Perioperative outcomes of laparoscopic, robotic, and open approaches to pheochromocytoma

Abstract

While multiple studies have demonstrated that minimally invasive surgical (MIS) techniques are a safe and efficacious approach to adrenalectomy for pheochromocytomas (PC), these studies have only been small comparative studies. The aim of this multi-institutional study is to compare perioperative outcomes between open and MIS, stratified by robotic and conventional laparoscopic, techniques in the surgical management of PC. We retrospectively evaluated patients who underwent adrenalectomy for PCs from 2000 to 2017 at three different institutions. Clinical, perioperative, and pathologic parameters were analyzed using t test, Chi square, and Fisher exact statistical measures. Of the 156 adrenalectomy cases performed, 26 (16.7%) were with an open approach and 130 (83.3%) using MIS techniques. Of the MIS procedures, 41 (31.5%) were performed robotically and 89 (68.5%) performed laparoscopically without robotic assistance. Demographic and clinical parameters were similar between the open and MIS groups. Patients, who underwent MIS procedure had a lower complication rate (p = 0.04), shorter hospitalization (p = 0.02), shorter operative time (p < 0.001), and less blood loss (p = 0.002) than those who underwent open surgical resection. Conventional laparoscopic and robotic operative approaches resulted in similar complication rates, length of hospitalization, and blood loss. Our study is one of the largest cohorts comparing the perioperative outcomes between conventional laparoscopic and robotic adrenalectomies in patients with PC. Our results support that MIS techniques have potentially lower morbidity compared to open techniques, while laparoscopic and robotic approaches have similar perioperative outcomes.

Introduction

Pheochromocytomas (PC) are metabolically active tumors derived from the neural crest cells in the adrenal medulla. Excessive catecholamine release from these tumors can present with a wide range of symptomology ranging anywhere from asymptomatic to polypharma-resistant hypertension, tachycardia, and palpitations.

Since the 1990s, laparoscopic adrenalectomy has largely replaced open surgery as the standard for adrenalectomy in well-selected patients at many centers [1]. Despite the decreased length in hospitalizations and fewer complications that laparoscopic surgeons offer, there are still concerns regarding its ability to safely resect large adrenal tumors and pheochromocytomas [2, 3]. Over the last 10 years, the advent of robotic adrenalectomy utilizing the da Vinci robotic surgical platform (Intuitive Surgical, Sunnyvale, CA, USA) has provided surgeons with a three-dimensional operative view and increased maneuverability using the EndoWrist^® technology further modifying laparoscopic surgical approaches. Initial reports cited the learning curve of robotic adrenalectomy, longer operative time, and increased healthcare costs as potential disadvantages when compared to traditional laparoscopy. There has also been suggestion that the lack of tactile feedback when using robotic-assisted laparoscopy can potentially result in undue tumor manipulation and inadvertent catecholamine release. This could increase the risk of hemorrhage, cardiovascular instability, and intraoperative hypertensive crisis [4].

While multiple studies have demonstrated comparable safety and efficacy between the robotic and laparoscopic approaches; these have all been small comparative cohort studies [5–9]. The aim of this multi-institutional study is to compare perioperative outcomes between open and minimally invasive surgery (MIS), furthermore stratifying MIS as robotic and conventional laparoscopic techniques, in the operative management of preoperatively recognized PCs.

Methods

Study population and parameters

After IRB approval was obtained at all participating centers, patients who underwent adrenalectomy for pathologically confirmed PC (PC) were retrospectively evaluated at University of Alabama at Birmingham (Birmingham, AL), New York Medical College (Valhalla, NY) and Hospital das Clínicas de São Paulo (São Paulo, Brazil), from 2000 to 2017. No partial adrenalectomies were included in our multi-institutional cohort or analysis. We obtained all available demographic, familial, and clinical data. Adjusted Charlson comorbidity index (CCMI) scores were calculated using an Excel-based application. Operative report, intraoperative anesthesia reports, Pheochromocytoma of the Adrenal gland Scaled Scores (PASS), inpatient complications, postoperative hospitalization, and pathologic reports were examined [10]. Hemorrhage was defined as a patient requiring transfusion in the setting of acute blood loss anemia. High-grade perioperative complications were defined as Clavien–Dindo III and greater. Hypertensive adrenal crisis was defined as symptomatic ≥ stage 2 hypertension and hemodynamic compromise requiring control with vasoactive agents. Pathologic specimen size was defined as the maximum diameter of the surgical specimen (cm), while tumor size was defined as the maximum diameter of the tumor (cm) itself. Our primary outcomes were perioperative morbidity that we assessed with variables including estimated blood, prevalence of high-grade perioperative complications, and length of hospitalization.

