Outcomes and Management Considerations in Patients on Dialysis Undergoing Laparoscopic Radical Nephrectomy for Renal-Cell Carcinoma

Abstract

Purpose:

To analyze the perioperative outcomes and management considerations in patients with dialysis-dependent end-stage renal disease (ESRD) undergoing laparoscopic radical nephrectomy for renal-cell carcinoma (RCC).

Methods:

There were 224 consecutive laparoscopic radical nephrectomies reviewed. Of those, 37 patients with ESRD were identified and compared with 187 patients with sporadic RCC. Evaluable parameters included age, sex, race, side of surgery, medical comorbidities, body mass index, American Society of Anesthesiologist (ASA) scoring, and age adjusted Charlson Comorbidity Index. All complications occurring intraoperatively and within the first 30 days were classified as per the Clavien classification system. Presurgical workup and transplant considerations were evaluated. Demographic and clinical characteristics were compared using Student t tests and chi-square tests for categoric variables.

Results:

Compared with non-ESRD patients, those with ESRD were younger and had smaller tumors. ASA was significantly higher in the ESRD group (P<0.001). Mean blood loss was similar between ESRD patients and non-ESRD patients. Overall complication rates were higher in patients with ESRD. Pathologic characteristics of ESRD renal masses included a higher proportion of papillary RCC.

Conclusion:

Patients with RCC associated with ESRD tend to have a higher ASA class and lower grade tumors. In addition, this population is at increased risk of surgical complications and more likely to need transfusions. Careful preoperative preparation and intraoperative anesthetic management are crucial to minimize patient morbidity and improve outcomes.

Introduction

Laparoscopic radical nephrectomy (LRN) is a standard and accepted option for the treatment of patients' organ-confined renal-cell carcinoma (RCC). The minimally invasive nature of the procedure affords shorter hospital stays, excellent cosmesis, reduced postoperative pain, and decreased recovery time, compared with the traditional open technique.¹ In addition, LRN for T₁ and selected T₂ RCC yields equivalent cancer control and lower morbidity compared with open approaches.¹

In 2014, an estimated 116,946 new cases of end-stage renal disease (ESRD) will be reported.² The pathologic complex of ESRD carries a number of associated risks, most notably those that are cardiovascular in nature. The adjusted all-cause mortality rate for cardiovascular sequelae in patients with ESRD is 6.4 to 7.8-fold higher than that of the general population.³ In addition, persons with renal failure often have other comorbidities, including diabetes, hypertension, altered clotting profiles secondary to uremia, and peripheral vascular disease, thus placing them at increased risk for surgical complications.

Little is written in the literature about the management and outcomes of laparoscopic surgical procedures in patients receiving dialysis. We reviewed our series of dialysis-dependent patients undergoing laparoscopic nephrectomy and compared their outcomes with nondialysis patients. We present the results of preoperative workup, kidney transplant considerations, unique management issues, and potential complications in this high-risk group undergoing a laparoscopic surgical procedure.

Methods

The study design and methods was revived and approved by the Institutional Review Board of Albert Einstein College of Medicine. We performed a retrospective study from our prospectively maintained database for all patients undergoing LRN by a single surgeon (RG) from January 2005 through August 2012 at Montefiore Medical Center in the Bronx, NY. Parameters included in our consideration included age, sex, race, side of surgery, medical comorbidities, American Society of Anesthesiologists (ASA) scoring, body mass index, age adjusted Charlson Comorbidity Index, length of surgical procedure, length of stay (LOS), estimated blood loss (EBL), laboratory parameters, and need for blood transfusion. The pathology reports were reviewed to obtain information on renal pathology (RCC vs non-RCC), size of the tumors, Fuhrman grade, and tumor stage. All slides were reviewed by a dedicated uropathologist, and the tumors were classified according to the 1997 International Union Against Cancer/American Joint Committee on Cancer consensus and the World Health Organization (2004) classification systems.

Renal function was evaluated using the Chronic Kidney Disease Epidemiology Collaboration estimated glomerular filtration rate formula. Estimated glomerular filtration rate (eGFR) was calculated by the Modification of Diet in Renal Disease formula. Units for eGFR were reported in mL/min/1.73 m², all units for creatinine clearance are reported in mL/min, and creatinine was reported in mg/dL. All complications occurring intraoperatively and within the first 30 days after the surgical procedure were classified per the Clavien classification system.

