystoscopy and ureteral balloon catheter placement is performed in the renal pelvis. A balloon catheter is helpful to opacify the diverticulum with injection of contrast through the ureteral catheter to guide the percutaneous access, especially when the neck of the diverticulum is narrow. Direct puncture of the diverticulum is then made under fluoroscopic guidance, and a guide wire is coiled within it. Ideally, a polytetrafluoroethylene-coated or hydrophilic safety wire is placed through the diverticular neck into the renal pelvis, but it is usually coiled in the diverticular cavity because the neck cannot be cannulated. Canales and Monga advocate dilation of the tract into the diverticulum, although not through the diverticular neck because the goal of the procedure is to ablate the cavity and the connection to the collecting system [44]. Dilation of the diverticular infundibulum could be viewed as counterproductive. Usually, one can find the diverticular neck with the nephroscope during the injection of indigo carmine through the ureteral balloon catheter. Auge et al described an alternative approach if guide wire passage into the main collecting system was unsuccessful after several attempts [46]. Once inside the cavity, they advance an 18 gauge percutaneous access needle directly through the inner or medial diverticulum wall into the renal collecting system and subsequently dilate to 30 Fr with a dilating balloon, creating a large 'neoinfundibulotomy' tract. This maneuver prevents the safety wire from being inadvertently withdrawn. With this technique, the connection between the diverticulum and the collecting system is enlarged rather than ablated. But in my experience, I never find a diverticulum without a diverticular neck. The urothelium of a calyceal diverticulum is usually fulgurated with electrocautery or a holmium laser if greater than 4 cm in diameter. If electrocautery is utilized, the safety wire should be insulated with an open-ended catheter to prevent inadvertent transmission of the current down the ureter. The nephrostomy catheter is placed through the calyceal diverticulum and neoinfundibulotomy and secured in the renal pelvis. There are controversies about the duration of a nephrostomy catheter. But it tends to be shorter because there is no difference in success rates according to nephrostomy catheter duration [44,46,47]. The results of percutaneous management of calyceal diverticula from the literature are presented in Table 2. In the cases reviewed, stone-free rates and symptom-free rates for percutaneous management are consistently 80% or greater. Minor complications during percutaneous ablation and calculus removal include hemorrhage, pneumothorax, persistent urinary extravasation, and mild extravasation of irrigant. Major complications include renal pelvis perforation with urinoma formation, pneumothorax or hemothorax requiring tube thoracostomy, and massive hemorrhage requiring balloon tamponade. As the table demonstrates, major complications are relatively uncommon.

Limitations exist primarily for an anteriorly located diverticulum. In this situation, if the diverticulum is in a superior anterior calyx, the ureteroscopic approach is recommended, whereas if the diverticulum is in a middle or lower anterior calyx, a laparoscopic approach is recommended [44]. The ureteroscopic approach may be an appropriate initial treatment option for patients with a small stone burden (<1.5 cm) or patients with comorbidities who are poor candidates for PNL.

Infundibular stenosis and hydrocalyx are usually an acquired condition associated with inflammation, renal tuberculosis, obstructive calculus, or prior renal surgery [18,51]. The hydrocalyx should be differentiated from a calyceal diverticulum because the treatments are different. At times, this distinction can be made only by a nephroscopy because the presence (hydrocalyx) or absence (calyceal diverticulum) of a renal papilla is diagnostic. The infundibular narrowing can be resolved in several ways. The least difficult approach is to dilate the infundibulum to 8 mm with an 8 mm ureteral dilating balloon passed over the working guide wire. Alternatively, the infundibulum can be cut under endoscopic control with a cold knife through a direct vision ureterotome. When the guide wire cannot be passed through the stricture, a round tipped rigid ureteroscope can be pushed in an antegrade fashion to traverse the stricture with injection of indigo carmine through the retrograde ureteral catheter (Fig. 4). According to anatomic studies by Sampaio, the incision should be made along the less vascular superior and inferior aspects of the middle calyceal infundibulum or the medial and lateral aspects of the upper calyceal infundibulum [52]. Reported series of endourologically treated infundibular stenosis are few. The first Korean case report of endoscopic infundibulotomy was published by Han et al in 1991 [53]. Also, Lang reported a 50% success rate in 6 patients with infundibular stenosis and caliceal diverticula containing stones [54]. Hwang and Park reported an 80% success rate in 10 patients with tuberculous infundibular strictures who had undergone a cold knife incision; follow-up was greater than 1 year [51]. Hwang et al reported long-term (more than 2 years follow-up) results with a success rate of 76% in 21 patients and better results in the strictures with stones than in the strictures with tuberculosis [55]. It appears that in contrast to the calyceal diverticulum, in which a successful outcome is obtainable in nearly 90% of cases, the infundibular stenosis is a more difficult entity to treat endourologically, with only a 50% to 76% success rate (Table 3).

Nephroureterectomy with bladder cuff excision has been considered the "gold standard" for upper-tract transitional cell carcinoma (TCC). In upper-tract TCC, endoscopic surgery could be indicated in those patients with a solitary kidney, bilateral disease, high surgical risk, or chronic renal insufficiency and incidental discovery of the tumor [57,58].

The smooth muscle covering the ureter and renal pelvis is much thinner than the bladder wall. Therefore, tumors in the ureter and renal pelvis can penetrate the wall earlier. Indeed, approximately 70% of upper-tract TCCs were of moderate or high grade, and more than half had some degree of invasion [59].

Endoscopic surgery (ureteroscopy and percutaneous nephroscopy) has been used in the treatment of upper-tract tumors since the 1980s [60-62]. Endoscopic surgery may be considered for those patients with low-grade, low-stage cancer malignancies under an intensive surveillance program. Percutaneous approaches show recurrence rates and disease-specific survival in the patients with low-grade tumors (Grades 1-2) of 26% to 28% and 96% to 100%, respectively [57]. Lee et al compared patients undergoing nephroureterectomy versus percutaneous resection and noted no statistically significant differences in survival for either treatment group (Grades 1, 2, and 3) [63]. Tumor grade was the most important prognostic indicator for pathologic stage, recurrence rates, and cancer-related deaths.