Comparative Genomic Hybridization Analysis of Adrenocortical Tumors
STAN SIDHU, DEBORAH J. MARSH, GEORGE THEODOSOPOULOS, JEANETTE PHILIPS, CHRISTOPHER P. BAMBACH, PETER CAMPBELL, CHRISTOPHER J. MAGAREY, COLIN F. J. RUSSELL, KLAUS-MARTIN SCHULTE, HANS-DIETRICH RÖHER, LEIGH DELBRIDGE, AND BRUCE G. ROBINSON
Cancer Genetics, Kolling Institute of Medical Research (S.S., D.J.M., G.T., B.G.R.), and Departments of Surgery (S.S., C.P.B., L.D.) and Pathology (J.P.), Royal North Shore Hospital, St. Leonards, New South Wales 2065, Australia; Departments of Molecular Medicine (S.S., D.J.M., G.T., B.G.R.), Pathology (J.P.), Medicine (B.G.R.), and Surgery (L.D.), University of Sydney, New South Wales 2006, Australia; Department of Surgery, Liverpool Hospital (P.C.), Liverpool, New South Wales 2170, Australia; Department of Surgery, St. George Hospital (C.J.M.), Kogarah, New South Wales 2217, Australia; Department of Surgery, Ward 37, Royal Victoria Hospital (C.F.J.R.), Belfast, United Kingdom BT12 6BA; and Department of General and Trauma Surgery, Heinrich Heine University (K .- M.S., H .- D.R.), Düsseldorf, Germany
Comparative genomic hybridization (CGH) is a molecular cy- togenetic technique that allows the entire genome of a tumor to be surveyed for gains and losses of DNA copy sequences. A limited number of studies reporting the use of this technique in adult adrenocortical tumors have yielded conflicting results.
In this study we performed CGH analysis on 13 malignant, 18 benign, and 1 tumor of indeterminate malignant potential with the aim of identifying genetic loci consistently impli- cated in the development and progression of adrenocortical tumors. Tissue samples from 32 patients with histologically proven adrenocortical tumors were available for CGH anal- ysis. CGH changes were seen in all cancers, 11 of 18 (61%) adenomas, and the 1 tumor of indeterminate malignant po- tential. Of the adrenal cancers, the most common gains were seen on chromosomes 5 (46%), 12 (38%), 19 (31%), and 4 (31%).
Losses were most frequently seen at 1p (62%), 17p (54%), 22 (38%), 2q (31%), and 11q (31%). Of the benign adenomas, the most common change was gain of 4q (22%).
Mann-Whitney analysis showed a highly significant differ- ence between the cancer group (mean changes, 7.6) and the adenoma group (mean changes, 1.1) for the number of ob- served CGH changes (P < 0.01). Logistic regression analysis showed that the number of CGH changes was highly predic- tive of tumor type (P < 0.01).
This study has identified several chromosomal loci impli- cated in adrenocortical tumorigenesis. Activation of a pro- tooncogene(s) on chromosome 4 may be an early event, with progression from adenoma to carcinoma involving activation of oncogenes on chromosomes 5 and 12 and inactivation of tumor suppressor genes on chromosome arms 1p and 17p. (J Clin Endocrinol Metab 87: 3467-3474, 2002)
A DRENAL MASSES ARE common and reported in 3-7% of autopsy and radiological series (1). The majority are benign adrenocortical adenomas, which may be functioning or nonfunctioning. Patients with functioning tumors are of- fered surgery routinely; however, patients with nonfunc- tioning tumors are offered surgery on the basis of symptoms and, if asymptomatic, on the basis of changes in tumor size, as this is currently the single best predictor for malignancy. The minimal size recommendation for surgery for inciden- tally discovered adrenal masses varies between institutions and ranges from 3-5 cm. Up to 10% of resected adrenocortical carcinomas lie in this size range (2). In contrast, adrenocor- tical malignancy is rare, with a reported incidence of 2 cases / million head of population (3). Presentation with pain and pressure symptoms secondary to a mass effect are common, with 30-50% of carcinomas secreting cortisol and androgens and less commonly estrogens and aldosterone. Once metas- tasis or local invasion has occurred, the patient’s prognosis is poor, and therefore the mainstay of successful treatment is surgical removal of the adrenal carcinoma while it is still confined to the adrenal gland (3), hence the seemingly ag-
gressive protocols regarding incidental adrenal tumors 3 cm or larger. A predictive genetic test discriminating malignant from benign disease for incidentally detected adrenal masses would represent a significant advance in the clinical man- agement of such patients.
Despite extensive work over the last decade, the genetic changes underlying the development and progression of sporadic adrenocortical tumors remains poorly understood (4). This is in contrast to the genetic abnormalities underlying several familial cancer syndromes that feature adrenocortical tumors as a manifestation. Briefly, the Beckwith-Wiedemann syndrome (OMIM 130650), characterized by the increased propensity to develop childhood tumors such as Wilms’ tumor, neuroblastoma, hepatoblastoma, and adrenocortical carcinoma, has been assigned to the chromosomal region 11p15.5. Several genes at this locus are implicated in Beck- with-Wiedemann syndrome, including CDKN1C, H19, and IGF-II (5). Adrenocortical carcinoma also features in the Li- Fraumeni syndrome (OMIM 151623) along with soft tissue sarcomas, breast cancer, leukemia, and brain tumors. The molecular basis for this condition is a germline mutation of the p53 tumor suppressor gene that maps to the short arm of chromosome 17 (6).
