Preoperative workup in the assessment of adrenal incidentalomas: outcome from 282 consecutive laparoscopic adrenalectomies

The term “incidentaloma” is widely used to define the age increasing prevalence of incidentally discovered adrenal masses in 3 to 10% of patients undergoing imaging modalities, mainly CT-scan and magnetic resonance imaging (MRI), for other clinical indications [1, 2]. These neoplasms represent the most frequent adrenal tumours encountered by physicians and they are accompanied with a negative clinical functional status, thus they are often defined as “challenging” in terms of preoperative diagnosis and management [1]. This is due to both their imaging characteristics, that sometimes do not permit the radiologists to formulate a diagnosis, and because fine needle aspiration biopsy (FNAB) does not always allow to distinguish benign from malignant primary adrenocortical tumours [3].

Despite these challenges, the goal of the clinicians dealing with an adrenal incidentaloma remains to remove functioning and malignant lesions without performing an useless adrenalectomy in non-functioning small benign tumours. In support of this approach, the most exhaustive guidelines concerning the correct management of adrenal incidentalomas have been proposed by the United States National Institute of Health (NIH) and by the American Association of Clinical Endocrinologists/American Association of Endocrine Surgeons (AACE/AAES) [4, 5]. When an adrenal incidentaloma is detected, according to these statements, the main clinical aims are to rule out the occurrence of any subclinical cortisol or aldosterone secreting lesion, to exclude a pheochromocytoma, and, finally, to resect all adrenal non-functioning lesions suspicious for adrenocortical cancer (ACC) at imaging.

If a secreting lesion is detected, the indications to treatment vary from a more aggressive attitude for pheochromocytoma to a non surgical management to be preferred in some cases of cortisol or aldosterone subclinical secreting lesions [5]. Additionally, two more points are clearly established. Firstly, adrenal masses identified as potential metastases during the staging of an extra adrenal cancer should be not considered as incidentalomas; secondly, most non secreting lesions larger than 4 cm must be removed due to the increased risk of an ACC irrespectively of their imaging status [5].

In this last decade laparoscopic adrenalectomy has gained gold standard status. It is widely used to resect almost all adrenal lesions, limiting open surgery to specific indication [6]. Although several experts have tried to define the ideal management of incidentalomas reaching the conclusion that, regardless of the surgical approach, open or laparoscopic, the indications to surgery through the years have not changed [7–10], some articles have demonstrated that the introduction of laparoscopy has caused an increase in the number of adrenalectomies performed on patients with incidentalomas [11–13]. In our study the final histology of 282 consecutive laparoscopic adrenalectomies performed at different institutions for incidentaloma, was evaluated. Patients have been then divided in two groups. Group A included patients who received a preoperative workup according to protocols [4, 5], while patients whose pre-surgical workup did not follow current guidelines were included in group B.

No deaths occurred in this audit. The general morbidity was 7.8% (22/282), see Table 1. The mean hospital stay was 4.1 days in group A patients vs. 4.6 days in group B (p = ns). Conversion rate to open surgery was significantly higher in group B (6/52 11.5%) than in group A (8/230 3.47%) (p = 0.02). Table 2 refers to group A (230/282 81.5%) whereas Table 3 reports final histology for group B (52/282 18.5%) incidental lesions. Among the lesions studied according to guidelines [4, 5], final histology matched the preoperative diagnosis for all (100%) subclinical cortisol and aldosterone secreting lesions. Remaining in the field of incidental functioning lesions it must be noted that:

However, by considering non-functioning lesions preoperatively classified as incidentaloma it was observed how:

Globally we found 1 stage I (T1, N0, M0), 9 stage II (T2, N0, M0) and 1 stage III (T3, N0, M0) cancers in group A. Conversely, 2 stage I cancers were found in group B.

In general, if we consider all non-functioning incidental lesions (group A + B) with suspicious imaging features (ACC vs. non-functioning benign adenomas) we observe that 1/89 (1.12%) lesion lesser than 4 cm, 2/34 (5.8%) lesions between 4 and 6 cm and 10/36 lesions larger than 6 cm (27.7%) were primary ACC. In general increasing dimensions were significantly associated with ACC (p = 0.001). However, if we compare lesions bigger than 6 cm with lesions between 4 and 6 cm and with lesions smaller than 4 cm, a significant correlation with cancer is shown (p = 0.03 and p = 0.01 respectively), whereas by comparing lesions between 4 and 6 cm with lesions smaller than 4 cm a significant probability of finding a cancer was not observed (p = 0.08).

