Endoscopic Retroperitoneal Adrenalectomy: Lessons Learned From 111 Consecutive Cases

Abstract

Objective

To evaluate the effectiveness of endoscopic retroperitoneal adrenalectomy (ERA).

Summary Background Data

Minimally invasive adrenalectomy has become the procedure of choice for benign adrenal pathology. Although the adrenal glands are located in the retroperitoneum, most surgeons prefer the transperitoneal laparoscopic approach to adrenal tumors.

Methods

Clinical characteristics and outcomes of 111 ERAs from January 1994 to December 1999 were evaluated.

Results

Ninety-five patients underwent 111 ERAs (79 unilateral, 16 bilateral). Indications were Cushing syndrome (n = 22), Cushing disease (n = 8), ectopic adrenocorticotropic hormone syndrome (n = 6), Conn’s adenoma (n = 25), pheochromocytoma (n = 19), incidentaloma (n = 11), and other (n = 4). Tumor size varied from 0.1 to 8 cm. Median age was 50 years. Unilateral ERA required 114 minutes, with median blood loss of 65 mL. Bilateral ERA lasted 214 minutes, with median blood loss of 121 mL. The conversion rate to open surgery was 4.5%. The complication rate was 11%. Median postoperative hospital stay was 2 days for unilateral ERA and 5 days for bilateral ERA. The death rate was 0.9%. At a median follow-up of 14 months, the recurrence rate of disease was 0.9%.

Conclusion

For benign adrenal tumors less than 6 cm, ERA is recommended.

Adrenal surgery entails various surgical approaches both in open and endoscopic surgery. The first endoscopic adrenalectomy was reported in 1992 by Higashihara et al. ¹ This report was followed by other early experiences with minimally invasive removal of adrenal tumors. ^2–7 Initially, the patient was placed in the supine position and the adrenal tumor was approached transperitoneally. However, laparoscopic removal of the left adrenal gland proved particularly difficult because access required retraction of surrounding organs such as the pancreatic tail, transverse and descending colon, and spleen. Therefore, most surgeons changed to the transperitoneal approach pioneered and advocated by Gagner et al, ⁸ in which the patient is in a lateral decubitus position.

At our institution, in 1995, we replaced laparoscopic transperitoneal adrenalectomy with the patient in a supine position by endoscopic retroperitoneal adrenalectomy (ERA) with the patient in the lateral decubitus position. A case-control study comparing our pilot series of ERA with conventional posterior adrenalectomy had shown that the endoscopic approach was associated with less blood loss, less postoperative pain, and reduced postoperative hospital stay. ⁹

The objectives of the present study were to evaluate the effectiveness of ERA for various pathologies and to determine its limitations after experience with 111 consecutive cases.

METHODS

The medical charts of all patients who had undergone adrenalectomy by the authors from January 1994 to December 1999 were reviewed. Patients in whom ERA had been attempted were selected for further analysis. Clinical characteristics, surgical details, and postoperative courses were recorded.

Preoperative hormonal analysis included a 1-mg dexamethasone suppression test and plasma or urinary catecholamine measurements in all patients. Plasma aldosterone and renin levels were assessed in patients with coexisting hypertension and hypokalemia. Adrenal androgens were assessed only in women with hirsutism or virilization. All patients had either computed tomography or magnetic resonance imaging of the abdomen. Patients with pheochromocytomas had I¹²³-meta-iodobenzylguanidine (MIBG) scanning to exclude extraadrenal or contralateral pheochromocytomas.

Patients with hypercortisolism were given corticosteroids during and after surgery. Patients with a known pheochromocytoma received alpha-adrenergic blockade. In those who developed compensatory tachycardia, beta-adrenergic blockade with plasma expansion when necessary was given for at least 10 days before surgery. In patients with pheochromocytomas, a pulmonary artery catheter was used to allow accurate hemodynamic monitoring of cardiac output and systemic vascular resistance during surgery. Patients with hypercortisolism resulting from Cushing’s adenomas or Cushing’s disease were given one dose of cefazolin at induction of anesthesia. A urinary catheter was installed only in patients undergoing bilateral adrenalectomy.

Endoscopic Retroperitoneal Adrenalectomy

Only patients with benign adrenal tumors less than 6 cm in diameter were selected for ERA. Previous perirenal surgery was a contraindication for ERA.

