Concurrent Angiomyolipoma and Renal Cell Neoplasia: A Study of 36 Cases Rafael E. Jimenez, M.D., John N. Eble, M.D., Victor E. Reuter, M.D., Jonathan I. Epstein, M.D., Andrew L. Folpe, M.D., Mariza de Peralta-Venturina, M.D., Pheroze Tamboli, M.D., I. David Ansell, M.D., David J. Grignon, M.D., Robert H. Young, M.D., Mahul B. Amin, M.D. Departments of Pathology of Emory University Hospital, Atlanta, Georgia (REJ, ALF, MBA); Indiana University, Indianapolis, Indiana (JNE); Memorial Sloan-Kettering Cancer Center, New York, New York (VER); John Hopkins Hospital, Baltimore, Maryland (JIE); Henry Ford Hospital, Detroit, Michigan (MPV); M.D. Anderson Cancer Center, Houston, Texas (PT); Nottingham City Hospital, Nottingham, United Kingdom (IDA); Harper Hospital and the Karmanos Cancer Institute, Detroit, Michigan (DJG); and Massachusetts General Hospital, Boston, Massachusetts (RHY) neoplasms studied were negative for HMB-45, Little is known about the association of angiomyo- whereas all 25 angiomyolipomas studied were lipoma and adult renal-cell neoplasia. We studied positive. the clinicopathologic features of 36 patients with concurrent angiomyolipoma and renal-cell neopla- KEY WORDS: Angiomyolipoma, Collision neo- sia from the consultation and surgical pathology plasms, Kidney, Oncocytoma, Renal cell carcinoma, files of nine institutions. HMB-45 immunoreactivity Renal cell neoplasia, Clear cell carcinoma. was analyzed in both neoplasms. Twenty-five spo- Mod Pathol 2001;14(3):157–163 radic cases of patients with angiomyolipoma and renal-cell neoplasia and 11 cases of patients with The occurrence of angiomyolipoma in association tuberous sclerosis, as defined by Gomez’ criteria, with a renal-cell neoplasm is uncommon, with only had mean ages of 59 and 53 years, respectively, and approximately 50 cases reported (1–37). Although female–male ratios of 2:1 and 5:1, respectively. The angiomyolipomas are extremely common in pa- mean size of the angiomyolipomas was 1 cm in the tients with the tuberous sclerosis complex (38), it is sporadic cases and 3 cm in those patients with tu- unclear whether patients with tuberous sclerosis berous sclerosis (medians: 0.5 and 3 cm, respec- have more renal-cell neoplasms than the general ؍ tively, P .002). The mean sizes of the renal-cell population. neoplasms were 5 cm in sporadic cases and 6 cm in Our understanding of the biology and clinical patients with tuberous sclerosis (medians: 4 and 5 behavior of angiomyolipoma and renal-cell neopla- ؍ cm, respectively; P .88). In both clinical settings, sia has recently been augmented by several impor- angiomyolipoma was more commonly the inciden- tant developments. There is strong evidence that tal tumor. Clear-cell (conventional) renal-cell carci- angiomyolipomas, for a long time considered to be noma was the most common renal-cell neoplasm in hamartomatous, are actually neoplasms (39, 40). both groups of patients, accounting for approxi- The morphologic spectrum of angiomyolipoma has mately two thirds of the tumors. In patients with been recently expanded to include an epithelioid tuberous sclerosis, 27% of renal-cell neoplasms variant, which is capable of a more aggressive be- were oncocytomas, compared with 8% in sporadic havior than the classic form (41–43). Further, the ؍ cases (P .15). Papillary neoplasia, chromophobe, recent description of the perivascular epithelioid and collecting-duct renal-cell carcinoma were cell family of tumors (44), to which angiomyoli- found only in sporadic cases. All of the 22 renal-cell poma belongs, has expanded our knowledge of en- tities that would fall into the differential diagnosis of angiomyolipoma and renal-cell neoplasms. Sim- Copyright © 2001 by The United States and Canadian Academy of ilarly, our knowledge of tuberous sclerosis has been Pathology, Inc. VOL. 14, NO. 3, P. 157, 2001 Printed in the U.S.A. strengthened by the definition of a constellation of Date of acceptance: December 5, 2000. Presented in part at the 89th United States and Canadian Academy of phenotypic diagnostic criteria and the discovery of Pathology Annual Meeting, New Orleans, LA, March 21–31, 2000. the genes involved in its pathogenesis (45–47). Last, Address reprint requests to: Mahul B. Amin, M.D., Emory University Hospital, 1364 Clifton Road, N.E., G-167, Atlanta, GA 30322; e-mail: our understanding of the biologic potential of [email protected]; fax: 404-712-0148. renal-cell neoplasms has been recently expanded 157 with the adoption of a widely accepted classifica- RESULTS tion (48) that is based on distinct histologic features that correlate closely with consistent genetic A total of 36 cases of coexistent renal-cell neopla- abnormalities. sia and angiomyolipoma were identified from nine We undertook a detailed clinicopathologic study institutions and/or consultation files of the authors. to ascertain the frequency of the concurrence of Eighteen cases from five institutions were identified angiomyolipomas and renal-cell neoplasms in the from a review of approximately 2160 nephrecto- same