Abstract
Introduction:
Noncontrast computed tomography (CT) is commonly utilized after percutaneous nephrolithotomy (PNL) to assess stone-free (SF) status. In addition to assessing SF status, CT is useful in the recognition of complications after PNL. We characterized complications demonstrated by postoperative CT scan and compared hospital re-admission rates based on whether or not CT was performed.
Methods:
We retrospectively reviewed records of 1032 consecutive patients from April 1999 to June 2010. Patients were divided into two cohorts based on whether they had a CT within 24 hours of PNL. Demographic data, CT findings, and need for re-admission for complication management were assessed.
Results:
Nine hundred fifty-seven patients (92.7%) underwent post-PNL CT. CT-diagnosed complications were perinephric hematoma in 41 (4.3%; 2 requiring embolization and 9 necessitating transfusion), pleural effusion in 25 (2.6%; 10 requiring intervention), colon perforation in 2 (0.2%), and splenic injury in 2 (0.2%). Of patients with postoperative complications, 33% required intervention. Among patients with a CT, 6 (0.6%) were readmitted despite negative postoperative CT (four perinephric hematomas, one calyceal-pleural fistula, and one pseudoaneurysm). The sensitivity of CT for diagnosing complications was 92.7%. Seventy-five patients (7.3%) did not undergo CT post-PNL. Of these, four (5.33%) were readmitted: three for perinephric hematomas and one for ureteral clot obstruction. Patients undergoing post-PNL CT were less likely to be readmitted because of missed complications (p=0.02).
Conclusions:
Serious post-PNL complications are uncommon, but their prompt diagnosis and treatment is imperative. In addition to identifying residual stones, CT is useful in diagnosing postoperative complications. Postoperative CT could potentially be considered for all patients undergoing PNL, particularly in complex cases such as patients with anatomical abnormalities (renal anatomic abnormality or retrorenal colon), patients requiring upper pole access (risk of thoracic, hepatic, and splenic complications), and patients requiring multisite access (higher risk of perinephric hematoma or need for transfusion).
Introduction
Although serious immediate complications after PNL are not common, their prompt diagnosis and treatment is important. 7 The most common significant complication of PNL is hemorrhage. Transfusions after PNL are necessary in 3% to 11% of cases. The need for multiple access sites also increases the transfusion risk (as much as 40% in one study). 8 Thoracic complications are another important complication after PNL, particularly when supracostal access is obtained. These complications include atelectasis, pleural effusion, pneumothorax, hemothorax, and hydrothorax. Supracostal puncture is associated with pulmonary morbidity in approximately 16% of cases. 7 Other serious complications of PNL include injuries to hollow viscera and trauma to the liver or spleen. Hollow viscera injury is more common in patients with renal anatomical abnormalities or retrorenal colon, while hepatic or splenic injuries are more likely to occur when a supracostal or upper pole access is used. 9
A potential benefit of CT after PNL is the ability to detect postoperative complications. However, the incidence of procedure-related complications diagnosed by CT and the sensitivity of CT for diagnosing such complications is not known. The goal of this report is to characterize the immediate postoperative complications diagnosed by CT, to document any missed complications, and to assess the overall accuracy of CT for diagnosing complications.
Materials and Methods
Study population
Our institutional review board-approved database was utilized to identify patients undergoing PNL for renal calculi at Indiana University Health Methodist Hospital between April 1999 and June 2010 (IRB# 03-096). Patients were divided into two cohorts: the first group consisted of patients who underwent a postoperative day 1 CT and the second group included patients who did not have a CT during their hospitalization. Only patients with CT performed during their initial hospitalization were included in the first group. Of 1032 consecutive patients undergoing PNL, 75 (7.26%) did not undergo CT imaging postoperatively. If the patient undergoing PNL had limited stone burden and, particularly if intracorporeal lithotripsy was not required during the procedure, CT was not performed, as it was judged that the likelihood of residual fragments was extremely low under this set of circumstances.
