Assessment of Perioperative Variables That Predict the Need for Surgical Drains Following Robotic Partial Nephrectomy Utilizing Quantitative Drain Creatinine Analysis

Introduction

The routine placement of surgical drains following partial nephrectomy has been questioned, with both selective and routine omission of drains following open and robotic-assisted partial nephrectomy (RAPN) having been shown to be safe.^1,2 Drainage following partial nephrectomy is intended to be a safety measure to recognize or prevent urinary fistula. However, a recent analysis of 1791 RAPN noted a urine leak rate of <1%. ³ It has been proposed that the increased discomfort secondary to postoperative drains may lead to prolonged hospital stay following partial nephrectomy.^1–3 Postoperative surgical drains have also been associated with complications such as peritonitis, evisceration, herniation, retained foreign bodies, and small bowel obstruction.^4–8

Body fluid analysis with drain creatinine values compared to simultaneously collected serum creatinine is the most reliable indicator of abdominal urine leakage.^9,10 Drain creatinine to serum creatinine ratios (D/S ratios) should be nearly identical in postoperative patients without this complication. ¹⁰ To our knowledge, there has not been an evaluation of postoperative D/S ratios following RAPN. It has been our practice to routinely check drain creatinine levels and use the D/S ratio directly following RAPN to determine the timing of drain removal. Using this practice, the majority of drains were removed on postoperative day 1 or 2 and were, therefore, unnecessary.^9,10 By separating groups based on D/S ratio, this study aims to identify factors that may aid in selecting patients who may require postoperative drains.

Materials and Methods

Following Institutional Review Board approval, a prospectively compiled single-surgeon RAPN database was reviewed. Patient demographics including age, body mass index (BMI), American Society of Anesthesiologists (ASA) classification, and Mayo Adhesive Probability (MAP) ¹¹ score were collected. Tumor characteristics including tumor radius, exophytic/endophytic classification, nearness to the collecting system, and laterality (RENAL score) ¹² were collected. Operative variables such as operative time (OR time), warm ischemia time (WIT), estimated blood loss (EBL), and collecting system (CS) entry status were recorded following each case. Length of hospital stay and postoperative complications were recorded using the Clavien-Dindo classification system. ¹³ Surgical margin status and final pathology were noted as well.

RAPN was performed intraperitoneally using the robotic Da Vinci Si Surgical System (Intuitive Surgical Corporation, Sunnyvale, CA) via a method outlined in previous studies. ¹⁴ Surgical technique remained consistent throughout the time frame. In short, intraoperative ultrasonography was routinely used for identification of tumor margins and approximatation of tumor depth before athermal tumor excision using cold scissors. If a CS violation was noted during excision, running 2-0 absorbable sutures were used to close the CS. Cortical renorrhaphy was performed using the well-described sliding Weck technique. ¹⁵ A renal bolster was used when deemed appropriate by the surgeon. Finally, a 19-F round full-fluted drain was placed in the perinephric bed and set to low suction using a 100 cc silicone bulb evacuator before abdominal closure.

Following surgery, all patients were on track to have their 16F foley catheter removed on postoperative day 1 and be discharged home if tolerated. Drain creatinine was obtained routinely on postoperative day 1. Drain creatinine was compared to same-day serum creatinine to calculate the D/S ratio. The drain was removed if the D/S ratio was <1.2. Patients returned 2–3 weeks following discharge for routine follow-up laboratory tests and wound checks. An elevated D/S ratio was defined as any value >1.2.

Statistical analyses were performed using SAS (version 9.3; SAS Institute, Inc., Cary, NC). The unpaired t-test was used to compare continuous variables. Fisher's exact test was used to compare categorical variables. Univariate analysis was used to identify perioperative factors that might necessitate drain placement.

Results

A total of 150 patients underwent RAPN between February 2008 and April 2015. One hundred forty patients had a drain placed at the time of surgery. Of these, 124 consecutive patients had a recorded postoperative drain creatinine value and were included in this study. One hundred three patients had D/S ratio <1.2 and 21 patients had D/S ratio >1.2. Detailed patient characteristics are listed in Table 1. Of all the tumors resected, 73 (58.1%) were clear cell renal cell carcinoma (RCC), 21 (16.9%) were papillary RCC, 9 (7.3%) were chromophobe RCC, and 22 (17.7%) were benign.

Table 2 compares patient, tumor, and perioperative variables of patients with postoperative D/S ratio over 1.2 compared to those under 1.2. Tumor size was statistically associated with having a postoperative D/S ratio >1.2. Patients with D/S ratio >1.2 had longer drain time (P = .001) and hospital stay (P = .036) when compared to the D/S ratio <1.2 group. Four patients were discharged with a drain, all of whom had a D/S ratio >1.2. The mean D/S ratio for this subset of patients was 1.99 (range 1.44–2.89). The mean drain time was 4.2 days (range 2–13). No patients required stent placement following surgery.

Univariate analysis was performed with several variables, as listed in Table 3. Renal mass size was the only variable that reached statistical significance. For T1b renal masses (i.e., 4–7 cm) there is a 2.78 greater chance of having a postoperative D/S ratio >1.2 when compared to D/S ratio <1.2 (OR 2.78; P = .041).