Statistical analysis

Continuous variables were assessed with unpaired Student t tests between two groups. Chi square and Fisher exact tests were used to assess categorical variables among different groups. All p values were two-sided with a predetermined alpha level set at 0.05.

Results

Between 2000 and 2017, 153 patients underwent 156 adrenalectomy cases for inclusion in this study (Table 1). Patients who underwent MIS and open adrenalectomies were similar in terms of age, gender, CCMI, tumor side, and BMI (Table 2). However, differences were noted between the conventional laparoscopic and robotic surgery groups. Patients who underwent conventional laparoscopic surgery tended to be younger (46.2 ± 17.9 versus 55.9 ± 15.4 years old, respectively, p = 0.004), more likely to have history of hypertensive crisis (39.3% versus 19.5%, respectively, p = 0.03), less likely to have a cardiovascular history (9.1% versus 24.4%, respectively, p = 0.02), and have a lower body mass index (BMI) (24.5 ± 4.9 versus 29.8 ± 6.5, respectively, p < 0.001) (Table 3).

Table 1.

Study population demographic and clinical data

Population characteristics
Patients	153
Adrenalectomy cases	156
No. male/female	70/83
Mean age ± SD (years)	48.34 ± 18.27
Mean BMI ± SD	26.54 ± 6.29
% Cardiac history	15.83
% Hypertension	59.72
% Crisis	33.97
% Family history	34.55
% Pediatric	5.00
Median CCMI (range)	1 (0–8)
Mean MAP at presentation (mmHg)	97.32 (12.71)
Average MAP at discharge MAP (mmHg)	91.41 (12.25)
Mean prescribed HTN meds admit ± SD	1.35 (1.38)
Mean prescribed HTN meds discharge ± SD	1.17 (1.16)
Mean prescribed meds admit ± SD	4.00 (3.56)
Mean prescribed meds discharge ± SD	6.45 (4.86)
Any complication (%)	46 (29.4)
High-grade complication (%)	15 (9.62)

Table 2.

Comparison of clinical and perioperative parameters between MIS and open adrenalectomy

Parameters	MIS (n = 130, mean ± SEM)	Open (n = 26, mean ± SEM)	p value
Age (years)	49.3 ± 17.7	43.7 20.7	0.153
Gender (male/female)	60/70	11/15	0.719
CCMI	1.8 ± 1.94	1.7 ± 2.4	0.888
Hypertensive crisis	43 (33.1%)	10 (38.5%)	0.597
CV history	17 (14.4%)	5 (23.8%)	0.329
Side (L/R/Bil)	51/71/8	6/17/1	0.337
BMI	26.4 ± 6.1	27.2 ± 7.8	0.608
OR fluids (ml)	3466 ± 1079	4843 ± 1957	0.006
EBL (ml)	150 ± 318	439 ± 423	0.002
OR time (min)	174.6 ± 76.4	260.4 ± 117.6	< 0.001
OR start mean BP (mmHg)	90.4 ± 19.4	94.0 ± 22.9	0.566
OR max SBP (mmHg)	171.8 ± 24.6	173.8 ± 25.4	0.798
OR min SBP (mmHg)	81.5 ± 17.7	83.3 ± 16.2	0.754
OR end BP (mmHg)	88.1 ± 14.7	77.0 ± 10.6	0.016
Any complication (%)	34 (26.2%)	12 (46.2)	0.041
High-grade complication (%)	10 (7.7%)	5 (19.2)	0.135
ICU stay (days)	1.2 ± 1.5	2.1 ± 2.0	0.056
Hospital stay (days)	3.5 ± 3.4	5.2 ± 3.0	0.021
Mass (g)	79.9 ± 106.6	215.0 ± 178.1	0.042
Specimen size (cm)	7.6 ± 2.1	8.7 ± 3.0	0.206
Tumor size(cm)	5.1 ± 5.12	7.5 ± 3.7	0.001
Positive margin	3 (6.7%)	3 (25%)	0.101
PASS	4.4 ± 3.2	4.6 ± 4.1	0.864
Familial disease	33 (27.8%)	5 (27.8%)	0.596

Table 3.