Operative technique
After cystoscopic placement of a ureteral catheter, percutaneous access was performed by the urologist as part of a single-stage procedure. Access was performed using an 18-gauge needle and bi-planar fluoroscopy. The nephrostomy tract was dilated to 30F in all cases with a NephroMax balloon (Boston Scientific Corp., Watertown, MA), and a 30F Amplatz sheath was utilized during all PNL procedures. Rigid and flexible nephroscopy were performed in all cases and the intrarenal collecting system was mapped in an effort to render the patient SF at the time of the initial procedure if possible. In most cases, a 10F cope loop nephrostomy tube (Cook Urological, Spencer, IN) was utilized for drainage after PNL. Following complicated PNL cases, as we have reported elsewhere, 20F re-entry Malcot catheters or 18F circle nephrostomy tubes were occasionally used. 10 If the PNL procedure involved supracostal access, the lung fields were imaged fluoroscopically at the conclusion of the procedure. If fluoroscopy was suspicious for hydrothorax or pneumothorax, patients underwent further imaging in the postanesthesia care unit with an upright portable chest X-ray. Access was considered supracostal if at least one puncture was supracostal. Patients in whom residual fragments were identified on post-PNL CT underwent second-look nephroscopy typically within 48 hours after the initial procedure.
Postoperative evaluation and follow-up
Typically, on the morning of postoperative day 1, patients underwent CT and these images were reviewed by the operating surgeon, as well as by a staff radiologist to assess for SF status and complications. The primary endpoint of the study was a comprehensive descriptive report of postoperative complications diagnosed by CT. Complication data were obtained from radiology reports and patient records. A standardized definition of each complication was used.
Patients were typically seen in the office for follow-up 6 weeks after PNL. Complications after discharge, such as emergency room visits, need for further hospitalization, or other postoperative treatment was documented and used for comparison between groups. Any complication-related event, defined as need for emergency room visit, re-hospitalization, or additional intervention specifically aimed at alleviating symptoms associated with a complication was reported.
Statistical analysis
A two-tailed Student's t-test, Chi-square test, or Fisher's exact test was utilized as appropriate. A p-value<0.05 was considered statistically significant.
Results
Descriptive data
1032 consecutive patients undergoing PNL were available for analysis. The mean age of the study group was 52 years (range 2–90). Average body–mass index (BMI) was 30.8 (range 13–75). CT was performed postoperatively on 957 patients (92.7%) (Table 1). Complications diagnosed by CT included perinephric hematoma in 41 (4.3%). Two required embolization and nine required blood transfusions (range 2–10 units, mean 2.94 units). Thoracic complications were diagnosed in 26 (2.71%), 10 of who required treatment. There were two colon perforations, with one requiring abdominal exploration and the other requiring placement of a retroperitoneal drain. There were two instances where the percutaneous access traversed the spleen, neither of which required additional treatment (Table 2). There were six patients (0.6%) who had no evidence of complications on CT post-PNL but required re-admission because of perinephric hematoma (four patients), hydrothorax (one patient), or pseudoaneurism (one patient). Thus, the sensitivity of CT for diagnosing post-PNL complications was 92.7%.
Mean values are presented.
CT=computed tomography; BMI=body–mass index; PCNL=percutaneous nephrolithotomy.
There were 75 patients (7.26%) who did not have a CT postoperatively. Patients who did not have CT performed after PNL tended to be a less complex group of patients with smaller stones, a lower BMI, and a lower rate of ectopic or renal fusion anomalies (Table 1). Among them, four (5.33%) required readmission: three because of a symptomatic perinephric hematoma and one as a result of intractable pain secondary to ureteral clot obstruction that necessitated insertion of an internal ureteral stent. No death was documented in either group.
Comparison between groups
The two groups were comparable in terms of age, prevalence of ectopic kidneys, and the prevalence of limb contractures or significant skeletal abnormalities (Table 1). BMI was significantly higher for patients who had a postoperative CT (p=0.04). Also, patients who had post-PNL CT had a larger stone burden than those who were not imaged after PNL (Table 1). Despite higher BMI and stone burden, patients who had post-PNL CT were less likely to be readmitted as a result of missed complications (p=0.02).
Discussion
Many urologists employ CT routinely post-PNL owing to its accuracy in assessing stone-free status. Pearle and colleagues 4 first reported that CT was equally as sensitive as flexible nephroscopy for detecting residual fragments after PNL. They reported that the routine use of CT after PNL reduced the need for second look nephroscopy in patients with large calculi from 32% to 12%.