All patients underwent preoperative medical evaluation for surgical clearance. In addition, all patients with ESRD undergoing nephrectomy were sent for cardiac evaluation. Further preoperative evaluations were undertaken on a case by case basis based on the patient's medical profile. Serum electrolytes, hemoglobin, hematocrit, platelets, and a coagulation profile were evaluated. Patients with ESRD receiving dialysis continued on their routine dialysis schedule and, in some instances, were admitted to the hospital before surgery.

Dialysis was performed on the day before surgery. Postoperatively, dialysis was initiated per the normal patient cycle unless warranted by postoperative laboratory results. There were no patients whose blood counts precluded them from heparin administration while concomitantly needing emergent dialysis. Therefore, heparin was administered in its usual dosage, unless bleeding was of concern; those patients had a delay in their routine HD cycle under surveillance by the nephrology service. All laparoscopic nephrectomies were performed via the transperitoneal approach, as described previously.⁴ All dialysis-dependent ESRD patients were transferred to the surgical step-down unit after surgery and were discharged home once they met standard discharge criteria.

Statistical analysis

We compared demographic, clinical characteristics between patients with ESRD vs those without ESRD using Student t tests for continuous normally distributed variables, Wilcoxon sign-rank test for continuous not normally distributed variables (e.g., LOS), and chi-square tests for categoric variables. All tests were two-sided with a significance level of alpha=0.05. We used logistic regression to fit a multivariate model, which examines associations between ESRD (yes vs no) and several demographic and clinical variables simultaneously. The final model included only variables that were associated statistically significantly with ESRD at P<0.05. All analyses were performed using STATA software.

Results

Patient demographics along with medical comorbidities and statistical associations for ESRD and non-ESRD patients are presented in Table 1. Table 2 represents a multivariate model describing the only significant differences in baseline characteristics between the ESRD and non-ESRD groups. In total, 224 consecutive patient records were evaluated. Of those, 37 (17%) patients with ESRD who were currently receiving dialysis underwent laparoscopic nephrectomy for suspected renal malignancy. Patients with ESRD were more frequently of black (62%) and Hispanic ethnic backgrounds (30%, P<0.002) than others.

Table 1.

Demographic and Comorbidity Information for End-Stage Renal Disease and Non–End-Stage Renal Disease Patients

	ESRD		Without ESRD		P-value
Characteristics	37		187
Age; mean SD	51.32	12.47	59.24	15.33	0.004
ASA; mean SD	3.29	0.57	2.54	0.64	<0.0001
BMI; mean SD	27.12	5.75	29.78	7.16	0.039
Race; n %					0.002
1: White	2	5.4%	42	22.5%
2: African American	23	62.2%	58	31.0%
3: Hispanic	11	29.7%	70	37.4%
4: Other/unknown	1	2.7%	17	9.1%
Sex, n %					0.72
F	19	51.4%	90	48.1%
M	18	48.6%	97	51.9%
Comorbidities
DM	10	27.0%	53	28.3%	0.87
HTN	29	78.4%	132	70.6%	0.34
CHF	5	13.5%	8	4.3%	0.03
CAD	5	13.5%	12	6.4%	0.14
Hypercholesterolemia	2	5.4%	23	12.3%	0.22
Lung disease	2	5.4%	6	3.2%	0.51
MI	1	2.7%	6	3.2%	0.87
Angina	1	2.7%	6	3.2%	0.87
Liver disease	1	2.7%	3	1.6%	0.65
Thyroid disease	3	8.1%	7	3.7%	0.24

ESRD=end-stage renal disease; SD=standard deviation; ASA=American Society of Anesthesiologists; BMI=body mass index; DM=diabetes mellitus; HTN=hypertension; CHF=congestive heart failure; CAD=coronary artery disease; MI=myocardial infarction.

Table 2.

Logistics Regression Model Showing the Primary Differences Between End-Stage Renal Disease and Non–End-Stage Renal Disease Patients

ESRD	Odds ratio	95% CI		P-value
Younger age	0.92	0.88	0.96	<0.0001
Higher ASA	6.36	2.43	16.59	<0.0001
Larger tumor	0.73	0.58	0.91	0.006

CI=confidence interval.