Comparison of clinical and perioperative parameters between laparoscopic and robotic adrenalectomy

Parameters	Lap (n = 89, mean ± SEM)	Robo (n = 41, mean ± SEM)	p value
Age (years)	46.2 ± 17.9	55.9 ± 15.4	0.004
Gender (male/female)	38/51	22/19	0.244
CCMI	1.6 ± 1.8	2.2 ± 2.2	0.127
Hypertensive crisis	35 (39.3%)	8 (19.5%)	0.026
CV history	7 (9.1%)	10 (24.4%)	0.024
Side (L/R/Bil)	33/48/8	18/23/0	0.133
BMI	24.5 ± 4.9	29.8 ± 6.5	< 0.001
OR fluids (ml)	3468 ± 981	3462 ± 1237	0.980
EBL (ml)	134 ± 243	173 ± 404	0.584
OR time (min)	157.9 ± 53.1	210.4 ± 103.0	0.004
OR start mean BP (mmHg)	95.2 ± 17.6	87.3 ± 20.1	0.138
OR max SBP (mmHg)	175.3 ± 25.4	169.8 ± 24.2	0.404
OR min SBP (mmHg)	86.1 ± 16.7	78.8 ± 17.9	0.124
OR end BP (mmHg)	88.4 ± 16.0	88.0 ± 14.1	0.917
Any complication (%)	23 (25.8%)	11 (26.8%)	0.905
High-grade complication	6 (6.7%)	4 (9.8%)	0.724
ICU stay (days)	1.5 ± 1.6	0.7 ± 1.2	0.002
Hospital stay (days)	3.7 ± 3.8	3.0 (2.3)	0.203
Mass (g)	75.9 ± 105.4	82.0 ± 108.9	0.847
Specimen size (cm)	6.9 ± 1.9	8.1 ± 2.1	0.031
Tumor size (cm)	4.6 ± 2.5	6.2 ± 8.4	0.245
Positive margin	1 (5.6%)	2 (7.4%)	1
PASS	4.8 ± 3.6	3.7 ± 2.1	0.699
Familial disease	29 (56.9%)	4 (9.8%)	< 0.001

In evaluation of perioperative parameters, the MIS group had faster operative time (174.6 ± 76.4 versus 260.4 ± 117.6 min, respectively, p < 0.001), less blood loss (150 ± 318 versus 439 ± 423 ml, respectively, p = 0.002) and required less intraoperative fluid (3466 ± 1079 versus 4843 ± 1957 ml, respectively, p = 0.006) when compared to patients undergoing open surgical approach to adrenalectomy. Intraoperative hemodynamics were consistent between MIS and open surgical cohorts, the exception being the MIS group that had a higher mean ending systolic blood pressure than the open surgical group (88.1 ± 14.7 versus 77.0 ± 10.6 mmHg, p < 0.02) (Table 2). Compared to the robotic approach, laparoscopy had a faster operative time (157.9 ± 53.1 versus 210.4 ± 103.0 min, respectively, p = 0.004). Otherwise, laparoscopic and robotic surgical approaches had similar blood, intraoperative fluid use, and intraoperative hemodynamics (Table 3).

While the MIS group had a lower perioperative complication compared to an open technique rate (26.2% versus 46.2%, respectively, p = 0.04), the laparoscopic and robotic approaches had similar perioperative complication rates. The rate of high-grade complications was similar regardless of approach. Additionally, patients who underwent MIS procedures also had shorter intensive care unit (ICU) stays (1.2 ± 1.5 versus 2.1 ± 2.0 days, respectively, p = 0.06) and shorter hospitalizations (3.5 ± 3.4 versus 5.2 ± 3.0 days, respectively, p = 0.02). The laparoscopic group had longer ICU stays than the robotic group (1.5 ± 1.6 versus 0.7 ± 1.2 days, respectively, p = 0.002), but no significant difference in length of hospitalization (3.7 ± 3.8 versus 3.0 ± 2.25 days, respectively, p = 0.203). Regarding tumor size, the MIS group had smaller tumors than those undergoing an open approach (79.9 ± 106.6 versus 215 ± 178.1 g, respectively, p < 0.04), but similar specimen size (7.6 ± 2.1 versus 8.7 ± 3.0 cm, respectively, p = 0.21). Of the MIS group, the laparoscopic approach had smaller specimen size than robotic (6.9 ± 1.9 versus 8.1 ± 2.1 cm, respectively, p = 0.03), but similar tumor size (4.6 ± 2.5 versus 6.2 ± 8.4 cm, respectively, p = 0.25). Positive margin on pathology was similar between both MIS and open (6.7% vs 25%, respectively, p = 0.1) and between robotic and laparoscopic approaches (5.6% versus 7.4%, respectively, p = 1). No association was found in PASS between the MIS and open groups (4.4 ± 3.2 versus 4.6 ± 4.1, respectively, p = 0.864) and laparoscopic and robotic subgroups (4.8 ± 3.6 versus 2.7 ± 2.1, respectively, p = 0.7).