Recently, however, the routine practice of post-PNL CT has been challenged. Osman et al 6 prospectively reviewed the role of different imaging modalities in 100 renal units undergoing PNL. A significant residual stone was defined as more than two 5-mm calyceal fragments. KUB and renal ultrasonography were able to detect 85.7% and 48.6% of significant residual fragments, while the sensitivity of noncontrast CT was 100%. The authors concluded that because of its expense CT should not be routinely performed postoperatively for radio-opaque stones.
An important additional benefit of unenhanced CT after PNL is its ability to detect postoperative complications promptly. Semins and associates studied CT findings in 197 patients undergoing PNL. 11 All patients underwent CT within 24 hours post-PNL. Atelectasis was documented in 88 patients (44.7%), 23 patients (11.6%) had other thoracic complications, and 15 patients (7.6%) had perinephric hematoma. There was one splenic injury. No injuries to hollow viscera were detected. The authors concluded that CT should be obtained post-PNL.
In the current study, we demonstrated the accuracy of CT in diagnosing complications with sensitivity of 92.7%. There were six patients who had a negative CT post-PNL but subsequently required re-hospitalization because of complications. In these patients, the complications (perinephric hematoma or hydrothorax) developed after removal of the nephrostomy tube, and thus the actual removal of the tube may have induced the complication.
Our study is inherently limited by its retrospective design. Patient selection for having a CT post-PNL was biased by surgeon preference, patient habitus, and stone characteristics as demonstrated in Table 1. When the surgical procedure was more difficult and potential complications were anticipated, the surgeon was more likely to obtain a CT postoperatively. As a specialized referral center, our patient population was complex and potentially prone to complications. For example, 396 patients (38.3%) had a BMI>30, 78 (7.55%) had an ectopic kidney or renal fusion anomaly, and 86 (8.3%) had significant skeletal abnormalities such as scoliosis and/or limb contractures (Table 1). This could potentially increase the incidence of postoperative complications. On the other hand, PNL is a high-volume procedure in our center and is done by experienced surgeons.
One technique that we commonly apply at our institution for supracostal punctures is the tubeless upper pole approach, which has been reported previously. 12 The technique uses lower pole nephrostomy drainage with an 8F or 10F catheter after a supracostal access. Patients deemed highly likely to be SF at the conclusion of PNL via a supracostal puncture can be selected for this approach. Avoiding a supracostal tube minimizes patient morbidity without compromising drainage or secondary access, if needed. Placement of a lower pole drain still allows for a second-look procedure when necessary. This approach may explain the low pleural complication rate in our series.
One of the major disadvantages of CT is the significantly larger dose of radiation delivered to the patient as compared with KUB or intravenous pyelography. 13,14 To address the issue of radiation exposure associated with CT, numerous centers are now using low-dose protocols in the evaluation of nephrolithiasis. 13 In our series we have used a standard noncontrast CT scan. While lower dose radiation protocols seem to yield comparable results to standard dose protocols, low-dose CT may be insufficient in detecting very small stones or stones in obese patients because of a reduced signal-to-noise ratio. 14 Sensitivity of low-dose CT in diagnosing post-PNL complications is currently unknown. Future studies are encouraged to assess the ability of low-dose CT in evaluating postoperative complications for kidney stone surgery.
Conclusions
Post-PNL complications are not common, but their prompt diagnosis and treatment is imperative. In the current study, postoperative CT had a sensitivity of 92.7% for diagnosing early postoperative complications. CT revealed significant complications in a timely manner, thereby leading to prompt treatment. Patients who had post-PNL CT were less likely to be readmitted because of complications. Postoperative CT is effective in assessing postoperative stone status as well as in diagnosing complications. It has the potential to allow for early intervention for postoperative complications before significant clinical deterioration occurs. Postoperative CT could be considered for all patients undergoing PNL, but should be strongly considered in patients with complexities such as anatomical abnormalities, patients with the need for upper pole/supracostal access, or in patients requiring multiple access tracts, as the risk for significant postoperative complications is considerably higher in these patients.
Footnotes
Acknowledgments
Dr. Gnessin's work was supported in part by the American Physician Fellowship program.
Disclosure Statement
Dr. Lingeman is a consultant and preceptor for Boston Scientific Corporation.