The patient was placed in the lateral decubitus position on a vacuum pad. The table was adjusted until the muscles between the costal margin and the iliac crest were taut. Through a 2-cm muscle-splitting incision in the midaxillary line, just caudal to the tip of the 11th rib, the retroperitoneal space was entered digitally. The lower pole of the kidney was palpated. A transparent disposable dissecting balloon (Origin Medsystems, Menlo Park, CA) was introduced and insufflated under visual control to create a retroperitoneal working space (Fig. 1). During insertion, the balloon was directed posteriorly to prevent penetration of the peritoneal sac. After insufflation, the balloon was deflated and replaced by a Hasson’s trocar. Carbon dioxide was insufflated to a maximum pressure of 12 mmHg and a 10-mm 0° laparoscope was introduced. On the vertebral side of the Hasson’s trocar, a 5-mm trocar was inserted close to the 12th rib. Through this trocar, a dissecting instrument was introduced. The peritoneal sac was dissected from the anterior abdominal wall to allow introduction of the third and fourth trocar on the medial side of the first trocar without perforating the peritoneal sac. All trocars were placed close to the costal margins to prevent impingement of the instruments on the iliac crest. A 0° optic was used routinely; a 25° optic was used when the view of the adrenal gland was limited.

Figure 1. Retroperitoneal endoscopic right adrenalectomy in the lateral decubitus position.

The lateroconal fascia, a thin avascular white–gray fascia that connects the peritoneum to the quadratus lumborum muscle, was opened in a craniocaudal fashion. The medial leaf of this fascia was retracted with a blunt instrument to allow identification of the upper renal pole. To expose the adrenal gland, all fat surrounding the upper renal pole was removed. After identification of the inferior rim of the adrenal gland, the ventral and lateral aspects of the adrenal gland were exposed.

In right adrenalectomy, an inferior adrenal artery arising from an upper renal artery was commonly found during dissection along the inferior rim of the adrenal. Ligation was performed close to the adrenal to avoid severing renal arterial branches, which can cause hypertension. The short right adrenal vein was exposed after dissection of the caval vein along the medial side of the adrenal. The adrenal vein was clipped and transected. In some patients the adrenal vein drained into an inferior hepatic vein. In several patients a second adrenal vein draining into the caval or hepatic vein was encountered. Middle adrenal arterial branches passing behind the caval vein to the adrenal gland and superior arterial branches descending from the phrenic artery were either cauterized or clipped.

In left adrenalectomy, dissection was performed in a similar fashion. Arterial adrenal branches were found inferiorly, medially, and superiorly. The left adrenal vein was exposed during dissection along the lower adrenal rim, ligated, and clipped.

For both left and right adrenalectomy, ligation of the adrenal vein took place in the final phase of the procedure. The adrenal gland was removed in a plastic bag to prevent tearing the brittle adrenal tissue. A drain was left behind only in case of diffuse oozing.

When conversion was necessary, we performed either lumbotomy or thoracolaparotomy through the ninth intercostal space, with the patient remaining in the same position.

Patients were allowed a normal diet and were mobilized on the first postoperative day. Patients were discharged from the hospital when they did not require intravenous medication, tolerated a normal diet, and were mobile. Follow-up data were collected through medical correspondence.

Analysis

Body mass index was determined by the quotient of body mass (kg) and the square of body length (m). Surgical time was defined as the period from the first incision to complete closure of all incisions. Hemodynamic instability was defined as systolic blood pressure exceeding 200 mmHg for at least 2 minutes and heart rate exceeding 130 beats per minute for at least 2 minutes. Postoperative stay was calculated by assigning the day of surgery as day 0. Wound infection was defined as a condition that required surgical drainage. Malignancy of adrenal tissue was defined as a van Slooten index greater than 8. ¹⁰

The mean or median values and standard errors of the continuous variables were calculated. Analysis of patients who required conversion to open adrenalectomy was performed on the basis of intention-to-treat. The Mann-Whitney test for nonparametric variables was used to determine differences between left, right, and bilateral procedures. P < .05 was considered to represent a significant difference.