Overall, there were a total of 46 complications (30%) throughout the cohort. 10% of these complications were high-grade (Clavien ≥ 3a). Three Clavien 4a complications were noted, two of which stemmed from postoperative myocardial infarction (MI). However, there were no Clavien 4b complications or deaths in our multi-institutional cohort. Postoperative hypertension was the most common post-adrenalectomy complication (30% of all complications) followed by pulmonary-related complications (26%), and hypotension (17%). Two adrenalectomies were converted from a laparoscopic to open approach. One of these conversions was complicated by a non-ST segment elevation MI (Clavien 4a), while the other patient’s postoperative course was complicated by respiratory failure and hypotension (Clavien 2).

Discussion

In this study of 153 patients with 156 PCs, we found that there was no significant difference in age, gender, CCMI, tumor size, or BMI among the patients that were treated with an open surgical approach to adrenalectomy compared to those treated with an MIS approach. The MIS group was found to have faster operative times (p < 0.001), less blood loss (p = 0.002), and shorter hospitalizations (p = 0.021). Overall, the MIS group had a lower perioperative complication rate of any grade than the open group (p = 0.041). However, most complications that occurred were either Clavien 1 or 2 including postoperative hypertension and minor pulmonary-related complications. Both groups had similar rates of high-grade complications.

On the other hand, patients undergoing a laparoscopic approach had different preoperative characteristics compared to robotic in terms of age, history of hypertensive crisis, cardiovascular history and BMI. The lower age and BMI seen in the laparoscopic group could be attributed to younger and less obese patients being deemed less complex surgical candidates and more amenable to a laparoscopic approach. Furthermore, the higher incidence of cardiovascular history and hypertensive crises in the laparoscopic group could be attributed to surgeon preference for the tactile feedback of laparoscopy to limit catecholamine spillage during resection. Furthermore, patients undergoing laparoscopic adrenalectomy had faster operative time compared to robotic (p = 0.004). This difference could be attributed either to surgeon experience or the docking time required for the robotic approach. Nevertheless, both laparoscopic and robotic approaches had similar rates of complications and length of hospitalization.

Our findings are congruent with the established literature on the advantages that laparoscopic adrenalectomy offers over an open approach [11–13]. In Lee et al.’s investigation of 669 patients in the Veterans Affairs National Surgical Quality Improvement Program, who underwent either laparoscopic or open adrenalectomy, they reported that open procedures had increased operative times (3.9 ± 1.8 h versus 2.9 ± 1.3 h, p < 0.001), greater length of stay (9.4 ± 11.0 days versus 4.1 ± 4.7 days, p < 0.0001), and increased 30-day morbidity rates (17.4% versus 3.6%, p < 0.0001) compared to the laparoscopic approach [12]. While these national databank analyses also provided excellent descriptions of patient-specific factors associated with perioperative complications after adrenal surgery including American Society of Anesthesiology (ASA) score and diabetes, they failed to address disease-specific data that could potentially affect perioperative outcomes. To address these shortcomings, Chen et al. analyzed 653 laparoscopic adrenalectomies performed across a 24-year period. They found that diagnosis of PC [odds ratio (OR), 4.31 (95% confidence interval (CI) 1.43–13.05), p = 0.01] and tumor size of 6 cm or greater [OR 2.47 (95% CI 1.05–5.78), p = 0.04] were independent risk factors associated with increased perioperative compilations [3]. Nevertheless, despite the risks associated with operating on PCs, our data further support an MIS approach as a safer option for adrenalectomy.

