Abstract
Background
Adrenal sarcomas are rare malignant tumors with structural and clinical similarities to sarcomatoid adrenocortical carcinoma. Preoperative diagnosis of tumors of the adrenal gland can be challenging and often misleading thus detaining patients from appropriate oncological strategies.
Objective
This analysis of a case series evaluated the predictive capability of the primary clinical diagnosis in case of malignancies of the adrenal gland.
Methods
Thirty two patients were treated from 2009 to 2015 at our clinic and analyzed retrospectively. All patients had computed tomography and/or magnet resonance imaging and a primary histopathological examination at our institution after surgery. Ten questionable cases were surveyed by a reference pathologist.
Results
Twelve out of 32 diagnoses had to be revised (37.5%). Only 15 out of 24 tumors primarily classified as adrenocortical carcinoma were finally described as primary adrenal cancer. We found two leiomyosarcomas, one liposarcoma, one sarcomatoid adrenocortical carcinoma, and one epitheloid angiosarcoma among 12 misleading diagnoses. Other tumors turned out to be metastases of lung, hepatocellular, and neuroendocrine tumors. Larger tumors were significantly more often correctly diagnosed compared to smaller tumors. Four patients of the group of revised diagnoses died whereas all patients with confirmed diagnoses survived during the follow-up.
Conclusion
Preoperative assessment of tumors of the adrenal gland is still challenging. In case of wrong primary diagnosis, the prognosis could be impaired due to inadequate surgical procedures or insufficient preoperative oncological treatment.
Background
Cancer of the adrenal gland is an extremely rare malignancy, with an incidence of 0.5–2 cases per million per year in the general population. The sarcomatoid adrenocortical carcinoma was first mentioned in 1987 1 . Irrespective of this specific type, adrenocortical carcinomas mixed with sarcoma or sarcoma-like components have been reported until now2,3. These tumors are poorly differentiated and badly aggressive with a high risk for loco-regional recurrence, rapidly growing metastases and death occurring within a short period of time 3 . According to the World Health Organization (WHO), variants of adrenocortical carcinomas (ACC) include oncocytic ACCs, myxoid ACCs, and ACCs with sarcomatous areas4‒6. Tumors showing both, carcinomatous and sarcomatous characteristics (so-called carcinosarcomas) are generally uncommon 7 . Due to the current WHO classification (2017), these neoplasms have been categorized as “sarcomatoid carcinomas” covering all carcinomas with “pleomorphic, sarcomatoid, or sarcomatous elements”3,7,8. Apparently, the sarcomatous part develops out of carcinoma by dedifferentiation 3 . Sarcomas of spindle cell type may develop from pheochromocytoma9,10. Others show a myofibroblastic type 11 . Some adrenal sarcomas, the epitheloid angiosarcoma, histologically resemble adrenal cancers and have to be distinguished from carcinomas by immunostainings12,13. Other primary adrenal sarcomas like leiomyosarcoma or liposarcoma are very rare3,9,14‒17. Anaplastic and sclerosing Kaposi’ssarcomas can develop in the adrenals with or without immunodeficiency18‒21. Alveolar sarcomas of the adrenals appear in children whereas adrenal Ewing sarcomas can be found in patients of any age22‒25. Finally, synovial sarcomas of the adrenal have been described 26 . Sarcomatoid adrenocortical carcinomas should strictly be separated from sarcomas of the adrenal gland, as they require different treatment.
Unfortunately, the presurgical diagnosis of adrenocortical cancer (ACC) is difficult and that is why they are often detected in an advanced stage 7 . According to literature, 60‒80% of ACC are functional and hypercortisolism can be found up to 60% of patients with ACC7,27‒30. Radiologic differentiation of ACC from a sarcomatoid carcinoma or sarcoma remains difficult in patients without elevated corticosteroids 31 . Contrast-enhanced computed tomography (CT) indicates heterogeneous enhancement with areas of high- and low-attenuation (55‒69 HU and 28‒30 HU) within the tumor 31 . As a preoperative biopsy is obsolete in ACC, in most patients the correct diagnosis is made postoperatively.
Furthermore, even the postoperative pathological diagnosis of ACC is challenging 32 . The Weiss scoring system, which was introduced 27 years ago and represents the most popular scoring system today, is based on the verification of at least three out of nine morphological parameters by microscopy6,32.
