Effect of Laparoscopy on the Indications for Adrenalectomy FREE

The use of laparoscopy for adrenalectomy has rapidly increased during the past decade. The first such procedure in our hospital was performed in 1997, and by 1999 almost two-thirds of the adrenalectomies were being done laparoscopically. Laparoscopic adrenalectomy (LA) is associated with less morbidity and shorter postoperative length of stay than an open approach and is now the criterion standard for removal of most benign adrenal lesions.^1- 3 It has been suggested that LA can also be used in resection of large adrenal tumors and malignant lesions.^4- 6 At the same time as the laparoscopic revolution, there has been enormous growth in the use of computed tomography and magnetic resonance imaging, leading to a substantial increase in the number of incidentally detected, clinically nonevident adrenal masses. This trend will probably continue with improving technology, as there remains a discrepancy between the prevalence of incidentalomas noted by computed tomography and the prevalence found in autopsies.⁷

The 2002 National Institutes of Health consensus statement emphasized removing an adrenal tumor if it exceeds 6 cm, if there is suspicion of malignancy based on imaging studies, or if there is evidence of a hormone-secreting adenoma or pheochromocytoma.⁸ Other thresholds for surgery have also been suggested.^9- 12 Evidence in favor of either surgery or a conservative approach is lacking for patients with smaller lesions or subclinical hyperfunction. In many instances, the treating physician faces uncertainty about the actual diagnosis and the patient and physician must make an educated decision about proceeding, sometimes with uncertain indications. With the perceived benefits of laparoscopic resection, it may be easier for a surgeon and more acceptable for a patient to undergo adrenalectomy. There are suggestions that indications have changed, but the literature is not conclusive.^13- 17 Early intervention avoids an extensive follow-up period, reduces the number of high-cost imaging studies, and reduces psychological stress on the patient. However, it may result in overtreatment, as most incidentalomas are small, benign, nonfunctional, and not likely to grow rapidly or become functional during either short-term or long-term follow-up.^10,18- 19

The objective of this study was to identify the impact of the introduction of laparoscopy on the indications for adrenalectomy and assess the changes in the volume of procedures over time in a single academic institution.

We performed a retrospective cohort study that included all patients older than 16 years who underwent adrenalectomy for a primary adrenal lesion at our hospital from September 1, 1987, through August 17, 2007. Any patient who underwent an adrenalectomy in conjunction with another procedure or who had a nonprimary adrenal issue was excluded. Patients were divided into 2 groups. The first had surgery between September 1, 1987, and May 10, 1997. This is the period during which all adrenalectomies were performed in an open fashion. The patients in the second group were operated on from May 11, 1997, through August 17, 2007, after the introduction of LA, but this group included patients operated on in both open and laparoscopic fashion.

Electronic medical records and medical charts were used to gather information. We specifically looked at the reason for the initial imaging study (adrenal disease suspected or incidental finding), the interval (in days) between diagnosis and operation, and the size of the adrenal tumor on each subsequent scan. Since not all imaging studies reported tumor size in more than 1 dimension, we took the largest measured dimension.

To try to definitively confirm secreting tumors, we gathered laboratory tests when the information was available. We used urine and plasma catecholamine levels to establish the diagnosis of pheochromocytoma. The tests for primary aldosteronism were aldosterone plasma levels and aldosterone to renin ratio, 24-hour urine aldosterone excretion following a high-sodium diet, and aldosterone levels from venous sampling. The overnight dexamethasone suppression test, plasma cortisol level, and 24-hour urine cortisol excretion were used to diagnose hypercortisolism. Plasma testosterone level was used to diagnose hyperandrogenism. Levels considered to represent positive laboratory results are given in Table 1.

Once we ascertained the size of the tumor and assessed the possibility of hormone production, we correlated this against the type of operation (laparoscopic or open) and the final pathology report, which included the size of the tumor and the histologic diagnosis. Again, for tumor size we used the largest measured dimension. We considered the following as straightforward indications for surgery: if the tumor was larger than 5 cm on an imaging study, if the tumor was positively secreting hormones on the basis of the results of proper laboratory tests (Table 1), and if there was a strong suggestion of pheochromocytoma or malignancy based on imaging studies (hyperintense T2-weighted magnetic resonance imaging signal, irregular aspect, metastases visible, or lymphadenopathy) or results of biochemical laboratory tests. Enlargement of the adrenal mass alone during the follow-up period was not considered an indication for surgery if the tumor did not exceed 5 cm.