With the advent of the da Vinci system, the popularity of robotic surgery across many traditionally laparoscopic surgeries has risen in popularity. However, the feasibility and safety of robotic adrenalectomy have only been studied in small cohort studies. Brandao et al. pooled 9 studies that included 277 robotic-assisted and 323 laparoscopic adrenalectomies into their meta-analysis. They found no statistical difference in postoperative complication rate between the two groups, but noted a significantly longer length of hospitalization [weighted mean difference (WMD) − 0.43; 95% CI − 0.56 to − 0.30; p < 0.01] and higher estimated blood loss (WMD − 18.21; 95% CI − 29.11 to − 7.32; p < 0.01) in the traditional laparoscopic group [14]. In the first study comparing laparoscopic to robotic adrenalectomy for PCs, Aliyev et al. compared 88 patients, who underwent either procedure. They noted no difference in the morbidity or mortality between the two techniques, but the laparoscopic approach was associated with a longer length of hospitalization than the robotic (1.8 ± 0.1 versus 1.2 ± 0.1 days, respectively, p = 0.036) [15]. Our analysis of 130 patients, who underwent either MIS approach, further contributes to this series as evidence of the safety and efficacy of a robotic approach compared to a laparoscopic approach.

Laparoscopic resection of large tumors was historically thought to be challenging due to concerns of malignancy, technical difficulty, and increased perioperative complications. However, Agcaoglu et al. postulated that the robotic approach’s increased range of motion and three-dimensional view may provide significant advantages to adrenalectomy. In their examination of 63 adrenal tumors ≥ 5 cm, they reported lower conversion to open rate (4% vs 11%, p < 0.043), shorter operative times (159.4 ± 13.4 versus 187.2 ± 8.3 min, p = 0.043) and shorter hospitalization stays (1.4 ± 0.2 versus 1.9 ± 0.1 days, p = 0.009) using a robotic compared to a laparoscopic approach. Interestingly, they were able to achieve 28 min of time saving and zero morbidity with the robotic technique. They attributed their success to surgeons already having robotic surgery experience for other various general surgery procedures [6]. This preference for utilizing the robot for large adrenal tumors is reflected in our cohort, with patients undergoing the robotic adrenalectomy having a larger specimen size than those who underwent a laparoscopic approach (p = 0.031).

With the addition of robotic adrenalectomy over the last several years, there have been concerns with the cost of robotic surgery with some studies estimating an increase of $950 compared to the laparoscopic approach [5]. In Probst et al.’s analysis of 28 matched pairs of open and robotic adrenalectomies, the cost of robotic procedure was higher than that of open surgery. However, the overall cost of hospitalization was lower for patients in the robotic group. They attributed this overall cost reduction to the decreased length of stay associated with robotic procedures (p < 0.01) [16]. Additionally, Feng et al. compared 122 patients who underwent either laparoscopic or robotic adrenalectomy. They found that both procedures had similar relative costs, operative times, and length of stay. Their cost analysis revealed that limiting extraneous robotic instruments and surgical team experience can greatly influence the cost of robotic adrenalectomy [17].

There were several limitations in our study that need to be addressed. First, as a multi-institutional non-randomized retrospective cohort study, our study is limited by selection bias. Operative approaches and postoperative care were made by each participating institution’s surgeon after consideration of patient factors including BMI, tumor size, location and comorbidities. Surgeon’s operative proficiency and each institution’s experience in each of the three approaches further contribute to the selection bias. Second, recall bias from incomplete institutional reporting on all perioperative and pathologic variables could have also influenced the results. This can be seen in the positive margin rate in the open group. While no difference was seen between the MIS and open approaches in their rate of positive margin, only 46% of patients in the open group reported their margin rate. Lastly, the small sample size of the open group limited the power of our results.

To our knowledge, this is the first multi-institutional study investigating perioperative outcomes in patients undergoing open, laparoscopic, robotic approaches to adrenalectomies in patients with PC. Our findings further corroborate the safety and efficacy of MIS approaches and the promising potential that robotic adrenalectomy has on the management of PCs. Future randomized studies with matched controlled cohorts are needed to help further elucidate the role that robotic surgery has in adrenalectomies, specifically in patients with PCs.

Conclusion

There is growing evidence that MIS techniques offer lower perioperative morbidity and decreased length of hospitalization compared to an open approach in the surgical management of PC. Furthermore, laparoscopic and robotic adrenalectomy offer similar perioperative outcomes with robotic adrenalectomy having a place in the management of large PC. However, further studies examining the surgical approaches to PC are still needed to corroborate these findings.