Due to the aggressive behavior, ACC has a poor prognosis, which underlines the importance of identifying diagnostic and prognostic markers to rule out other differential diagnosis especially a sarcoma 33 . As mentioned, a preoperative biopsy should be avoided in ACC, but could be useful in sarcomas to decide if neoadjuvant chemotherapy is indicated. Due to the lack of reliable diagnostic markers, the histopathological diagnosis after successful surgery may differ from the preoperative assumption causing a precious loss of time while the patient does not receive adequate therapy influencing outcome and survival.
The aim of the present study was an analysis of the frequency of misdiagnosis of adrenal tumors in preoperative imaging and its impact on outcome and tumor recurrence. Primary endpoint was the overall survival. Secondary endpoint was the time until tumor recurrence and evaluation of clinical risk factors.
Methods
We present retrospective data of a series of 32 patients who had an adrenalectomy at our clinic from September 2009 until the end of 2015. Inclusion criteria were a simple or multi-visceral adrenal resection due to preoperatively suspected malignancy. All patients underwent preoperative imaging and/or additional biopsy. We identified 10 questionable cases out of 32 patients and revised by a reference pathologist to especially identify sarcomatoid ACCs. We defined the clinical diagnosis as the radiomorphologic diagnosis only based on the preoperative imaging. The first histopathological diagnosis was defined and described in the first record of the institute of pathology of the Charité. The final diagnosis was derived from either the final report from the institute of pathology of the Charité or from the statement of the reference pathologist if this report was requested and. Other clinical parameters evaluated in this study were gender, age at diagnosis, size of tumor, site of tumor, hormonal activity, preoperative imaging (CT or magnet resonance imaging (MRI)), confirmed infiltration due to preoperative imaging, biopsy results, operating procedure (adrenalectomy only or multi-visceral resection), Weiss score, adrenal origin confirmed by pathology, metastases, loco-regional recurrence, survival until recurrence, and overall survival. Patients (n = 32) were grouped to those, whose clinical diagnosis was finally confirmed (CONFIRMED, n = 20) and those whose diagnosis was ultimately revised (REVISED, n = 12). The data were retrospectively analyzed using our digital patient documentation system. Statistical analysis was performed by IBM SPSS Statistics® (Version 23, Armonk, NY, USA). Significances of non-parametric data were tested by Mann–Whitney-U-test. Variances of nominal data were tested by cross tables and Fisher’s-exact-test. Kaplan–Meier-analyses were performed for survival analysis. Level of significance was defined as p < 0.05.
Results
We identified 32 patients, who underwent adrenal surgery due to suspected adrenal carcinoma at our clinic from September 2009 until December 2015. Histopathology confirmed the preoperative diagnosis in 17 patients (CONFIRMED, 53.1%), whereas the diagnosis had to be revised in 15 patients (REVISED, 46.9%). A reference pathologist (W.S.) retrospectively reviewed 10 questionable cases. Five out of those 10 histological diagnoses (50.0%) were defined by the reference pathologist.
Table shows the preoperative clinical diagnosis, the first histological result and the final histological diagnosis of group REV (n = 15). Each line replaces one case.
Demographic Data for the total cohort, and both groups CONFIRMED and REVISED.
(Absolute numbers and percentage, median, minimum and maximum).
N = 32.
Analysis was performed for the groups CONFIRMED and REVISED without any significant difference concerning demographic data, except for tumor size (Figure 1). There was a trend towards a longer recurrence-free survival in the group CONFIRMED (53.7 months, 95% CI: 29.5–77.9) compared to REVISED (35.1 months, 95% CI: 16.4–53.8), but without significance (p = 0.406, Figure 2). The estimated overall survival of CONFIRMED was 99 months (95% CI: 87.6–110.0) and 75.7 months within REVISED (95% CI: 52.5–98.9, p = 0.107, Figure 2). This difference was not statistically significant. Boxplots. Tumor Size in (mm) according to the groups. Left: Kaplan–Meier-Analysis. Estimated recurrence-free survival in the patients, whose primary clinical diagnosis was confirmed, and the patients with revision of the diagnosis (n = 32). Right: Kaplan–Meier-Analysis. Estimated overall survival grouped to those patients, whose primary clinical diagnoses was confirmed, and those, whose primary diagnosis changed during histopathological work-up (n = 32).