RESULTS

From January 1994 to December 1999, we performed adrenalectomies in 141 patients: 95 ERAs, 11 laparoscopic transperitoneal adrenalectomies, and 35 open adrenalectomies. The distribution of approaches during the study period is shown in Figure 2. In the first minimally invasive adrenal surgeries, laparoscopic transperitoneal adrenalectomy was performed with the patient in a supine position. The left adrenal gland was approached through the transverse mesocolon laterally to the inferior mesenteric vein, which had been the standard conventional anterior open approach to adrenal tumors of moderate size. ¹¹ However, the laparoscopic transmesocolic approach to the left adrenal gland proved difficult because simultaneous retraction of both the pancreatic tail and descending colon was necessary to expose the adrenal gland. ⁷ Therefore, this approach was abandoned in 1995 and ERA was started. Laparoscopic transperitoneal adrenalectomies with the patient in the lateral decubitus position were performed in only two patients in 1999. One patient with food-dependent Cushing syndrome had a 7.5-cm right adrenal tumor and a 5-cm left adrenal tumor. The right adrenal was removed by a laparoscopic transperitoneal approach, the left by an endoscopic retroperitoneal approach. The other patient who had a transperitoneal adrenalectomy for a left 4-cm Cushing’s adenoma had previously undergone lumbotomy for a renal concrement.

Figure 2. Number of open, transperitoneal laparoscopic, and retroperitoneal endoscopic adrenalectomies, 1994–1999.

Seventy-nine patients with unilateral disease and 16 patients with bilateral disease underwent ERA. Patient and adrenal tumor characteristics are listed in Table 1. Conn’s adenoma was the most prevalent indication. Bilateral adrenalectomy for pheochromocytoma was performed in one patient with multiple endocrine neoplasia (MEN) IIA syndrome, one patient with MEN IIB syndrome, and one patient with von Recklinghausen’s disease. Subtotal adrenalectomy was not possible because grossly normal adrenal tissue was not present in these three patients.

Table 1. PATIENT CHARACTERISTICS

ACTH, adrenocorticotropic hormone.

One 19-year-old woman with MEN IIB syndrome and another 17-year-old patient with von Hippel-Lindau’s disease underwent unilateral adrenalectomy. The contralateral adrenal gland was left in place in both patients because MIBG scintigraphy did not show accumulation of the radiopharmaceutical, and a computed tomography scan showed it to be of normal size. At follow-up after 15 and 6 months, respectively, there was no sign of disease recurrence in either patient.

Two patients had extraadrenal pheochromocytomas (paragangliomas). In these patients, the normal adrenal gland was removed in continuity with the paraganglioma.

Of all adrenal tumors removed endoscopically, two exceeded 6 cm at gross examination. A 7-cm cortisol-producing adrenal tumor in one patient turned out to be malignant. Another patient with MEN IIA syndrome had a left 8-cm pheochromocytoma. In both instances, the size of the adrenal tumors was underestimated on preoperative imaging. Adrenal tumors as small as 0.1 cm were found in patients who underwent bilateral adrenalectomy for Cushing’s disease.

The size of the incidentalomas varied from 2.5 to 6 cm. Although surgery for incidentalomas is recommended only in patients with adrenal tumors exceeding 4 cm in diameter, three incidentalomas of 2.5, 3.0, and 3.5 cm were removed. The indication for adrenalectomy in these patients was the patient’s concern about the nature of the enlarged adrenal.

Two patients with Cushing’s disease had only unilateral adrenalectomy by the endoscopic retroperitoneal approach because of mild Cushing’s disease after unsuccessful pituitary surgery or irradiation. Six patients had Cushing syndrome resulting from ectopic adrenocorticotropic hormone (ACTH) production from carcinoid tumors (n = 3), neuroendocrine pancreatic tumor (n = 1), and unknown primary tumors (n = 2).

Unilateral adrenalectomy required less than 2 hours; bilateral adrenalectomy lasted 3.5 hours (Table 2). There were no significant differences in either surgical time or blood loss between left and right adrenalectomy. Hospital stay was significantly longer after bilateral adrenalectomy because of medical correction of endocrine imbalance (P = .013). There was no correlation between body mass index and surgical time, blood loss, or hospital stay.

Table 2. SURGICAL DETAILS

Hemodynamic instability was observed only temporarily in 3 of the 19 (15%) patients with pheochromocytoma; it was treated by nitroprusside infusion without negative sequelae.

Peritoneal tears occurred in one third of all patients, particularly during the first 40 cases. Although peritoneal perforation to some extent reduced the working space, conversion was never necessary. In some cases, an additional instrument was introduced to provide retraction of the peritoneal sac.