Abbreviations: CGH, Comparative genomic hybridization; CNC, Carney complex; MEN1, multiple endocrine neoplasia type 1.
Two familial syndromes featuring benign adrenal tumors are multiple endocrine neoplasia type 1 (MEN1; OMIM 13110) and the Carney complex (CNC; OMIM 160980). MEN 1 is associated with adrenal adenomas in 30% of cases and, more rarely, with adrenocortical carcinoma in addition to the more prevalent tumors involving the parathyroid, pituitary, or pancreatic glands. The underlying abnormality is a mu- tation in the MEN1 gene, which maps to 11q13 (7,8). In CNC, patients develop spotty cutaneous pigmentation, atrial myx- omas, and nodular adrenocortical dysplasia and have an abnormal genetic loci map to 2p16 (9) and 17q22-24 (10, 11). Germline mutation in the gene encoding for the protein kinase A type Ia regulatory subunit mapping to 17q22-24 has been shown to be responsible for the CNC phenotype in up to half (40-45%) of these patients.
The characterization of mutations in genes responsible for these familial cancer syndromes has prompted several groups to look for mutations in key candidate genes in spo- radic adrenocortical tumors. Most studies have shown a low level of mutation for the genes tested (4). To date, the most common change seen in the majority of adrenocortical car- cinomas is insulin-like growth factor II overexpression due to uniparental isodisomy at the 11p15.5 locus (12). Further- more, mutations in the p53 tumor suppressor gene are seen in approximately one quarter of adrenocortical cancers (13, 14). These changes are seen infrequently in benign adreno- cortical lesions (13, 14). As a consequence of the poor yield from the candidate gene approach, other investigators have applied the technique of comparative genomic hybridization (CGH) to screen for genetic aberrations in adrenocortical tumors (15-17). CGH is a molecular cytogenetic technique that allows the entire genome of a tumor to be surveyed for gains and losses of DNA copy sequences. Regions of DNA copy gain may harbor putative tumor oncogenes, whereas regions of DNA copy loss may contain tumor suppressor genes (18). This then allows further fine mapping of regions considered important in the development of specific tumor types.
Previous studies have demonstrated an increased fre- quency of DNA copy number changes in large malignant adrenocortical tumors compared with small benign tumors; however, there is little consensus regarding regions of con- sistent chromosomal aberration. This is in contrast to pedi- atric adrenal CGH studies, which have shown a high level of concordance for chromosomal regions implicated in tumor- igenesis (19, 20). In this study we performed CGH analysis on 13 malignant, 18 benign, and 1 adrenal tumor of inde- terminate malignant potential with the aim of identifying loci consistently implicated in the tumorigenesis of adult adre- nocortical tumors.
Subjects and Methods
Patients and tumors
Ethics approval for the study was obtained from the Northern Sydney Area Health Service ethics committee and ethics committees of partic- ipating institutions. The study was performed in accordance with the ethical standards of the Helsinki Declaration of 1975. Tissue samples from 32 patients with sporadic, histologically proven adrenocortical tumors were available for analysis. Patients with adrenal tumors as part of a known familial syndrome were excluded on clinical grounds. There
were 12 males and 20 females, with a mean age of 51 yr. Patient data are summarized in Table 1. Tumor tissue was obtained at surgery from the central part of the lesion and snap-frozen at -80 C in liquid nitrogen. DNA extraction was performed from fresh-frozen tissue using standard proteinase K and phenol-chloroform protocols (21). At the time of DNA extraction, a representative sample of tumor tissue was excised and sent for histological examination. Only tumors in which at least 80% of tumor cells were obtained from the site of DNA extraction were used for CGH analysis. Five of 37 (14%) tumors were excluded on this basis for CGH analysis. The presence of more than 2 of the following histological criteria was used to diagnose malignancy in the absence of metastasis or local tumor invasion: necrosis, vascular and capsular invasion, diffuse growth pattern, more than 5 mitoses/50 high power field, atypical mitoses, the presence of more than 75% eosinophilic tumor cells, and nuclear and cellular atypia. Benign tumors were classified on the basis that they did not show any of these suspicious histological features, whereas tumors with 1-2 of these features were classified as being of indeterminate malignant potential (22, 23). There were 13 adrenocortical cancers, 18 benign adenomas, and 1 tumor of indeterminate malignancy available for analysis.
CGH
CGH was performed in accordance with previously described pro- tocols (18, 24). In brief, tumor DNA (probe) was labeled with the green fluorochrome, fluorescein 12-deoxy-UTP (NEN Life Science Products, Boston, MA), by nick translation (18). The probe was then cohybridized with normal DNA [labeled with spectrum red-deoxy-UTP (Vysis, Inc., Downers Grove, IL)] and unlabeled Cot-1 blocking DNA (Life Tech- nologies, Inc., Gaithersburg, MD) to a denatured metaphase spread of chromosomes from a karyotypically normal healthy male and incubated at 37 C in a moist chamber for 3 d. After hybridization, slides were washed, stained with 4,6-diamidino-2-phenylindole (Vysis, Inc.), and stored in a light-impermeable box. For each hybridization, a control experiment was performed using normal male DNA (Promega Corp., Madison, WI) labeled with fluorescein 12-deoxy-UTP and cohybridized with the standard reference DNA labeled with spectrum red-deoxy-UTP (Vysis, Inc.). In addition, separate reverse labeling experiments were performed to confirm chromosomal gains and losses when results were equivocal.