Demographic Data of patients with adrenocortical carcinoma and with retroperitoneal sarcoma (SARCOMA).
(Absolute numbers and percentage, median, minimum and maximum).
N = 32.
Discussion
Histopathology could not confirm preoperative diagnoses in 47% of patients with adrenal malignancy or malignancies involving the adrenal gland. Fifty percent of questionable pathological reports, which were seen by a reference pathologist were revised after reevaluation. Twelve out of 23 clinically diagnosed ACC were misdiagnosed and turned out to be sarcomas in four cases. One ACC was not detected before resection (Table 1). We must note, that even in specialized centers, large retroperitoneal tumors within or close to the adrenal gland might be misdiagnosed preoperatively. The discrimination of these tumors seems to be difficult, a fact, which may lead to delayed or inadequate oncological therapy.
Clinical and differential diagnoses of retroperitoneal tumors are challenging as the retroperitoneum hosts a wide range of heterogeneous lesions 34 . The median age of onset of ACCs and liposarcomas is in the fifth and sixth decade whereas leiomyosarcomas normally appear in elderly patients28,34,35. ACC’s have a genetic predisposition in seldom cases and are associated with the germline mutation of TP53 and therefore the Li Fraumeni syndrome as well as the Beckwith–Wiedemann syndrome in pediatric patients 28 . In adult patients, the Multiple Endocrine Neoplasia I, Lynch syndrome, familial adenomatous polyposis, neurofibromatosis type I and the Carney complex bare a higher risk for ACCs 28 . Sarcoma do not occur hereditarily but can be associated with immunodeficiency 35 . In general, the clinical presentation of retroperitoneal tumors besides functional ACCs is uncharacteristic, depends on its localization and may occur after a certain latency period, in which its size had grown enough to provoke severe sight effects by compression, displacement, or invasion of the surrounding structures34,36. Functional ACCs show the clinical image of hypercortisolism in up to 80% with a glucocorticoid-mediated mineralocorticoid receptor activation followed by hypokalemia, hypertension, and clinically severe muscle weakness and a Cushing’s syndrome28,29. Adrenal androgens are the second most secreted hormones by ACCs leading to hirsutism, virilization, and menstrual irregularities28,29. However, 20–40% of ACCs are non-functional aggravating the differentiation from other retroperitoneal lesions which all cause similar unspecific symptoms28‒30. Those complaints might be discomfort, abdominal pain, weight loss, the palpation of the tumor mass, and in case of a close localization to the aorta an unspecific abdominal pulsing27,28,36. In this study, a third of all revised cases and a quarter of the confirmed cased were functional active tumors (Table 2). Nevertheless, hormone diagnostic among the revised can be misleading especially as values were not strongly elevated and even three pheochromocytomas were not clearly detecting. Furthermore, the clinical presentation of this questionable cases was also unspecific.
Therefore, a detailed hormone work-up, CT, and MRI are mostly able to differentiate between adrenal adenomas, non-adenomas, or a further retroperitoneal entity34,36‒40. ACCs are normally large tumors (>4 cm), occur in 2 to 10% bilaterally and show a heterogeneous enhancement in CT-scan (>10 HU) because of internal hemorrhage, necrosis, and calcifications28,41‒43. In contrast, adrenal adenomas are smaller (<4 cm), homogenous (≤10 HU) without calcification, necrosis or hemorrhage showing a greater wash-out than ACCs 28 . The parenchyma of ACCs appears isointense to hypointense on T1-weighted MRI images and hyperintense on T2-weighted MRI images compared to the liver28,42. If the imaging shows intralesional fat signs, that might be suspicious for liposarcomas, extramedullary hematopoiesis, and teratoma as differential diagnosis 34 . Calcifications are associated with leiomyosarcomas, teratomas, undifferentiated pleomorphic sarcoma, or neurogenic tumors and a hyperintense signal in T2-weighted MRI sequences with delayed contrast enhancement characterizes a myxoid stroma which is related to neurogenic tumors, myxofibrosarcomas, and myxoid liposarcomas 34 . Nevertheless, necrosis might appear in nearly all dedifferentiated, primary, or secondary malignancies 34 . Therefore, a well-differentiated liposarcomas might show a typical picture with larger size (>10 cm), septa and nodular enhancement, and can be easily distinguished from an ACC. However, low-differentiated sarcomas with necrosis and calcification might much harder be delimited from a malignant adrenal lesion 34 . In our series, imaging was not significant in sarcomas, metastases, and ACCs probably due to low differentiation