This retrospective study, involving six independent surgical units, covers a period of nineteen years. For this reason it is important to highlight limits such as the lack of a central laboratory assessment for some patients, the differences in imaging quality due to technical evolution, the distinct scanning equipment that different centres have used and finally the historical absence of guidelines to treat adrenal incidentalomas before 2002. Furthermore it is strictly a surgical series, in which, if final histology is in any case provided, on the other hand no information is given about incidental lesions that were not resected. Nevertheless, evaluating the points raised in the introduction and observing the results, in our opinion some conclusions might be suggested. By considering 230/282 group A incidentalomas, we found 3/24 (12.5%) pheochromocytomas to have a borderline elevation of metanephrine levels and 2/24 lesions (8.3%) showing normal metanephrine levels, being a pheochromocytoma at final histology. Nevertheless it has to be remarked how in these group of lesions, studied according to protocols [4, 5], a preoperative diagnosis for 39/39 cortisol secreting lesions, 18/18 aldosterone secreting lesions and 19/24 (79.1%) pheochromocytomas was obtained. Thus, if we consider only subclinical functioning lesions at final histology presenting as an incidentaloma, a correct preoperative diagnosis was reached in 76/81 (93.8%) lesions. This appears especially important for pheochromocytomas. These lesions must not only be resected [1–5] but diagnosed and treated by an α adrenergic blockade before surgery, to avoid intraoperative hemodynamic instability [18, 19].

However, if we consider non-functioning incidental lesions, the preoperative detection rate decreases. In fact, although sometimes suggested, a specific imaging, CT scan or MRI, pattern for adrenal incidentalomas, allowing a preoperative diagnosis, has been never defined [10, 20]. Only 8/13 (61.5%) ACC were actually preoperatively addressed by imaging in this series. It may therefore be concluded that, despite the accuracy of preoperative evaluation, about 20.9% of incidental pheochromocytomas and 38.5% of primary cancers miss a preoperative diagnosis. An acceptable 22/24 (91.6%) and 11/13 (84.6%) preoperative detection for pheochromocytoma and primary ACC respectively would however be achieved for both functioning and non-functioning lesions if a further 3/24 borderline pheochromocytomas and 3/13 indeterminate cancers at imaging were considered. These results would also be in line with recent papers [19, 21, 22] in which both the efficacy of preoperative biochemical testing, or CT and MRI imaging, led preoperative diagnosis to match final histology in a rate ranging from 80 to 90% of resected incidental lesions.

An 18.5% rate (52/282 group B) of lesions classified as “incidental” and described in Table 3 has been on the other hand resected without an accurate preoperative study. This occurred in 49/52 patients (94.2%) treated from 1993 to 2000, when most of the centers participating in our audit were starting their laparoscopic adrenalectomy experience. This disputable attitude has however some plausible explications. The first one, as well as the lack of common endocrine coordination for the participating surgery units, was surely represented by the absence of defined guidelines to treat adrenal incidentalomas before 2002 [4, 5]. The second reason was the trend to use small incidental lesions to improve the laparoscopic adrenalectomy learning curve. This attitude, justified by the unpredictable risk of cancer in any non functioning lesion, has been however emphasized in other papers [11–13] to explain the increased number of adrenalectomies that conversely, would have been never performed before the laparoscopic age. A concept supported by the observation that if CT scan tends to underestimate the real dimension of adrenal lesions [23], on the other hand the improvement in imaging quality [11] has allowed the identification of smaller and smaller non functioning lesions. Although this aggressive surgical attitude has led to the detection of one 1/30 (3.3%) adrenocortical cancer among all lesions smaller than 4 cm, it must be highlighted the occurrence of unnecessary resection in 25/52 (48.5%) non-functioning adenomas smaller than 4 cm (Table 3); this value is significantly higher than the 63/230 (27.3%) lesion rate observed in group A (p = 0.005). This result, together with an higher intraoperative complication rate, led to a number of conversion to open surgery significantly higher in group B than in group A. In this perspective, both the overall morbidity rate of 10.9% and the mortality rate of 0.3% described for laparoscopic adrenalectomy [24], have to be reminded. Moreover the preoperative diagnosis of 1/52 (1.9%) subclinical pheochromocytoma has been missed.

While the respect of current guidelines is undisputed about the indication to resect incidental non-functioning lesions larger than 4 cm, the observation that an incorrect protocol on 52/282 patients has led to discover a small incidental cancer in 1/30 lesion lesser than 4 cm (3.3%) should cause some concern. Although our data exceeds previous papers reporting to 2% the detection rate of primary cancer coming from incidental non-functioning lesions smaller than 4 cm. [7–9, 25, 26], one instance of incidental cancer in a 2 cm lesion is probably not enough to suggest a shift in surgical indication, especially if we consider the entire series of 282 incidentalomas. In that case the cancer rate in incidentalomas smaller than 4 cm would be 1/142 (0.7%), in close agreement with the previously published range. Additionally, according to recent papers, resecting incidental adrenal lesions smaller than 4 cm appears to be not only unnecessary but also cost-ineffective [27]. However in this direction, although a wait and see policy is accepted, a clear consensus about the follow-up of these small lesions is still missing [4, 5, 21], sometimes reaching critical positions [28].

We might therefore conclude, returning to the points raised in the introduction, that a high preoperative detection rate of lesions presenting as an adrenal incidentaloma has been achieved in patients studied according to guidelines. Conversely, the unnecessary number of adrenalectomies performed in understudied patients, causing higher intra- and peri-operative morbidity, has not determined any significant change in the rate of primary adrenocortical cancer detected.