Digital image analysis and interpretation
Digital images were captured using an Olympus Corp. BX50 epiflu- orescence microscope (Olympus Corp., Tokyo, Japan) fitted with dif- ferent single bandpass filter sets for 4,6-diamidino-2-phenylindole (blue), green, and red fluorescence. The microscope was equipped with a monochrome charge-coupled device camera (Cohu, Inc., San Diego, CA) interfaced to a quantitative image-processing system (QUIPS, Vysis, Inc.). For each tumor sample 15-20 metaphase spreads were captured, and image analysis and interpretation were performed using QUIPS software.
The average green/red profile along each chromosome was deter- mined, and only metaphases containing high quality green and red fluorescence intensities along each chromosome arm were included in the final analysis. After exclusions, the observations from 9-10 meta- phase spreads for each tumor sample were pooled to obtain the mean green/red ratio. Green/red ratios of 1.20 or more were considered gains of genetic material, and ratios of 0.80 or less were considered losses of genetic material. Heterochromatic regions of the chromosome including the telomeres, centromeres, and paracentromeric regions and the Y chromosome were not included in the analysis (24).
Statistical analysis
Mann-Whitney analysis, logistic regression analysis, and x2 analysis were used to compare the number of aberrations between tumor types and examine discriminating factors between the tumor groups. Pearson correlation analysis was used to examine the relationship between tu- mor size and the number of genetic changes.
| Patient | Age | Sex | Tumor type | Size (cm) | Hormonal profile | Follow-up (months)/ findings | No. of CGH changes | Gains | Losses |
|---|---|---|---|---|---|---|---|---|---|
| 1 | 74 | F | Cancer | 9 | Nil | 24/NED | 15 | 3q22-qter, 4, 5, 6q, 7, 9p, 12p, 12q14- 21, Xq | 1p31-pter, 11q, 13, 15, 17, 22 |
| 2 | 53 | M | Cancer | 5.5 | Nil | 6/DOD | 11 | 1p13-31, 1q, 3q, 8q | 3p, 8p, 13, 16p, 17p, 18, 19р |
| 3 | 72 | F | Cancer | 9 | DOC | 12/NED | 4 | X | 15, 17, 22 |
| 4 | 73 | M | Cancer | 10 | ADH | 97/NED | 10 | 6p, 12p, 14, 19q | 2q31-qter, 3p, 4q, 5q, 11, 17p |
| 5 | 28 | F | Cancer | 8.5 | Cortisol, DHEAS | 100/NED | 6 | 4, 5, 12, 19, 20 | 17p13-pter |
| 6 | 66 | F | Cancer | 6 | Nil | 66/NED | 6 | 1q, 5p, 3qcent-3q25, 9p | 1p34-pter, 2q34-qter |
| 7 | 37 | F | Cancer | 9 | Cortisol, DHEAS | 9/MET | 7 | 2, 5, 12, 18q | 1p32-pter, 11q, 16p |
| 8 | 47 | F | Cancer | 14 | Cortisol, DHEAS | 32/MET | 13 | 2q, 4, 5, 6q, 7, 9p, 13 | 1p32-pter, 16, 17, 19, 20q, 22 |
| 9 | 72 | F | Cancer | 12 | Nil | 18/NED | 1 | 19 | |
| 10 | 60 | M | Cancer | 8 | Estradiol | 100/MET | 6 | 9q, 12q, 17q, 19 | 1pcent-1p31, 1qcen-1q31 |
| 11 | 42 | M | Cancer | 11 | Aldosterone | 33/REC | 6 | 9q31-qter, 14, 19 | 1pcent-1p31, 2q, 4 |
| 12 | 58 | F | Cancer | 9 | Nil | 10/NED | 3 | 1pcent-1p31, 2q, 13q | |
| 13 | 62 | M | Cancer | 14 | Nil | 12/NED | 10 | 4, 5, 12, 7p, 20p | 1p31-pter, 11q24-qter, 9q, 17p, 22 |
| 14 | 62 | M | Indeterminant malignant potential | 11 | Aldosterone | 18/NED | 1 | 22 | |
| 15 | 51 | M | Adenoma | 5 | Nil | 24/NED | 1 | X | |
| 16 | 66 | M | Adenoma | 4 | Cortisol | 20/NED | 1 | 3q | |
| 17 | 47 | M | Adenoma | 5 | Nil | 20/NED | 0 | ||
| 18 | 54 | M | Adenoma | 4 | Cortisol | 39/NED | 0 | ||
| 19 | 55 | F | Adenoma | 4 | Nil | 22/NED | 1 | 9q | |
| 20 | 53 | F | Adenoma | 3 | Cortisol | 36/NED | 0 | ||
| 21 | 49 | F | Adenoma | 3.7 | Cortisol | 18/NED | 2 | 3q26-qter, 17p12-pter | |
| 22 | 36 | F | Adenoma | 3 | Aldosterone | 12/NED | 0 | ||
| 23 | 46 | F | Adenoma | 2 | Aldosterone | 24/NED | 3 | 18p | 17q, 19q |
| 24 | 23 | F | Adenoma | 2.5 | Aldosterone | 20/NED | 1 | 4q | |
| 25 | 52 | M | Adenoma | 1.5 | Aldosterone | 41/NED | 0 | ||
| 26 | 54 | M | Adenoma | 1.3 | Aldosterone | 27/NED | 4 | 17 | 1pcent-1p31, 2q22-32, 5q |
| 27 | 52 | F | Adenoma | 2 | Aldosterone | 15/NED | 1 | 4q | |
| 28 | 38 | F | Adenoma | 4 | Cortisol | 22/NED | 2 | 4 | 14 |
| 29 | 71 | F | Adenoma | 3 | Nil | 18/NED | 1 | 5 | |
| 30 | 47 | F | Adenoma | 2 | Aldosterone | 12/NED | 0 | ||
| 31 | 50 | F | Adenoma | 3.5 | Nil | 17/NED | 2 | 4.5 | |
| 32 | 70 | F | Adenoma | 3 | Aldosterone | 19/NED | 0 |
DOC, Deoxycorticosterone; ADH, antidiuretic hormone; DHEAS, dehydroepiandosterone sulphate; NED, no evidence of disease; DOD, dead of disease; REC, local recurrence; MET, metastases.