of the tumors showing an unclear image of inhomogeneous mass with calcifications and necrosis. In the three cases of pheochromocytomas, the smaller size of the lesions might have played a role. If a tumor does not clearly arise from a solid organ, it can be defined as a primary retroperitoneal lesion. Those lesions are suspicious for sarcoma and patients should be referred to a qualified center for percutaneous biopsy, which represents the gold standard for preoperative diagnosis in retroperitoneal sarcomas34,37. In contrary, a preoperative biopsy is not recommended in case of an ACC because of the risk of tumor spilling unless there is evidence of metastatic disease requiring a histopathologic proof 39 . Therefore, physicians must rely on preoperative clinical presentation and radiological imaging 38 . If the tumor is located inside the adrenal gland, the differentiation between sarcoma and adrenal tumors seems easy. However, in case of unclear relation, which is more likely in large tumors, the discrimination between ACC, metastases, a rare primary sarcoma of the adrenal gland, or a retroperitoneal sarcoma remains difficult. Forty-seven percent of our patients were preoperatively misdiagnosed. The preoperative assessment did not differ between those, whose diagnosis was confirmed or revised (Table 2). Furthermore, patients with sarcoma or an ACC received nearly the same preoperative work-up and unfortunately, not all patients with sarcoma underwent preoperative biopsy mainly due to the suspicion of an ACC (Table 3). Ten to fifteen percent of ACCs present as incidentalomas. These asymptomatic adrenal tumors are more frequently detected with the increasing use of imaging, while the prevalence of ACC among adrenal incidentalomas varies between 1–11%44‒47. ACCs have a variable tumor size, and a small tumor size does not exclude malignancy 46 . We found that especially smaller tumors were significantly more often misdiagnosed. In addition, there was no significant difference in the diameter of ACCs and sarcomas (Figure 1; Table 3). Gender, tumor side, and tumor infiltration of other organs were not significantly associated with the fact, that a preoperative clinical diagnosis was either confirmed or revised (Table 2, 3).
Furthermore, the histological diagnosis of ACC remains challenging in the absence of metastatic disease. Nine histological characteristics are usually assessed as part of the Weiss score32,48,49. The Weiss system comprises nine issues of invasive behavior, cell architecture and mitotic characteristics, but is limited in borderline tumors, pediatric patients and oncocytic and myxoid variants 6 . Sarcomatoid ACCs present histopathological features of a conventional ACC as well as sarcomatoid lesions of a sarcoma, which are not accounted in the Weiss system and may lead to confusion when dealing with these malignancies 6 . Especially in case of strong dedifferentiation a retroperitoneal tumor impedes the diagnosis of adrenocortical origin 28 . There might be a mix of sarcomatoid and more mature parts requiring numerous immunostainings and molecular tests for exact differentiation3,7,50,51. In our study we resected 23 tumors, which we clinically suspected to be ACCs, but only 11 of them were histologically classified as ACCs. ACC is a rare and aggressive malignancy with an incidence of 0.5–2 per million per year and its overall prognosis is already very poor 52 . However, myxoid or sarcomatoid histological features or an increased Ki-67 index are associated with a worse outcome 53 . Especially in tumors with a high proliferation rate, a rapid initiation of adequate therapy is essential.
The suggested treatment of stage I-III ACC is complete surgical resection. Negative resection margins and an intact tumor capsule are important prognostic factors 54 . Smaller tumors (<6 cm or stage I-II) without any evidence of local invasion can be resected by a minimally invasive approach with given expertise39,55‒57. The recommended surgical strategy for large ACC implies an open procedure with complete tumor excision with en-bloc excision of perirenal fat and if necessary infiltrated adjacent organs to achieve negative resection margins39,50,54. Ipsilateral nephrectomy by routine did not deliver a significant benefit 57 . Twenty-five percent of patients present with tumor thrombus in the adrenal or renal vein or vena cava requiring thrombectomy or vascular replacement55‒58. The role of a consequent lymphadenectomy remains unclear but should be performed in stage III tumors59‒64. Dependent on the Ki-67 index, adjuvant treatment with mitotane is indicated. Stage IV ACCs necessitate chemotherapy. Even after full surgical resection, the risk for recurrence is up to 85% and adjuvant mitotane is recommended in case of high risk of recurrence63‒65.