All but four patients had complete removal of the affected adrenal gland. Partial adrenalectomy was purposely performed in two patients with Conn’s adenomas, in one patient with a 3-cm Cushing’s adenoma, and in one patient with an adrenal cyst. One of these patients had agenesis of the contralateral kidney with suspected concomitant absence of the adrenal gland. After partial adrenalectomy, this patient was cured of disease and did not require substitution of adrenocortical hormones. In the other patients, partial adrenalectomy was performed because the lesion was located at the periphery of the adrenal gland and the remaining gland was macroscopically normal.

Two patients underwent ERA during pregnancy. One patient had an uneventful bilateral procedure for MEN IIA syndrome at 18 weeks of gestation and delivered a healthy baby after 40 weeks. The other pregnant patient had a 6-cm pheochromocytoma diagnosed at 28 weeks of gestation. At 35 weeks, a healthy baby was delivered through cesarean section, followed by endoscopic removal of the right pheochromocytoma. As suggested by preoperative magnetic resonance imaging, this involved a paraganglioma located between the renal artery and vein, necessitating lumbotomy.

Five ERAs were converted (4.5%). Bleeding from excessive traction on the right adrenal vein was the cause of conversion in one patient with a left pheochromocytoma and in another patient with Cushing syndrome. The conversions for bleeding occurred among the first 30 cases. The two procedures in patients with paragangliomas were converted to thoracolumbotomy because lumbotomy did not allow sufficient access. One paraganglioma was located between the aorta and the caval vein; the other was found between the artery and vein of the right kidney. Both the location and the presence of multiple short venous branches into the caval and renal vein necessitated open surgery. The remaining conversion was for unexpected malignancy in a patient with a 5-cm left incidentaloma. Several enlarged lymph nodes were encountered during adrenal dissection that contained malignant tissue on perioperative microscopy. The endoscopic procedure was converted to lumbotomy to allow periaortic lymph node dissection.

The total complication rate was 11%. Hematoma (n = 3) and urinary tract infection (n = 3) were the most common complications. One hematoma was drained under ultrasonographic guidance for persisting pain in the surgical field. Other complications included paralytic ileus after lumbotomy, depression, and candidemia. This patient had near-terminal asthmatic disease, precluding open surgery, and developed a hematoma after bilateral adrenalectomy for severe Cushing syndrome resulting from ectopic ACTH production by an unknown primary tumor. The hematoma spread to the peritoneal cavity as a result of a laceration of the peritoneal sac during right adrenalectomy and became infected by Candida. Despite systemic fungicide treatment and surgical abdominal lavage, this patient died on the 20th postoperative day. This was the only postoperative death (0.9%).

Pathology and Follow-Up

In one patient, microscopic examination of the adrenal tissue revealed a van Slooten index of more than 8. This patient had a cortisol-producing tumor with a diameter of 7 cm. Adjuvant mitotane treatment was not given because capsular and vascular invasion was absent. Although at 6 months of follow-up the abdominal computed tomography scan was normal, local recurrence and bone metastases were found 3 months later. Port-site metastases were not present. The patient died 14 months after adrenalectomy.

In one patient with a 5-cm left incidentaloma, adenocarcinoma was found in the resected adrenal and lymph nodes. It was not clear that this tissue was an adrenocortical cancer. Enlarged lymph nodes in the retroperitoneum and mediastinum and a pulmonary lesion were found at follow-up, suggesting that the incidentaloma was a metastasis of an unknown primary adenocarcinoma.

Median duration of follow-up was 14 months (range 2–68). All patients who had Conn’s adenomas removed were normokalemic without potassium-preserving medication. Eighteen of the 26 patients (69%) had normal blood pressure without medication. Recurrence of disease was not observed in any of the patients with pheochromocytomas.

Two of the six patients with hypercortisolism induced by ectopic ACTH died, one of perforated diverticulitis at 11 months and the other of hepatic failure resulting from extensive metastases at 14 months.

Open Adrenalectomy

In 1994, we performed the first minimally invasive adrenal procedures. During 1994, 13 open adrenalectomies were performed for pheochromocytomas, bilateral disease, and adrenal tumors larger than 6 cm. With growing experience, pheochromocytomas and bilateral adrenal disease were also approached endoscopically. However, adrenal tumors exceeding 6 cm were not considered for endoscopic retroperitoneal surgery.