Results
CGH analysis
CGH analysis of our tumor cohort revealed DNA copy number changes in 13 of 13 (100%) adrenocortical cancers, 11 of 18 (61%) adenomas, and the 1 tumor of indeterminate malignant potential. Of the malignant tumors the mean num- ber of CGH changes was 7.6 (range, 1-15), with an equal distribution of gains and losses. The benign adenomas showed a mean of 1.1 changes (range, 0-4), and these, too, showed an almost equal distribution of gains (53%) and losses (47%).
The results of DNA copy number changes for adrenal cancers and adrenal adenomas are shown in Figs. 1 and 2. Of the adrenal cancers, the most common gains were seen on chromosomes 5p (46%), 5q (38%), 12p (38%), 12q (38%), 19 (31%), and 4 (31%). The minimal region of gain on chromo- some 12 was 12q14-21, whereas no smaller regions could be
delineated on chromosomes 4, 5, and 19. Losses were most frequently seen at 1p (62%), 17p (54%), 22 (38%), 2q (31%), and 11q (31%). The minimal regions of loss on these chro- mosomes were 1p34-pter, 17p13-pter, 2q34-qter, and 11q24- qter, respectively. No minimal region of loss was defined for chromosome 22, primarily because of its small size. Of the benign adenomas, the most common gain was 4q (22%), followed by chromosome 5 (11%), whereas 2 (11%) losses were seen on chromosome arm 3q. The most common losses of chromosome arms 1p and 17p in our malignant group were reflected in one adenoma each respectively in the be- nign cohort.
Statistical analysis
Mann-Whitney analysis showed a highly significant dif- ference between the malignant and benign tumors for the number of observed CGH changes (P < 0.01). Logistic re-
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
X
Y
gression analysis showed that the number of CGH changes was highly predictive of tumor type (P < 0.01). The cohort was examined by x- analysis for the discriminating variables CGH changes less than or equal to 3 and CGH changes greater than 3, and this was found to significantly discrim- inate between the malignant and the benign group (P < 0.01).
The mean size of the malignant tumors was 9.6 cm (range, 5.5-14 cm) compared with the benign tumor mean size of 3.1 cm (range, 1.3-5 cm). Examination of the number of CGH changes in relation to tumor size using Pearson correlation analysis (Fig. 3) revealed a strong relationship between in- creasing tumor size and the number of genetic aberrations (P < 0.01). This relationship was independent of benign or malignant tumor type.
Discussion
In our series there was a marked difference between the number of genetic events observed in adrenocortical carci- nomas vs. adenomas. Chromosomal gains and losses were equally distributed in both groups; however, carcinomas showed, on the average, 7 times the number of changes compared with adenomas. We detected CGH changes in 61% of adenomas, which is similar to the 71% and 75% rates of CGH changes reported by Zhao (16) and Dohna (17), re- spectively, but less than the 28% rate of change reported by Kjellman (15). The most common change seen in our ade- noma cohort was a gain in chromosome 4. This was also observed in approximately one third of the adrenocortical carcinomas, suggesting that activation of a protooncogene(s) on chromosome 4 is an early event in adrenocortical tumor- igenesis. In previous studies Kjellman (15) had demonstrated a gain of chromosome 4 in 4 of 8 carcinomas, but did not observe any such change in the adenomas studied. The pub- lished results of Zhao (16) and Dohna (17) showed no gains
of chromosome 4 in 23 adenomas (mean size, 4 cm) and 1 gain of chromosome 4 in 8 adenomas (mean size, 4.5 cm), respec- tively, as well as gains of chromosome 4 in 2 of 12 (17%) carcinomas and 4 of 12 (33%) carcinomas, respectively. The gain of chromosome 4, at least in the adrenocortical carci- nomas, seems to be a moderately consistent finding among the 3 series previously reported.
Of the carcinomas, the most significant copy number changes were losses of all or part of chromosome arms 1p and 17p in half the cohort as well as loss of all or part of chro- mosome arms 2q, 11q, and chromosome 22 in one third of the group. Gains were predominantly observed on chromo- somes 4, 5, 12, and 19, and these were seen in 31-46% of the cancers. The changes we report have been described previ- ously, albeit in different series (Table 2). Overall, our expe- rience is most consistent with the data from Kjellman (15), with an equal distribution of gains and losses in the carci- noma cohort. In that study losses were reported predomi- nantly on chromosome arms 17p (50%), 11q (50%), and chro- mosome 2 (50%), and gains were seen primarily on chromosomes 4 (50%) and 5 (50%). Loss of chromosome arm 1p was seen in 1 of 8 (13%) carcinomas compared with 8 of 13 (62%) carcinomas in our series. However, the series of Zhao (16) reported loss of 1p in 8 of 12 carcinomas (67%), with the common region of deletion being 1p21-1p31. We ob- served loss of this region in 3 of 13 (23%) carcinomas, with 5 of 13 (38%) carcinomas displaying distal 1p loss. We report similar results as Zhao with respect to losses of chromosomes 2 and 11q and gains on chromosome 12; however, we had markedly differing results with respect to chromosome 17. Gains of chromosome arms 17p and 17q were observed in one quarter to one third of the Zhao adenoma cohort and in 17-25% of the Zhao cancer cohort, in contrast to the 54% loss of 17p that we observed in our cancer group. The CGH
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
X
Y
results reported by Dohna (17) least resemble our experience. They showed a predominance of gains in their cancer group, being observed consistently on chromosomes 5, 7, 8, 9, 12, 14, 16, 17, 20, and 22. Of these, gains of chromosomes 5 and 12 were observed in one third of our series.