Ten patients had RAPN done without the placement of a drain. Excluding one patient in the subanalysis who had a prior orthotopic liver transplantation and suffered a Clavien-Dindo grade V complication secondary to pneumonia, the patient characteristics for the drainless procedures were mean age 58.2 (range 22–75), BMI 27.7 (range 18.7–35.2), ASA score 2.67 (range 2–3), preoperative creatinine 1.06 (range 0.6–1.8), tumor size 2.2 cm (range 1.5–5.0 cm), RENAL nephrometry score 5.7 (range 4–10), and MAP score 0.78 (range 0–4). Operative and perioperative characteristics for the drainless cases were mean WIT 5 minutes (range 0–24), total operative time 175 minutes (range 140–216), EBL 300 mL (range 200–500), and hospital stay 2.2 days (range 1–4). One patient had CS entry.

Discussion

The reported urinary leakage rate (urinoma/fistula) in the modern day partial nephrectomy literature ranges from 2% to 18.5% depending on the definition used.^1,16 However, a recent multicenter analysis of 1791 RAPN noted a urinary fistula rate of only 0.78%. ³ The authors theorized that higher surgical volume, improvement in preoperative imaging, and refinement of surgical technique have led to the decreased urine leak rates in the study. More specifically, the authors attributed the low leakage rate to the sliding Weck renorrhaphy and the improved visualization and optimized suturing angles offered by robotic technology. As in our study, large tumor size, renal hilar tumor location, operative time, and need for CS repair were predictive of urinary leakage. Godoy et al. ¹ was the first study to question routine drain placement after open partial nephrectomy. This group selectively omitted drain placement in roughly 15% of patients with a median tumor size of 2.0 cm. Their conclusion was that this practice was safe and feasible but should be reserved for small renal cortical tumors when the CS was not violated.

Abaza and Prall published their series on routine, not selective, omission of drain placement following RAPN. ² Their results included drain omission in 93% of 160 patients with a mean tumor size of 3.5 cm and RENAL nephrometry score of 7.8. Only two patients (1.3%) experienced urine leak. More controversially, they omitted drain placement in patients when CS violation was identified intraoperatively. All but five patients (97%) were discharged on postoperative day 1, which was attributed to less postoperative pain in the absence of a surgical drain.

While these studies may hint that omitting drain placement can be done safely following partial nephrectomy, they do not prove that drain omission is better than drain placement. However, it is important to be reminded that drain placement is not without harm. While complications are rare, case reports have been published to show drain involvement in retained foreign bodies, small bowel obstruction, intestinal evisceration, and postoperative intraperitoneal infection.^4–9,17

Peyronnet et al. ¹⁸ has published the most recent and largest series to date to answer this question. This multicenter retrospective study compared three centers that practiced routine drain omission to five centers that routinely placed postoperative drains. They found urinary fistula with RAPN to be rarer than previous believed. Additionally, complication rates and postoperative imaging rates were no different between the groups. Interestingly, there was a statistically significant difference in length of hospital stay, which was lower for the group without drains.

The D/S ratio is the most reliable test for abdominal urinary leakage.^9,10 An analysis of colorectal surgery patients noted that the D/S ratio ranged from 29.7 to 89 times higher in colorectal surgery patients with diagnosed postoperative urinary tract injury. ⁹ Moreover, the drain/body fluid ratios have been shown to be identical in postoperative patients without this complication. ¹⁰ There is no scientific explanation of why the D/S ratio should be over 1 in the absence of urinary leakage.

To our knowledge, no one has described the routine testing of drain creatinine in the setting of postoperative RAPN. While an elevated D/S ratio does not definitely mean there is a urinary fistula, a urinary fistula would definitely result in an elevated D/S ratio. Our results indicate that D/S ratio elevations after RAPN are more likely to occur with larger renal masses. Interestingly, with CS violation, our results demonstrated no difference between normal D/S ratio and elevated D/S ratio. These results could be valuable for clinical decision making if a selective approach to drain omission was taken.

As with any test, sensitivity and specificity are important to consider. Our definition of a D/S ratio over 1.2 as a sign of urinary leakage may seem rigid. However, we believe this value errs on the side of having a high sensitivity and thus identifying potential early urinary fistulas. It is possible that a lower D/S ratio sacrifices specificity. For instance, Peyronnet et al. ¹⁸ showed drainless RAPN to have a urine leak rate of roughly 1%. If they were to check drain creatinine, many would have likely been identified as D/S ratio >1.2, but it is also probable that the majority of these would be clinically insignificant.

This study relies on the premise that urinary fistulas develop as a result of surgical technique. In other words, the D/S ratio—if elevated—will be so after surgery as long as there are functioning renal units. Conversely, if the D/S ratio is not elevated, then it should not be elevated any time thereafter. It is possible that this is an incorrect assumption, but it would be difficult to prove otherwise.

Limitations to this study are intrinsic in its retrospective design and inherent selection bias. Drain omission was performed in select cases at the surgeon's discretion; thus, this study includes a subset of the true RAPN sample. Most cases for which drain placement was avoided were of smaller, peripherally located renal masses without perceived CS violation. Therefore, this study may overestimate the true incidence of elevation in drain creatinine following RAPN. Another limitation is that the sample size lacked statistical power and prevented any multivariate testing. Furthermore, imaging was not routinely performed in the perioperative period to confirm diagnosis of urinary fistula.

Conclusion

D/S ratio analysis demonstrates that most RAPN can safely be left without a surgical drain. An elevated D/S ratio is more likely to occur following RAPN in larger masses (those approaching or larger than 4 cm) compared to small renal masses. An elevated postoperative D/S ratio, indicative of possible urinary leakage, will likely contribute to prolonged surgical drain time and longer hospital stay.