patient, whether any particular histologic mies performed for a renal mass, suggesting an type of renal-cell neoplasm is more frequently as- incidence of slightly less than 1% (data were un- sociated with angiomyolipomas, and whether the available for the remaining four institutions). Clin- clinical presentation or behavior of either neo- ical features are summarized in Table 1. In one plasm is different compared with when they oc- case, the renal-cell neoplasm preceded a contralat- cur in isolation. Furthermore, we studied whether eral angiomyolipoma by 11 months, whereas in 35 there was any difference in the presentation or cases, both neoplasms were diagnosed simulta- behavior between patients with and without tu- neously. Additionally, one of these patients had a berous sclerosis. partial nephrectomy 6 years later for a contralateral angiomyolipoma. In 19 cases, both neoplasms oc- curred in the left kidney, in 10 they occurred in the right kidney, and in 4, the side was not specified. MATERIALS AND METHODS One patient had a renal-cell neoplasm in the left Surgical pathology records of nine institutions side and bilateral angiomyolipomas, and two pa- and the consultation files of the authors were tients had bilateral renal-cell neoplasms and an searched for the diagnosis of renal angiomyoli- angiomyolipoma in the left kidney. One of the pa- poma. Patients that had coexistent, precedent, or tients with bilateral renal-cell neoplasms had subsequent renal-cell neoplasia were included in autosomal-dominant polycystic kidney disease. the study. Eleven patients (31%) had tuberous sclerosis. Of Clinical information was obtained from the pa- these, only six had a clinical diagnosis of tuberous tients’ medical records and treating physicians and sclerosis; in the remaining five, the diagnosis was included age at presentation, gender, any personal made based on the presence of multiple angiomyo- or family history of tuberous sclerosis, and lipomas (47). The mean ages for the sporadic and tuberous sclerosis-related cases, respectively, were follow-up information. A patient was considered to ϭ have tuberous sclerosis if there was a clinical diag- 59 and 53 years (P .365), and the female-to-male nosis of the disease or, according to Gomez’ criteria ratios were 2:1 and 5:1. Twenty-eight cases were (47), if there were multiple angiomyolipomas. He- matoxylin and eosin–stained sections of each neo- TABLE 1. Clinical Features of 34 Cases of Coexistent plasm were reviewed. Renal epithelial neoplasms Renal Cell Neoplasia and Angiomyolipoma Sporadic Setting Tuberous Sclerosis were classified according to the Rochester classifi- Variable cation (48). Other pathologic features noted in- (n ϭ 25) (n ϭ 11) cluded size of the tumor and multifocality. Fuhr- Mean age (y) 59 53 F:M 2:1 5:1 man nuclear grade and 1997 UICC/AJCC TNM Side of angiomyolipoma (n) pathologic stage were determined for malignant Right 9 1 renal-cell neoplasms. Size, histologic type, and Left 13 8 Bilateral 0 1 multifocality of angiomyolipoma were recorded. Not specified 3 1 Because many of the cases were seen in consulta- Side of renal cell neoplasm (n) tion and all of the cases were retrieved retrospec- Right 8 2 Left 12 8 tively from nine institutions, gross sampling of tu- Bilateral 2 0 mors was not performed in any predictable, Not specified 3 1 standardized manner. Multifocality (n) Angiomyolipoma 0 11 At least one block from each neoplasm (available Renal cell neoplasm 7 0 for 22 renal-cell neoplasms and 25 angiomyolipo- Surgery type (n) mas) was retrieved, and sections were submitted for Total nephrectomy 21 7 Partial nephrectomy 3 4 immunohistochemical staining for monoclonal Partial and radical 10 HMB-45 (DAKO, Carpenteria, CA; dilution, 1:40). nephrectomy Statistical analysis was done with SPSS software, Size angiomyolipoma (cm) Mean 1.0 3.3 Version 7.0, for Windows 98. Fisher’s Exact test was Median 0.5 3.0 used to analyze frequency data, whereas the Stu- Size renal cell neoplasm (cm) dent’s t test was used for numerical data. P values of Mean 5.2 5.6 less than .05 were considered significant. Median 3.8 5.0 158 Modern Pathology managed with total nephrectomy (Fig. 1), seven TABLE 2. Distribution of Cases by Histologic Subtype and Clinical Setting with partial nephrectomy, and one with a radical and a contralateral partial nephrectomy. In both Histologic Subtype Sporadic Setting Tuberous Sclerosis sporadic and tuberous sclerosis–associated pa- Benign Papillary adenoma 2* — tients, angiomyolipoma was more commonly the Metanephric adenoma — 1 incidentally found tumor. This was also true within Oncocytoma 2† 3 the category of tuberous sclerosis–associated pa- Malignant Clear-cell RCC 15* 6 tients (those with clinical history of tuberous scle- Papillary RCC 2 — rosis versus those assigned as tuberous sclerosis Chromophobe RCC 4 — based on Gomez’ criteria of multifocality). Collecting-duct RCC 1 — For angiomyolipomas occurring sporadically, the Unclassified RCC — 1 mean and median sizes were 1.0 and 0.5 cm, re- RCC, renal cell carcinoma.