The variation in reported results may reflect the method- ological limitations of CGH and the variability in its use and interpretation. It is accepted that CGH is unable to detect chromosomal rearrangements such as reciprocal transloca- tions, inversions, and uniparental isodisomy (24). This was evident in our findings, in that loss of 11p was seen in one adrenal cancer in contrast to the well documented 11p15 allelic rearrangements seen in 85% of adrenocortical cancers (loss of the maternal allele and duplication of the paternal allele) (25). The sensitivity of the technique is hampered by contamination of tumor material with normal cells and is also highly dependent on the ploidy of the tumor cells as well as the size and level of copy number changes, specifically in the cases of amplification. Based on validation experiments, the crucial detection size is estimated to be in the range of 10-20 Mb (26, 27). In our series we attempted to minimize normal cell contamination by only including tumors in which at least 80% tumor cells were identified on a histological section taken from the site of DNA extraction. As a consequence, 5 of 37 (14%) tumors were excluded from undergoing CGH analysis. Only 1 of 3 previous series reported this procedure before CGH analysis, and, perhaps significantly, this was the series that had data most consistent with our own (15). The recommendations regarding the number of metaphases an- alyzed to give a final profile for each tumor vary between institutions. The rationale behind averaging the results from several metaphases is to limit random noise in the hybrid- ization that may give misleading results (24). Previous adult adrenal CGH studies have included the profiles from a min-
imum of 2 (16), 3 (15), and 7 (17) metaphases. To minimize background noise and increase our ability to recognize real copy number changes, we included the results from 9-10 metaphases in the final CGH analysis and interpretation for each patient. Furthermore, verification of changes when re- sults were equivocal were made by separate reverse labeling experiments (24, 28). In this regard we feel we have applied a very rigorous and robust protocol to the examination of this cohort of adrenocortical tumors for chromosomal imbalances.
The other possible explanation for the apparent discrep- ancies reported between adult adrenal CGH series are the geographical and ethnic differences between the study pop- ulations. Certainly our series contained patients from a va- riety of different racial and ethnic backgrounds, partly due to the collaborative nature of the project and partly due to the inherent multicultural nature of Australian society. More than 75% of our cohort, however, was of Caucasian descent, so in this respect we doubt that racial origin would make our study population significantly different from the patients studied by other groups who were reported from three Eu- ropean centers (15-17).
To date, few genes have been implicated in adrenocortical tumorigenesis. Gicquel et al. (25) have demonstrated unipa- rental isodisomy at the 11p15 region, with IGF-II gene over- expression in more than 90% of malignant adrenal tumors studied in their series. p53 mutation analysis in adrenocor- tical cancers has yielded a 25% mutation rate when exons 5-8 have been amplified and sequenced (13, 14). Our results suggest that important tumor suppressor genes implicated in adrenocortical tumorigenesis lie on chromosome arms 1p and 17p. The candidate tumor suppressor gene p73 lies in the region of distal 1p, which was deleted in 5 of 13 of our malignant tumors. The incidence of p73 mutations has not
Number of CGH changes vs Adrenal Tumor Size
16
14
12
10
CGH changes
8
6
4
2
0
+
0
2
4
6
8
10
12
14
16
-2
Adrenal Size(cm)
| Locus of genetic change | Current series (n = 13) | Kjellman (15) (n = 8) | Zhao (16) (n = 12) | Dohna (17) (n = 14) |
|---|---|---|---|---|
| 1p- | 8 (62%) | 1 (13%) | 8 (67%) | 2 (14%) |
| 2q- | 4 (31%) | 3 (38%) | 5 (42%) | 0 |
| 4p+ | 4 (31%) | 3 (38%) | 3 (25%) | 3 (21%) |
| 4q+ | 4 (31%) | 4 (50%) | 0 | 1 (7%) |
| 5p+ | 6 (46%) | 4 (50%) | 3 (25%) | 8 (57%) |
| 5q+ | 5 (38%) | 4 (50%) | 4 (33%) | 7 (50%) |
| 11q- | 4 (31%) | 4 (50%) | 5 (42%) | 0 |
| 12p+ | 5 (38%) | 1 (13%) | 2 (17%) | 6 (43%) |
| 12q+ | 5 (38%) | 1 (13%) | 6 (50%) | 12 (86%) |
| 17p- | 7 (54%) | 4 (50%) | 1 (8%) | 0 |
| 19+ | 4 (31%) | 3 (38%) | 0 | 6 (43%) |
| 19- | 3 (23%) | 0 | 0 | 0 |
| 22- | 5 (38%) | 3 (38%) | 0 | 1 (7%) |
been evaluated in adrenocortical tumors; however, its likely role in the development of other tumors, including neuro- blastomas, breast, colorectal, gastric, and lung cancers, has not been as promising as when it was first reported (29). Our results, considered in conjunction with the 25% rate of p53 mutations reported in adrenocortical cancers, also suggest a role for other potential candidate tumor suppressor genes that map to chromosome 17p, such as HIC-1, CRK, and ABR (30,31). A recent report demonstrating loss of heterozygosity at 17p13 loci in 83% of malignant adrenal tumors lends fur- ther support to our results (32). It was also shown in this study that 17p loss of heterozygosity was a strong predictor of relapse in localized adrenal tumors with Weiss scores less than 4 on prolonged follow-up. The recently reported tumor suppressor gene PRKAR1A, which maps to 17q22-24, is mu- tated in at least 50% of CNC patients, but has not been studied in patients with sporadic adrenocortical tumors (11). In our series, 17q was lost in 3 of 13 (23%) cancers, suggesting a possible role for this gene in a subset of sporadic adreno- cortical tumors.