Categorical variables were compared between the 2 groups by χ² tests, while 2-sample t tests or Wilcoxon rank sum tests, as appropriate, were used for continuous variables. Linear and logistic regression models were used to compare outcomes between the 2 periods controlling for American Society of Anesthesiologists (ASA) class and comorbidities. A 2-sided P ≤ .05 was considered statistically significant. All analyses were conducted with SAS version 9.1 (SAS Institute Inc, Cary, North Carolina).

A total of 295 patients underwent adrenalectomy for primary adrenal disease between September 1, 1987, and August 17, 2007. Twenty-four patients were excluded because they were younger than 16 years, and 2 patients were excluded because insufficient information could be retrieved. A total of 269 patients were included in our study.

Table 2 gives the patient characteristics at the time of operation. Sixty-six patients (24.5%) were operated on up to May 10, 1997 (group 1), averaging 6.8 cases per year. Two hundred three patients (75.5%) were operated on after that date (group 2), averaging 19.8 cases per year. This represents an almost 3-fold increase in the number of adrenalectomies over the first period. The 2 groups were similar in terms of age (mean [SD], 50 [14] and 52 [14] years in groups 1 and 2, respectively) and sex (62.1% female vs 66.5% female) distributions. All operations were performed in an open fashion in group 1 compared with 71 (35.0%) of the operations in group 2. Six of the laparoscopic cases were converted to an open approach but were kept in the laparoscopic group on an intention-to-treat basis. The 2 groups differed in ASA class distribution and comorbidities. A higher proportion of patients in group 2 were in ASA class III (P = .04), and there were more patients with metabolic-endocrine or gastrointestinal comorbidity in group 2 (P = .002 and P = .008, respectively).

Table 3 shows the main outcomes. There was no significant difference in the assessed appropriateness of surgical indication between the 2 periods. In group 1, 59 cases (89.4%) had an easily identifiable indication for surgery and 7 (10.6%) had indications that did not meet our defined criteria. Group 2 had 182 cases (89.7%) with a clear indication and 21 (10.3%) that did not meet our defined criteria (P = .95). The 2 patient groups were similar in terms of tumor size and proportion of incidentally discovered tumors.

The second period was then divided into two 5-year intervals (May 11, 1997, through December 31, 2002, and January 1, 2003, through August 17, 2007). A large increase in volume, from 81 to 122 procedures, was seen between the 2 intervals. A new surgeon had been hired in 2002 and accounted for 5 of the 81 cases in the first time interval and 52 of the 122 procedures in the second. When these cases were excluded, the number of procedures was not significantly different between the two 5-year intervals, with 76 cases in the first and 70 cases in the second. There were no differences in indications or postoperative diagnosis between the first and second halves of the second period. Tumor size was significantly different between the two 5-year intervals, with a mean of 5.6 cm in the first and 4.4 cm in the second 5 years (P = .02).

In group 2, 4 patients were operated on for tumor enlargement, but these tumors did not reach the 5.0-cm threshold set as our criterion for resection. There was a significant difference in tumor size between the open and laparoscopic cohorts of group 2. The mean tumor size was 7.2 cm for open resection and 3.6 cm for laparoscopy (P < .001). Tumor size information was lacking in pathology reports of 3 patients. Two were excluded from the size analysis and one was assigned a size on the basis of the most recent imaging study. Group 2 had a longer time interval between the day the diagnosis was made on an imaging study and the day of surgery (median, 91 vs 39.5 days; P = .006). For this analysis, 4 patients were excluded because the exact date of the imaging study could not be found.

Postoperative diagnoses are shown as secondary outcomes in Table 4. There was no difference in distribution of postoperative diagnoses. Table 5 shows the final pathological findings of the patients who underwent adrenalectomy for incidentaloma. There was no difference in tumor size or postoperative diagnosis between the 2 groups for the incidentally discovered tumors. Of the lesions that were diagnosed after adrenal disease was suspected, 23 of 34 (68%) turned out to be hormonally active adenomas or pheochromocytomas in group 1 vs 61 of 85 (72%) in group 2, and there was no significant difference (P = .19).

When adjusted for the imbalanced factors (ASA distribution, metabolic-endocrine, and gastrointestinal comorbidity), the differences in indication, tumor size, reason for imaging study, and postoperative diagnosis remained not significant and the difference in the time interval between diagnosis and surgery remained statistically significant (P = .004).