Wide excision with negative margins also is the only curative modality for retroperitoneal sarcomas. It is defined as a retroperitoneal quasi-compartmental resection with en-bloc visceral resections of adjacent organs and tissues 66 . Even if the role of preoperative radiotherapy and chemotherapy is not completely established and still investigated, it should be considered when radical resection is technically uncertain or in case of higher grading. Therefore, the standard preoperative approach requires preoperative histopathological proof, different to the guidelines for ACC 66 . Delayed diagnosis or misdiagnosis of patients with retroperitoneal sarcoma may also result in increased tumor size or metastases which is associated with a worse prognosis. As the treatment of retroperitoneal sarcomas is complex, outcome improves when patients are treated in high-volume sarcoma centers66,67. Despite complete initial resection, more than 50% of patients with a retroperitoneal sarcoma will relapse and adjuvant chemotherapy might be necessary67,68. In our cohort, there was no difference regarding the surgical procedure despite misdiagnosing or depending on the entity (Table 2, 3). Preoperative misdiagnosing did not influence the overall survival (Figure 2).
We found one case of sarcomatoid adrenocortical cancer. According to WHO classification 2004, oncocytic ACCs, myxoid ACCs, and ACCs with sarcomatous areas are regarded as rare variants of ACC and only few cases of this unusual sarcomatous variant have been previously reported3,6,7. They should be designated carcinosarcoma or sarcomatoid carcinoma as they show forms of mesenchymal malignancy6,51. The limited number of reported cases does not allow any precise definition, but adrenocortical carcinosarcomas and sarcomatoid carcinomas have similar clinical outcome and a worse prognosis compared to ACC 6 . This aggressive behavior may be influenced by progressive biological dedifferentiation from a pre-existing better differentiated ACC, in analogy to the anaplastic carcinoma of the thyroid gland6,69,70. The one patient detected in our cohort showed poor survival and died within 14 months of follow up.
Conclusion
The limitation of this study is the small number of cases, the retrospective character, and the heterogeneity of presented patients. However, we wanted to outline that even in a high-volume center, the prevalence of retroperitoneal sarcomas and ACCs is low and the differentiation between both entities challenging by clinical and radiological methods as well as by histopathology. Biopsy of an ACC is not recommended but could be helpful in case of a sarcoma. Nevertheless, disease-free survival in both entities depends on radical resection and surgery should be performed in questionable cases. In case of an equivocal pathological report, a reference pathologist should be consulted as soon as possible to allow early initiation of adjuvant treatment.
Footnotes
Acknowledgments
Parts of the presented work have been presented at the national conference “Symposium der Deutschen Gesellschaft für Endokrinologie” 2016 and at the "Joint Meeting of the German Association of Endocrine Surgeons (CAEK) and the British Association of Endocrine and Thyroid Surgeons (BAETS)" in Berlin in 2016 in co-operation with the center of the university of Wuerzburg and the support and of Prof. Dr. Quinkler, Prof. Dr. Fassnacht, Dr. Korinna Jöhrens, and Ms. Wiebke Herrman. The current data were analyzed and forwarded without the support of the second center and the co-authors of 2016 whereas it only includes oupatients treated at our clinic.
Conflicting interests
The authors declare that there is no conflict of interest.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Ethical approval
The Ethics committee of Charité does not require approval for reporting case series. This study was completed in accordance with the Helsinki Declaration as revised 2013.
Informed consent
Verbal informed consent was generally obtained from all subjects for research projects. Written informed consent was partly given for research reasons and research register. Full written consent especially for this project could not be obtained as patients have already died or been lost for follow up as data comprises the period from 2009 till 2015.
Contributorship
E.M.D. and N.R. analyzed and interpreted the patient data. H.B. and W.S. performed the histological interpretation and evaluation; examination of the kidney, and was a major contributor in writing the manuscript. M.M. and M.B. participated in drafting the manuscript. JP had the supervision of the presented work. All authors read and approved the final manuscript.