The MEN1 tumor suppressor gene maps to 11q13, and 31%
of our malignant tumors had lost all or part of chromosome 11q. However, point mutations in the MEN1 gene have rarely been demonstrated in adrenocortical tumors, suggesting that other potential tumor suppressor genes on 11q may play a role in adrenal tumorigenesis (33, 34). Alternatively, epige- netic inactivation, e.g. methylation, possibly coupled with loss of the remaining allele may lead to functional loss of MEN1, even in the absence of mutations. There are numerous genes within the gained regions of chromosomes 4, 5, and 12 as seen in our study, for which an oncogenic function may be assumed. Those shown to have a role in other human tumors include SKP2 mapping to 5p13 (35) and MDM2, CDK4, and SAS mapping to 12q13-q21 (36, 37). K-ras, which maps to 12p11, has been shown not be mutated in adult adrenocortical cancers (13, 38).
Defining the molecular events involved in adrenocortical tumorigenesis may lead to better prognostic markers and therapeutic regimens. The clonal composition of adrenocor- tical tumors has been well characterized (39, 40). It is known that adrenocortical adenomas may be either monoclonal or polyclonal, whereas adrenocortical cancers are monoclonal, suggesting a stepwise progression from adenoma to carci- noma, as a particular clone of cells gains a growth advantage over competing subclones, by advantageous genetic events. Our results support this concept, in that fewer abnormalities were detected in the adenomas vs. the carcinomas, and 12 of 14 (86%) distinct changes in the adenomas were also seen in the carcinomas. In a monoclonal population of cells, the likelihood of detecting chromosomal instability, if it exists, with a technique such as CGH increases. However, as the cell population becomes more diverse, as in the polyclonal sit- uation, it becomes difficult to assign real gains or loss of genetic material because the hybridization signals from dif- ferent subpopulations of cells tend to draw the average value from 9-10 metaphases toward the null value.
Pediatric adrenal CGH analyses have demonstrated ex- tensive genetic aberrations in both adenomas and carcino- mas, with a striking gain of 9q in 5 of 5 adenomas and 14 of 15 carcinomas reported in 2 series (19, 20). In addition, child-
hood adrenal cancers seem to be associated with a high incidence of germline p53 mutations (41, 42), and it is be- lieved that the tissue of origin is fetal adrenal cortex (20).
In a recent review of the molecular basis of adrenocortical tumors, Kjellman and colleagues (43) suggested a progres- sion model for adrenocortical tumorigenesis. Early genetic events allowing the formation of an adenoma from normal adrenal tissue were amplification of chromosomes 17 and 9q, whereas later events leading to malignant transformation of adrenal adenomas were high IGF-II overexpression, p53 point mutation, and loss of 2p16, 11q13, and 1p21-31. This model was based partly on the findings of 2 adult and 2 pediatric adrenal CGH series (15, 16, 19, 20). We believe that the genetic changes reported in the pediatric CGH experience are not relevant to the adult situation. Therefore, the only evidence for amplification of chromosomes 17 and 9q as early events in the adult adrenal tumor progression model comes from the results of Zhao et al. (16), as even Kjellman’s own data did not reflect these findings in their cohort of adenomas or carcinomas (15).
Our data support the concept of a progression model. We suggest that activation of a protooncogene(s) on chromo- some 4 may be an early event in adrenocortical tumorigen- esis, with progression from adenoma to carcinoma involving gain of function of protooncogenes on chromosomes 5 and 12 and loss of function of tumor suppressor genes on chro- mosome arms 1p and 17p. We have also found that genetic aberrations are more commonly seen with the malignant adrenal phenotype and with increasing tumor size, and that the presence of four or more CGH alterations in one tumor is strongly suggestive of the malignant phenotype.
Acknowledgments
Chris Francis is thanked for technical assistance with karyotyping. Anna Guinea is acknowledged for statistical assistance.
Received December 21, 2001. Accepted April 4, 2002.
Address all correspondence and requests for reprints to: Prof. Bruce G. Robinson, Cancer Genetics, Kolling Institute of Medical Research, Royal North Shore Hospital, St. Leonards, New South Wales 2065, Aus- tralia. E-mail: bgr@med.usyd.edu.au.
This work was supported by the Royal Australasian College of Sur- geons Research Foundation and the National Health and Medical Re- search Council of Australia. At the time of this writing, S.S. was a recipient of the Sir Roy McCaughey Research Grant and D.J.M. was an R. D. Wright Fellow.