Since the early 1990s, the use of laparoscopy for removal of benign and malignant adrenal tumors has steadily increased because of its superiority over open adrenalectomy in terms of morbidity and length of stay.^1- 3 These advantages, along with a better cosmetic result, have made it more acceptable for patients and physicians to take the step to surgery. The National Institutes of Health has set a threshold for surgery at 6 cm and advises surgery for hormone-producing tumors and lesions suggestive of malignancy; however, the size threshold is subject to debate.^8- 12

With improving quality and wider availability of imaging techniques such as magnetic resonance imaging and computed tomography, there has been an increase in the number of incidentally found, clinically silent adrenal masses.^7- 9 Adrenal incidentalomas are likely to be small, benign, and nonfunctional during follow-up and therefore are unlikely to meet criteria for resection.^10,17- 18 Patients and physicians must choose either surgery early in the process or extensive and costly follow-up with possible diagnostic uncertainty. Better surgical results with LA and an increased availability of laparoscopy, combined with an increased prevalence of incidentalomas, have raised concern that the threshold for adrenalectomy may have been lowered. The sparse literature available is not conclusive about possible changes in indications for adrenalectomy.^13- 17

There has been a documented increase in the number of adrenalectomies over time.^13- 17 The 3-fold increase we found in our institution is substantially larger than the national trend between 1988 and 2000.¹⁷ As the early and later groups had equal proportions of incidentalomas, we can say that part of the increase in adrenalectomies in our hospital was due to more widespread use of improved imaging techniques. Furthermore, in 2002, a new endocrine surgeon was hired, bringing his referral pattern with him and accounting for 57 adrenalectomies. While this accounts for a significant portion of the increased volume, there was still a 2-fold increase in volume without these cases. This reflects a combination of increased use of imaging and our increased growth as a referral center for endocrine disease.

Whereas we saw a vast increase in adrenalectomies over time, the feared lowering of the threshold for surgery did not occur. With data available from imaging and laboratory studies, we looked critically at the indications for surgery to determine whether more operations were performed with a questionable indication after LA became available. It is reassuring that the introduction of LA in our hospital did not lead to more unnecessary surgery and that the indications for surgical intervention did not change.

Patients in the group treated later (group 2) had a significantly longer interval between the initial diagnostic imaging and surgery. We hope this implies that, after the introduction of LA, referral for or the decision to proceed with surgery took place after a longer and more thorough workup than before. Furthermore, mean tumor size remained the same during the 2 periods. This is consistent with other studies.^13- 14 The distribution of postoperative histologic diagnoses remained the same over time.

Interestingly, we found a higher number of patients in ASA class III and more patients with endocrine-metabolic gastrointestinal comorbidity in the later group. It may be owing to the availability of laparoscopy that the surgeons in our hospital were willing to perform the operation on patients who were slightly sicker, possibly accounting for a percentage of the increase in the number of adrenalectomies.

This study has its weaknesses. In general, data would have been more accurate had we performed a prospective study instead of a retrospective one, since that would have overcome the difficulties of retrieving data from 2 decades ago. Also, to define whether there was an indication for surgery, we had to choose somewhat arbitrary cutoff values for laboratory results and tumor size on imaging. With different reference values and tumor sizes, our results may have been different; however, we were consistent and used the same cutoff values and sizes for the 2 periods. Therefore, the results are still useful for comparison. Furthermore, we primarily looked at potential changes in indications for surgery but did not look at the postoperative clinical results, and hence we cannot comment on changes in outcomes over time. Finally, the small number of patients and the large difference in size between the 2 groups make it difficult to identify small changes as significant.

The introduction of LA in our hospital occurred concurrently with a 3-fold increase in adrenalectomies, probably as a result of more widespread use of imaging modalities and the growth of our hospital as an endocrine referral center. It is reassuring to demonstrate that there was no change in the indications for adrenalectomy with the advent of laparoscopy.

Correspondence: David L. Berger, MD, Department of General Surgery, Massachusetts General Hospital, 15 Parkman St, Wang 465, Boston, MA 02114 (dberger@partners.org).

Accepted for Publication: February 9, 2008.

Author Contributions: Dr Berger had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Henneman and Berger. Acquisition of data: Henneman. Analysis and interpretation of data: Chang and Hodin. Drafting of the manuscript: Henneman. Critical revision of the manuscript for important intellectual content: Henneman, Chang, Hodin, and Berger. Statistical analysis: Chang. Obtained funding: not applicable. Administrative, technical, and material support: Hodin and Berger. Study supervision: Berger.

Financial Disclosure: None reported.

Additional Contributions: We thank Deborah McGrath, RN, for technical assistance with the study.