Twenty-two patients underwent open adrenalectomies from 1995 through 1999. Eight of these patients had adrenocortical cancers with diameters from 8 to 20 cm. Five adrenal tumors larger than 7 cm and suspected to be malignant were benign on microscopic examination of the specimen. Three patients had sporadic pheochromocytomas with diameters of 7, 10, and 15 cm. The remaining six patients had Cushing’s adenoma (7 and 9 cm), adrenal metastasis of hepatocellular cancer, adrenal metastasis of renal cancer, Castleman’s disease, and a 5-cm pheochromocytoma with spontaneous hemorrhage.

DISCUSSION

In less than 6 years, the adoption of endoscopic techniques has greatly changed our approach to adrenal pathology. At present, only 20% of adrenal tumors are still removed by open surgery at our institution. In our opinion, the indication for open surgery is still an adrenal tumor suspected to be malignant. Radical resection of the tumor is of paramount importance in the surgical treatment of adrenal cancer. Considering that most adrenal cancers are large tumors, the retroperitoneal space does not provide sufficient room for safe removal. The role of transperitoneal laparoscopic removal of adrenal cancers remains controversial. Because Gajraj and Young ¹² have demonstrated clearly that the risk of malignancy is 1 in 10,000 for adrenal tumors less than 6 cm, we have adopted 6 cm as the upper limit of adrenal tumors amenable to a retroperitoneal endoscopic approach. However, primary adrenal cancers as small as 3 cm have been described. ¹³ Therefore, we cannot rely solely on tumor size to differentiate between benign and malignant adrenal tumors. In our opinion, adrenal tumors larger than 6 cm can be approached safely using minimally invasive techniques when imaging studies have shown homogeneous density of the tumor, a smooth surface, and lack of invasion of surrounding tissues.

For tumors larger than 6 cm without malignant features, we advocate a transperitoneal approach, with the patient in a lateral decubitus position in accord with the technique described by Gagner et al, ¹⁴ because the retroperitoneal space is too small to conduct safe dissection of large tumors.

Recurrence of adrenal cancer occurred in one patient after endoscopic removal of a 7-cm corticosteroid-producing adrenal tumor. It remains under debate whether the endoscopic technique contributed to cancer recurrence. Because the adrenal tumor had not been breached, either macroscopically or microscopically, and the tumor had been placed in a plastic bag for extraction, direct implantation of tumor cells in the surgical field appeared unlikely. To our knowledge, three other cases of recurrence of adrenal cancer after laparoscopic adrenalectomy have been reported. ^15–17 Hamoir et al ¹⁵ described a patient with a 12-cm androgen-producing right adrenal tumor that had been removed through a subcostal laparotomy after 4 hours of troublesome laparoscopic dissection. Diffuse peritoneal adrenal cancer was found at follow-up. A second laparotomy was undertaken to excise all tumor deposits and to install intraperitoneal chemotherapy, which was followed by intravenous chemotherapy. Foxius et al ¹⁶ observed peritoneal carcinomatosis 6 months after laparoscopic removal of a right 2.7-cm aldosterone-producing adrenal lesion. Despite splenectomy and omentectomy followed by mitotane chemotherapy, the patient died 20 months after surgery. Suzuki et al ¹⁷ reported a similar case of abdominal dissemination 19 months after laparoscopic adrenalectomy for a cortisol-producing adrenal tumor. We are not aware of reports of port-site metastases from adrenal cancer. The advantages of ERA are a limited need to retract the liver, spleen, or pancreas to expose the adrenal gland and the avoidance of the peritoneal cavity. Transperitoneal adrenalectomy with the patient in the lateral decubitus position requires mobilization of the right lobe of the liver on the right side and detachment of the spleen from the lateral abdominal wall on the left. Both liver and spleen are retracted during transperitoneal adrenalectomy. Although bleeding from tears in the liver can usually be solved by compression, splenic injury necessitated splenectomy in two patients among 60 laparoscopic transperitoneal donor nephrectomies and 11 laparoscopic transperitoneal adrenalectomies at our institution.