References
1. Angeli A, Osella G, Ali A, Terzolo M 1997 Adrenal incidentaloma: an over- view of clinical and epidemiological data from the National Italian Study Group. Horm Res 47:279-283
2. Dackiw APB, Lee JE, Gagel RF, Evans DB 2001 Adrenal cortical carcinoma. World J Surg 25:914-926
3. Wajchenberg BL, Albergaria Pereira MA, Medonca BB, Latronico AC, Cam- pos Carneiro P, Ferreira Alves VA, Zerbini MC, Liberman B, Carlos Gomes G, Kirschner MA 2000 Adrenocortical carcinoma: clinical and laboratory ob- servations. Cancer 88:711-736
4. Reincke M, Beuschlein F, Slawik M, Borm K 2000 Molecular adrenocortical tumourigenesis. Eur J Clin Invest 30:63-68
5. Maher ER, Reik W 2000 Beckwith-Wiedemann syndrome: imprinting in clus- ters revisited. J Clin Invest 105:247-252
6. Malkin D, Li FP, Strong LC, Fraumeni Jr JF, Nelson CE, Kim DH, Kassel J, Gryka MA, Bischoff FZ, Tainsky MA 1990 Germ line p53 mutations in a familial syndrome of breast cancer, sarcomas, and other neoplasms. Science 250:1233-1238
7. Skogseid B, Larsson C, Lindgren PG, Kvanta E, Rastad J, Theodorsson E,
Wide L, Wilander E, Oberg K 1992 Clinical and genetic features of adreno- cortical lesions in multiple endocrine neoplasia type 1. J Clin Endocrinol Metab 75:76-81
8. Chandrasekharappa SC, Guru SC, Manickam P, Olufemi SE, Collins FS, Emmert-Buck MR, Debelenko LV, Zhuang Z, Lubensky IA, Liotta LA, Crabtree JS, Wang Y, Roe BA, Weisemann J, Boguski MS, Agarwal SK, Kester MB, Kim YS, Heppner C, Dong Q, Spiegel AM, Burns AL, Marx SJ 1997 Positional cloning of the gene for multiple endocrine neoplasia-type 1. Science 276:404-407
9. Stratakis CA, Carney JA, Lin JP, Papanicolaou DA, Karl M, Kastner DL, Pras E, Chrousos GP 1996 Carney complex, a familial multiple neoplasia and lentiginosis syndrome. Analysis of 11 kindreds and linkage to the short arm of chromosome 2. J Clin Invest 97:699-705
10. Kirschner LS, Carney JA, Pack SD, Taymans SE, Giatzakis C, Cho YS, Cho-Chung YS, Stratakis CA 2000 Mutations of the gene encoding the protein kinase A type I-a regulatory subunit in patients with the Carney complex. Nat Genet 26:89-92
11. Stratakis CA 2001 Clinical genetics of multiple endocrine neoplasias, Carney complex and related syndromes. J Endocrinol Invest 24:370-383
12. Gicquel C, Bertagna X, Schneid H, Francillardleblond M, Luton JP, Girard F, Lebouc Y 1994 Rearrangements at the 11p15 locus and overexpression of insulin-like growth factor-II gene in sporadic adrenocortical tumors. J Clin Endocrinol Metab 78:1444-1453
13. Ohgaki H, Kleihues P, Heitz PU 1993 p53 mutations in sporadic adrenocor- tical tumors. Int J Cancer 54:408-410
14. Reincke M, Karl M, Travis WH, Mastorakos G, Allolio B, Linehan HM, Chrousos GP 1994 p53 mutations in human adrenocortical neoplasms: im- munohistochemical and molecular studies. J Clin Endocrinol Metab 78: 790-794
15. Kjellman M, Kallioniemi OP, Karhu R, Hoog A, Farnebo LO, Auer G, Larsson C, Backdahl M 1996 Genetic aberrations in adrenocortical tumors detected using comparative genomic hybridization correlate with tumor size and malignancy. Cancer Res 56:4219-4223
16. Zhao J, Speel EJ, Muletta-Feurer S, Rutimann K, Saremaslani P, Roth J, Heitz PU, Komminoth P 1999 Analysis of genomic alterations in sporadic adreno- cortical lesions. Gain of chromosome 17 is an early event in adrenocortical tumorigenesis. Am J Pathol 155:1039-1045
17. Dohna M, Reincke M, Mincheva A, Allolio B, Solinas-Toldo S, Lichter P 2000 Adrenocortical carcinoma is characterized by a high frequency of chromo- somal gains and high-level amplifications. Genes Chromosomes Cancer 28: 145-152
18. Kallioniemi A, Kallioniemi OP, Sudar D, Rutovitz D, Gray JW, Waldman F, Pinkel D 1992 Comparative genomic hybridization for molecular cytogenetic analysis of solid tumors. Science 258:818-821
19. Figueiredo BC, Stratakis CA, Sandrini R, DeLacerda L, Pianovsky MA, Giatzakis C, Young HM, Haddad BR 1999 Comparative genomic hybridiza- tion analysis of adrenocortical tumors of childhood. J Clin Endocrinol Metab 84:1116-1121
20. James LA, Kelsey AM, Birch JM, Varley JM 1999 Highly consistent genetic alterations in childhood adrenocortical tumours detected by comparative genomic hybridization. Br J Cancer 81:300-304
21. Mathew CG, Chin KS, Easton DF, Thorpe K, Carter C, Liou GI, Fong SL, Bridges CD, Haak H, Kruseman AC 1987 A linked genetic marker for multiple endocrine neoplasia type 2A on chromosome 10. Nature 328:527-528