Information regarding dialysis was available from our internal medical records in 32 patients, while 5 patients had external records that were imported to our medical data systems. The duration of dialysis before nephrectomy for the tumors ranged from 1 to 19 years, with a mean of 8 years. Patients with ESRD had a greater ASA score (3.3 vs 2.6, P<0.001). Preoperative hematocrit in the ESRD group was an average 38.6 and was similar to the average of the non-ESRD group at 38.8. Percent change in hematocrit from pre- to postoperative values was 0.07% and 0.08% for ESRD and non-ESRD, respectively.

Oncologic outcomes are reported in Table 3. RCC was confirmed in 155 (69%) patients. Clear-cell RCC was the most frequent tumor type noted in both groups (ESRD=40.5% vs non-ESRD=44.9%). Although the prevalence of papillary RCC was higher in the ESRD group (32.4%) in comparison with non-ESRD (18.7%), this difference was not statistically significant (P=0.30). The majority of the patients with ESRD presented with pT₁ lesions when compared with patients with non-ESRD (70.3% vs 38.5%, P=0.001). In addition, mean tumor size was larger in patients without ESRD in comparison with those with ESRD: 6.1 cm vs 3.2 cm, P=0.001. These findings are likely a result of the use of partial nephrectomy for smaller lesions in the non-ESRD cohort.

Table 3.

Pathologic Outcomes for End-Stage Renal Disease and Non–End-Stage Renal Disease Patients Undergoing Laparoscopic Radical Nephrectomy

	ESRD		Without ESRD		P-value
Tumor characteristics
Pathology					0.30
Medullary RCC	0	0.0%	6	3.2%
Clear-cell RCC	15	40.5%	84	44.9%
RCC-papillary	12	32.4%	35	18.7%
Chromophobe	1	2.7%	12	6.4%
Non-RCC path	9	24.3%	50	26.7%
Patients with RCC	28		137
Tumor size (cm) avg, SD	3.12	3.34	6.1	3.06	0.0001
Fuhrman tumor grade					0.15
1	2	5.4%	10	5.3%
2	10	27.0%	76	40.6%
3	15	40.5%	43	23.0%
4	1	2.7%	8	4.3%
Tumor stage					0.001
pT₁	26	70.3%	72	38.5%
pT₂	1	2.7%	34	18.2%
pT₃	1	2.7%	30	16.0%
pT₄	0	0.0%	1	0.5%

RCC=renal-cell carcinoma.

Of patients with ESRD, 76% were found to have malignancy. There was no statistically significant difference in Fuhrman grade between the two groups. One patient with ESRD who had a nephrectomy for a T_1a lesion was found to have recurrence, while five non-ESRD patients were found to have recurrence after the initial path showed node positive features. There were no recurrences within the renal fossa. Of the 37 patients with ESRD, 23 of them had the nephrectomy to attain candidacy for renal transplantation. As of August 2014, 13 of the 27 patients with ESRD have been grafted while 10 remain on the active transplant list.

Information regarding patient complications and postoperative parameters are presented in Table 4. No major intraoperative complications occurred in any patients, and there were no open conversions. Nine of 37 in the ESRD group and 27 of 187 in the non-ESRD group experienced postoperative complications. The overall complication rate was higher among patients with ESRD (24.3% vs 14.4%). Clavien grade III and IV complications were more commonly reported in the postoperative period in patients with ESRD, but this did not achieve statistical significance. Five patients with ESRD received at least 1 unit of blood within the postoperative period with two of the five needing more than 4 units, while one patient without ESRD received a single transfusion.

Table 4.

Complications and Postoperative Parameters in ESRD and Non-ESRD Patients Undergoing Laparoscopic Radical Nephrectomy

	ESRD		Without ESRD		P-value
Clavien Grade					0.38
0	28	75.7%	160	85.6%
1	1	2.7%	7	3.7%
2	6	16.2%	17	9.1%
3 & 4	2	5.4%	3	1.6%
PACU_Tach	2	5.4%	31	16.6%	0.08
PACU_Hyp	1	2.7%	2	1.1%	0.43
preop_HCT	38.64	5.66	38.79	5.04	0.87
Postop_HCR	35.52	4.99	34.41	4.71	0.21
HTC Change %	0.07	0.11	0.11	0.08	0.028
Serum CR	8.35	3.73	1.22	0.66	<0.0001
MDRD_eGFR	7.65	2.64	69.95	33.49	<0.0001
EBL	108.88	100.69	171.69	148.13	0.035
OR Time	125.93	48.37	148.56	58.80	0.04
LOS; median IQR	4	3 to 5	4	3 to 5	0.76

EBL=estimated blood loss; OR=operating room; LOS=length of stay.