Avoidance of the peritoneal cavity is a clear advantage of the retroperitoneal approach and is valuable in patients who have undergone previous abdominal surgery. Another indication for the retroperitoneal approach is the presence of an enlarged liver, which makes retraction difficult in patients with right adrenal tumors. Studies with longer follow-up are required to establish whether retroperitoneal adrenalectomy leads to fewer late bowel obstructions and less chronic abdominal pain from intraperitoneal adhesions than transperitoneal adrenalectomy.

The first reports on successful ERA appeared in 1995. ^18,19 Both recommended placing the patient in a prone position. The advantage of the prone position is that bilateral adrenalectomy can be performed without turning and repositioning the patient. However, conversion to open surgery, particularly in case of profuse bleeding, is more complex in the prone position than in the lateral decubitus position. In our experience, turning and redraping the patient does not require more than 10 minutes. After initial experience with balloon dissection, Walz et al ^19,20 modified their technique and now create the retroperitoneal space by blind digital dissection. Although this method is less expensive than balloon dissection, we still use a balloon because it provides rapid creation of the retroperitoneal space without anatomical distortion.

The anatomy of the retroperitoneal space has been considered to provide fewer landmarks than the intraperitoneal space. However, the lateroconal fascia, the quadratus lumborum muscle, and the kidney provide sufficient guidance to find the adrenal gland. Removal of the fat surrounding the upper renal pole is essential to expose the adrenal gland. In our experience, perioperative ultrasonography has never been necessary to help localize the adrenal gland. A disadvantage of ERA is the limited maneuvering space. Proper patient positioning, sufficient mobilization of the peritoneal sac, and avoidance of bulky 10-mm instruments are necessary to provide good visualization of the surgical field. In our series, no single endoscopic procedure required conversion for insufficient exposure.

In ERA, the adrenal vein is usually ligated in the last phase of the procedure. This appears to conflict with the traditional dogma for pheochromocytomas that the adrenal vein should be ligated before manipulating the pheochromocytoma. However, serious hypertensive crises or cardiac arrhythmias were not encountered in 19 patients undergoing surgery for pheochromocytomas. Timely preoperative alpha- and beta-adrenergic blockade and limited manipulation of the pheochromocytoma appear sufficient to prevent perioperative hemodynamic instability. Another concern about the endoscopic treatment of pheochromocytomas is the unrecognized presence of pheochromocytoma in the contralateral adrenal gland or along the aortic axis. However, MIBG scintigraphy combined with either computed tomography or magnetic resonance imaging of the abdomen precisely localizes multiple pheochromocytomas. ^21,22 Even in 1993, at the brink of endoscopic adrenalectomy, the open extraperitoneal flank approach to pheochromocytomas was being advocated. ²³

Subtotal adrenalectomy has been advocated by Walz et al ²⁰ to preserve adrenal cortical tissue. This is particularly important in patients with bilateral disease because supplementation of adrenal hormones can be avoided, and Addison’s crises are potentially lethal. Although most endocrine surgeons favor complete adrenalectomy, other endocrine disorders such as parathyroid adenomas and solitary adenomas can be treated successfully by removing only the lesion.

Several studies comparing various endoscopic approaches to adrenal pathology have been performed. Fernandez Cruz et al ²⁴ performed the only randomized study in patients with hypercortisolism undergoing either transperitoneal or retroperitoneal endoscopic adrenalectomy. Transperitoneal adrenalectomy was associated with increased hypercapnia and more profound hypertension. In another comparative study by Fernandez Cruz et al, ²⁵ surgical time, blood loss, postoperative pain, and hospital stay were similar after either transperitoneal or retroperitoneal endoscopic adrenalectomy. Duh et al ²⁶ studied 23 transperitoneal laparoscopic adrenalectomies with the patient in the lateral decubitus position and 14 posterior endoscopic adrenalectomies with the patient in the prone position. Surgical time was approximately 3.5 hours in both groups, and the hospital stay was similar for both approaches. Baba et al ²⁷ assessed the efficacy of 33 transperitoneal laparoscopic adrenalectomies with the patient in the supine position versus 5 ERAs in the lateral decubitus position and 13 ERAs in the prone position. Only three trocars were required for the retroperitoneal prone approach, whereas at least four trocars were used for the others. Surgical time and blood loss were less for the retroperitoneal approaches. In one of our earlier studies, transperitoneal laparoscopic adrenalectomy with the patient in the supine position was compared with ERA in the lateral decubitus position. ²⁸ Surgical time, blood loss, and hospital stay were significantly less with the retroperitoneal approach. Therefore, ERA appears preferable to the transperitoneal approach with the patient in the supine position. The surgical time and blood loss in the present series are in the same range as reported by Gagner et al ¹⁴ in a study of 100 consecutive transperitoneal laparoscopic adrenalectomies with the patient in the lateral decubitus position. However, all comparative studies involved consecutive series, except for the randomized study of Fernandez Cruz et al. ²⁴ Therefore, the technique used later in the experience of the authors would benefit from their experience with the techniques used initially.