22. Lack EE 1995 Tumours of the adrenal gland and extra-adrenal paraganglia. In: Rosai J, ed. Atlas of tumour pathology, 3rd Series. Washington DC: Armed Forces Institute of Pathology; vol 19:123-152
23. Weiss LM, Medeiros LJ, Vickery Jr AL 1989 Pathologic features of prognostic significance in adrenocortical carcinoma. Am J Surg Pathol 13:202-206
24. Kallioniemi OP, Kallioniemi A, Piper J, Isola J, Waldman FM, Gray JW, Pinkel D 1994 Optimizing comparative genomic hybridization for analysis of DNA sequence copy number changes in solid tumors. Genes Chromosomes Cancer 10:231-243
25. Gicquel C, Raffinsanson ML, Gaston V, Bertagna X, Plouin PF, Schlum- berger M, Louvel A, Luton JP, Lebouc Y 1997 Structural and functional abnormalities at 11p15 are associated with the malignant phenotype in spo- radic adrenocortical tumors: study of a series of 82 tumors. J Clin Endocrinol Metab 82:2559-2565
26. Joos S, Scherthan H, Speicher MR, Schlegel J, Cremer T, Lichter P 1993 Detection of amplified DNA sequences by reverse chromosome painting using genomic tumor DNA as probe. Hum Genet 90:584-589
27. Bentz M, Plesch A, Stilgenbauer S, Dohner H, Lichter P 1998 Minimal sizes of deletions detected by comparative genomic hybridization. Genes Chromo- somes Cancer 21:172-175
28. Weiss MM, Hermsen MA, Meijer GA, van Grieken NC, Baak JP, Kuipers EJ, van Diest PJ 1999 Comparative genomic hybridisation. Mol Pathol 52:243-251
29. Han S, Semba S, Abe T, Makino N, Furukawa T, Fukushige S, Takahashi H, Sakurada A, Sato M, Shiiba K, Matsuno S, Nimura Y, Nakagawara A, Horii A 1999 Infrequent somatic mutations of the p73 gene in various human cancers. Eur J Surg Oncol 25:194-198
30. Wales MM, Biel MA, el Deiry W, Nelkin BD, Issa JP, Cavenee WK, Kuerbitz
SJ, Baylin SB 1995 p53 activates expression of HIC-1, a new candidate tumour suppressor gene on 17p13.3. Nat Med 1:570-577
31. Morris C, Benjes S, Haataja L, Ledbetter DH, Heisterkamp N, Groffen J 1995 Spatial organization of ABR and CRK genes on human chromosome band 17p13.3. Oncogene 10:1009-1011
32. Gicquel C, Bertagna X, Gaston V, Coste J, Louvel A, Baudin E, Bertherat J, Chapuis Y, Duclos JM, Schlumberger M, Plouin PF, Luton JP, Le Bouc Y 2001 Molecular markers and long-term recurrences in a large cohort of patients with sporadic adrenocortical tumors. Cancer Res 61:6762-6767
33. Schulte KM, Heinze M, Mengel M, Simon D, Scheuring S, Kohrer K, Roher HD 1999 MEN I gene mutations in sporadic adrenal adenomas. Hum Genet 105:603-610
34. Schulte KM, Mengel N, Heinze M, Simon D, Scheuring S, Kohrer K, Roher HD 2000 Complete sequencing and messenger ribonucleic acid expression analysis of the MEN I gene in adrenal cancer. J Clin Endocrinol Metab 85: 441-448
35. Zhang H, Kobayashi R, Galaktionov K, Beach D 1995 p19Skp1 and p45Skp2 are essential elements of the cyclin A-CDK2 S phase kinase. Cell 82:915-925
36. Smith SH, Weiss SW, Jankowski SA, Coccia MA, Meltzer PS 1992 SAS amplification in soft tissue sarcomas. Cancer Res 52:3746-3749
37. Khatib ZA, Matsushime H, Valentine M, Shapiro DN, Sherr CJ, Look AT
1993 Coamplification of the CDK4 gene with MDM2 and GLI in human sarcomas. Cancer Res 53:5535-5541
38. Moul JW, Bishoff JT, Theune SM, Chang EH 1993 Absent ras gene mutations in human adrenal cortical neoplasms and pheochromocytomas. J Urol 149: 1389-1394
39. Gicquel C, Leblond-Francillard M, Bertagna X, Louvel A, Chapuis Y, Luton JP, Girard F, Le Bouc Y 1994 Clonal analysis of human adrenocortical carci- nomas and secreting adenomas. Clin Endocrinol (Oxf) 40:465-477
40. Beuschlein F, Reincke M, Karl M, Travis WD, Jaursch-Hancke C, Abdelh- amid S, Chrousos GP, Allolio B 1994 Clonal composition of human adreno- cortical neoplasms. Cancer Res 54:4927-4932
41. Wagner J, Portwine C, Rabin K, Leclerc JM, Narod SA, Malkin D 1994 High frequency of germline p53 mutations in childhood adrenocortical cancer. J Natl Cancer Inst 86:1707-1710
42. Varley JM, McGown G, Thorncroft M, James LA, Margison GP, Forster G, Evans DG, Harris M, Kelsey AM, Birch JM 1999 Are there low-penetrance TP53 alleles? Evidence from childhood adrenocortical tumors. Am J Hum Genet 65:995-1006
43. Kjellman M, Larsson C, Backdahl M 2001 Genetic background of adreno- cortical tumor development. World J Surg 25:948-956