Specific Clavien grade III and IV complications in the ESRD group included postoperative exploratory laparotomy for unstable hematocrit and hematoma and reintubation for respiratory distress. Clavien III–IV complications in the non-ESRD group included hematuria necessitating cystoscopy and clot evacuation, pneumothorax managed with a chest tube, and small bowel obstruction necessitating exploratory laparotomy. No significant difference was observed between dialysis-dependent patients and non–dialysis-dependent with regard to operative time, EBL, and length of hospital stay.

Discussion

We identified patients with ESRD receiving dialysis who were undergoing LRN and compared their outcomes with a group of patients without ESRD. Logistic regression analysis revealed that with the exception of age, tumor size, and ASA classification, the two study groups were similar with regard to baseline characteristics (Table 2). Our study is, to the best of our knowledge, the largest reported series of LRN in patients with ESRD.

In our study, clinical parameters such as operative time and perioperative blood loss were comparable for both groups. Postoperative complications were not significantly more common in patients with ESRD and were more generally exacerbations of preexisting medical conditions. There were also no significant differences in length of hospital stay. We found that the EBL was similar between both groups, but patients with ESRD were more likely to need a blood transfusion. On pathologic review, patients with ESRD in this series were more likely to have papillary RCC.

Our results are similar to findings from other authors who have reported on laparoscopic nephrectomy in patients with ESRD. Dunn and colleagues⁵ reported on outcomes after laparoscopic nephrectomy in 11 patients with autosomal dominant polycystic kidney disease and reported that 14% needed transfusion. Similarly, Fornara and associates⁶ compared 19 patients with ESRD undergoing laparoscopic nephrectomy with 20 consecutive non-ESRD patients and reported an increased transfusion rate (32% vs 0%). The authors suggested this difference may be related to a low starting serum hemoglobin and not to increased blood loss or bleeding diatheses.

More recently, Bird and coworkers⁷ reported on outcomes after laparoscopic nephrectomy in 16 patients with ESRD. Complication rates according to the Clavien classification system were reported as 31.3% in ESRD patients and 6.3% for non-ESRD patients, although this did not achieve significance; P=0.05). LOS, however, was found to be significantly longer in patients with ESRD (median: 90 vs 70 hours, P=0.001). Intraoperative blood loss did not differ between the two groups.

Renal masses in our ESRD population were most often found among patients undergoing evaluation for renal transplantation. Our transplant service mandates that any person receiving hemodialysis for more than 4 years undergo abdominal imaging to rule out RCC. Whereas lesions with minimal suspicion in the non-ESRD populations are sometimes observed, lesions with any index of suspicion in the ESRD population under review for renal transplant undergo intervention. Time from nephrectomy to transplant ranged from 1 month to 6 years.

Clearance for transplant from an RCC perspective was at the discretion of the urology team and was predicated on the pathology reports. Benign pathology was cleared immediately, T₁ node negative lesions were cleared at the 3-month mark, T₂ and T₃ lesions were cleared on a case by case basis and patients were sent routinely for at least one postoperative imaging modality. Of interest, there has been one reported recurrence in our ESRD cohort from a 1.7 cm upper pole lesion with negative margins and nodes on initial pathology. This patient received a transplant 1.5 years after her nephrectomy. Three years after nephrectomy, she was found to have RCC metastatic to the liver confirmed by biopsy. This patient falls into the rare category of <4% of patients who harbor disease outside the kidney with initial lesions less that 4 cm. No other recurrences in the ESRD population have been noted as of August 2014. In addition, there were no port tract seeding events identified.

The most common tumor type in our patients was acquired cystic disease-associated RCC (ACD-associated RCC), seen in 56% of ESRD patients with RCC. Papillary RCC comprises up to 18% of the renal epithelial tumors in the general population but has been reported to form a larger proportion (42%–71%) of the tumors in ESRD.⁸ Among our cases, we found papillary RCC to constitute 32% of tumor masses in ESRD. The biologic behavior of RCCs in ESRD has been reported to be less aggressive than those in sporadic or non-ESRD patients.⁹ A possible reason suggested for this less aggressive behavior are that these patients are usually under constant medical care, and as such, under closer radiologic scrutiny. This is one possible reason for the fact that tumors in patients with ESRD tend to be smaller (mean 3.2 cm in our series). A more likely reason, however, as mentioned above, is the widespread use of partial nephrectomy in the management of small renal masses in the general population.