Conclusions about the preferred endoscopic approach to adrenal pathology can be drawn only based on a prospective trial randomizing patients to the transperitoneal or retroperitoneal approach, with experienced surgical teams performing both procedures. ERA appears safe and effective in patients with benign adrenal lesions smaller than 6 cm, but expertise in adrenal surgery and thorough knowledge of retroperitoneal anatomy are required.

Discussion

Prof. A. Fingerhut (Poissy, France): This is a large series, probably one of the largest series of retroperitoneal adrenalectomy in the world. I congratulate you on that. If I can have my slide, please. I will try to show you the theoretical differences between the transabdominal and the retroperitoneal laparoscopic approaches. I would just like to go through them and ask a few questions. Better landmarks are an advantage in the transabdominal approach, because you go through the abdomen with a familiar anatomy; in the retroperitoneal approach, you might have a problem. Some people have advocated ultrasound to find small tumors, and I did notice in the paper, that there were at least two tumors that were 1 and 2 mm in diameter. Another advantage is better vascular control. You did mention that when you converted, you went through the lumbar space. I admire that because of the shortness of the right adrenal vein. That was very good, but I would like to have a comment on that. My third question concerns the 6-cm cutoff value you stated. Why 6 cm? In the literature it is 5, so why not 7? Just a comment on that, please. And last but not least, you use a balloon, and this might have the same criticism which we encounter in the extraperitoneal hernia repair. Does it increase costs? Is it worth it? Is it really necessary?

Prof. H. Bonjer (Rotterdam, The Netherlands): To go back to the landmarks, we recently reviewed an interesting survey of general surgeons and urologists. General surgeons have a preference for the transperitoneal route, and urologists prefer the retroperitoneal route because they are more familiar with it. So I think it is a matter of familiarity. You certainly need some time to get used to it. But if you stay close to the kidney and tease away the fat surrounding the superior pole, you will find the adrenal gland. The control of the adrenal vein is really straightforward when the patient is in lateral decubitus position, because you already have created your space. You only have to cut in between the ribs and you are right there. For the really large adrenal tumors, we always use a thoracolumbotomy to get good access to the caval vein and to the adrenal vein. Concerning the 6-cm limit, you have got to draw a line somewhere, and as you can read in the article, we also violated this rule. We have had a few tumors of 7 or 8 cm. For large adrenal tumors, you should certainly move to the transperitoneal approach, and it is also our preferred approach to laparoscopic donor nephrectomy. We do not do that extraperitoneally.

The balloon, I agree, is a costly device. There are semidisposable balloons that are less costly. We have done it without the balloon, but it is far more difficult. This is a type of procedure performed by our senior surgical residents, and for the teaching situation we take the cost.

Prof. C. Proye (Lille, France): Your technique is appealing although very demanding, as I could witness when you visited us in Lille. Laparoscopic adrenalectomy is straightforward on the right side but can be painstaking on the left side, where small tumors are embedded in the fatty tissue overlying the renal fossa, and we know that the left adrenal lies immediately at the anterior aspect of the posterior abdominal wall. So maybe the retroperitoneal approach could be better than the laparoscopic approach on the left side, at least in obese patients.

It is obviously already the preferred approach in patients with multiple previous abdominal operations and, as in your experience, in cases who are pregnant, especially in the multiple endocrine neoplasia type II setting.

Prof. Bonjer (Closing Discussion): I completely agree with you. The most important thing in endoscopic adrenal surgery is to put the patient in lateral decubitus position, and I agree with you that the transperitoneal route on the right-hand side is very straightforward as well. On the left-hand side, you have to mobilize either the splenic flexure or the spleen, and then I think there is a very distinct advantage for the retroperitoneal access.