The population of patients with ESRD is unique given the extent of their comorbidities and the physiologic impact of dialysis. Therefore, in addition to medical clearance by internal medicine physicians, all ESRD patients were referred for cardiology evaluation regardless of previous cardiac history. Cardiac evaluation included echocardiography, electrocardiography, and routine cardiac laboratory work. Although not recorded in our database, seven patients sent for cardiac evaluation were excluded from nephrectomy secondary to complex and serious cardiac comorbidities. Of the 37 patients who ultimately had nephrectomies, 36 of the 37 followed through with cardiology referral. Three of the 36 were found to have abnormalities on an echocardiography.

Further workup in two of these three revealed significant coronary artery disease, and those two patients underwent percutaneous interventions before radical nephrectomy. The single patient who was not evaluated by the cardiology team was cleared for the surgical procedure by internal medicine physicians only. The patient had no history or complaints of cardiac pathology. Two days postoperatively, the patient was worked up for significant shortness of breath and lower extremity edema. The echocardiography showed abnormal wall motion and ejection fraction of 36%. The patient received a diagnosis of congestive heart failure in consultation with house staff cardiology. We feel strongly that given the wide range of biologic and physiologic effects seen in ESRD and HD patients, all patients must be evaluated by cardiologists and doing so protects the urologist from encountering complications that might surface for the first time in the perioperative period.

The anesthesia team was also proactive in their patient evaluation before surgical intervention. Before the day of operation, the anesthesia team, through the medical records, routinely evaluated recent patient blood pressure trends. Any patient deemed to have labile blood pressure had an arterial “A” line placed for close monitoring during the procedure. Eleven of 37 patients with ESRD had A-lines placed. On the morning of the intervention, routine blood work was performed to ensure electrolytes and blood parameters were within normal limits in the patients with ESRD. Two patients in the ESRD cohort had their operation shifted to later in the day to accommodate an extra dialysis session based on blood work findings.

Finally, all patients with ESRD were placed on a 3 mL/kg drip of crystalloid on arrival to address likely dehydration from hemodialysis the day before. The 3 mL/kg regimen represents a balance between ensuring hydration without overloading the patient receiving hemodialysis. There were no anesthesia-related complications reported or observed for the entire cohort of patients undergoing nephrectomy.

To minimize postoperative complications, patients with ESRD were sent to a monitored setting for 24 hours. Emphasis was placed on evaluation for bleeding, which was the most common postoperative complication encountered in the ESRD group (14%). In addition to postoperative laboratory work, serial abdominal examinations were undertaken, and interestingly, the one patient who was taken back to the operating room for an unstable hematoma was identified through an abdominal examination 7 hours postoperatively. Laboratory values 4 hours before showed a 5-point drop in the hematocrit and 1.5 point decrease in the hemoglobin from preoperative values. Although surgical procedures in patients with ESRD may be associated with an increased risk of bleeding partly because of the qualitative platelet dysfunction associated with uremia, with meticulous attention to hemostasis, the need for blood products may be minimized.

In our series, there were no significant differences in pre-/postoperative hemoglobin/hematocrit in the ESRD group or the non-ESRD group. Hemodialysis adequacy is an important perioperative consideration and has been shown to improve surgical outcomes.¹⁰ Loss of erythropoietin secondary to advanced renal disease is a major factor that can lead to clinically significant anemia in patients with ESRD, and that anemia may potentially aggravate existing cardiac dysfunction in susceptible patients.¹¹ Therapy with iron supplementation and blood transfusions as necessary may optimize outcomes.

In treating patients receiving hemodialysis, there are a number of general perioperative care issues to consider. First, intravenous access and blood pressure monitoring should avoid the arm that bears the arteriovenous fistula or its future site. Along a similar line, particular attention should be paid to positioning on the surgical table so there is risk of thrombosis of the arteriovenous fistula. Fornara and colleagues⁶ reported that postoperative thrombosis of the arteriovenous fistula developed in 2 of their 19 patients with ESRD. Second, the clinician must be vigilant and proactive in managing electrolyte abnormalities postoperatively. Attention must be paid to close monitoring of hyperkalemia, replacement of calcium for symptomatic hypocalcemia, and the use of phosphate binding agents for hyperphosphatemia.

Third, as mentioned above, consideration should be given to both the type and rate of intravenous fluids. Patients receiving hemodialysis who have not eaten in more than 8 hours are at particular risk for multiple systemic derangements, and aggressive administration of fluids could lead to pulmonary and or tissue edema. We suggest light maintenance fluids augmented by boluses of colloid/crystalloid as needed to maintain stability. Hemodialysis is usually best delayed until the risk of hemorrhage and shifting fluids has fallen (usually 24 hours postoperatively). Finally, care should be taken with the use of narcotics for postoperative pain control because they may have prolonged effects despite hepatic clearance.

A final point worth mentioning in this series is the surgical approach. While all the LRN in this cohort were performed transperitoneally, a retroperitoneal approach is a feasible option and gaining popularity for small tumors or in patients with previous abdominal surgery. Benefits of this approach include avoiding a previously manipulated abdomen. Limitations usually pertain to surgeon expertise because the loss of landmarks and limited working space make this approach technically challenging.

Limitations to this study include its retrospective nature despite being gleaned from a prospective database. In addition, the ESRD group of patients analyzed was also smaller than the comparative non-ESRD group. We believe, however, that our findings remain valid, despite possible contamination.

Conclusion

LRN for the management of renal tumors in dialysis-dependent patients with ESRD is feasible and safe in the hands of a skilled laparoscopic surgeon. We think that part of the feasibility hinges on the way these patients are worked up preoperatively and how management is undertaken in the postoperative period. Having proper understanding of the entity of ESRD while encouraging multiple specialties to work in concert can help prevent adverse outcomes before they occur.

Footnotes

Disclosure Statement

No competing financial interests exist.

Abbreviations Used

References

Bhayani

, Clayman

, Sundaram

, et al. Surgical treatment of renal neoplasia: Evolving toward a laparoscopic standard of care. Urology, 2003; 62:821–826.

Collins

, Foley

, Chavers

, et al. United States Renal Data System 2011 Annual Data Report: Atlas of chronic kidney disease & end-stage renal disease in the United States. Am J Kidney Dis, 2012; 59(suppl 1):A7, e1–420.

Azizzadeh

, Sanchez

, Miller

3rd , et al. Glomerular filtration rate is a predictor of mortality after endovascular abdominal aortic aneurysm repair. J Vasc Surg, 2006; 43:14–18.

Feder

, Patel

, Melman

, et al. Comparison of open and laparoscopic nephrectomy in obese and nonobese patients: Outcomes stratified by body mass index. J Urol, 2008; 180:79–83.

Dunn

, Portis

, Elbahnasy

, et al. Laparoscopic nephrectomy in patients with end-stage renal disease and autosomal dominant polycystic kidney disease. Am J Kidney Dis, 2000; 35:720–725.

Fornara

, Doehn

, Miglietti

, et al. Laparoscopic nephrectomy: Comparison of dialysis and non-dialysis patients. Nephrol Dial Transplant, 1998; 13:1221–1225.

Bird

, Shields

, Aziz

, et al. Transperitoneal laparoscopic radical nephrectomy for patients with dialysis-dependent end-stage renal disease: An analysis and comparison of perioperative outcome. Urology, 2010; 75:1335–1342.

Denton

, Magee

, Ovuworie

, et al. Prevalence of renal cell carcinoma in patients with ESRD pre-transplantation: A pathologic analysis. Kidney Int, 2002; 61:2201–2209.

Amin

, Amin

, Tamboli

, et al. Prognostic impact of histologic subtyping of adult renal epithelial neoplasms: An experience of 405 cases. Am J Surg Pathol, 2002; 26:281–291.

10.

Hemodialysis Adequacy 2006 Work Group. Clinical practice guidelines for hemodialysis adequacy, update 2006. Am J Kidney Dis, 2006; 48(suppl 1):S2–S90.

11.

Trainor

, Borthwick

, Ferguson

. Perioperative Management of the hemodialysis Patient. Semin Dial, 2011; 24:314–326.