Genomic Alterations and Complex Subclonal Architecture in Sporadic GH-Secreting Pituitary Adenomas

Genomic Alterations and Complex Subclonal Architecture in Sporadic GH-Secreting Pituitary Adenomas Abstract Purpose The molecular pathogenesis of growth hormone-secreting pituitary adenomas is not fully understood. Cytogenetic alterations might serve as alternative driver events in GNAS mutation–negative somatotroph tumors. Experimental Design We performed cytogenetic profiling of pituitary adenomas obtained from 39 patients with acromegaly and four patients with sporadic gigantism by using array comparative genomic hybridization analysis. We explored intratumor DNA copy-number heterogeneity in two tumor samples by using DNA fluorescence in situ hybridization (FISH). Results Based on copy-number profiles, we found two groups of adenomas: a low–copy-number alteration (CNA) group (<12% of genomic disruption, 63% of tumors) and a high-CNA group (24% to 45% of genomic disruption, 37% of tumors). Arm-level CNAs were the most common abnormalities. GNAS mutation–positive adenomas belonged exclusively to the low-CNA group, whereas a subgroup of GNAS mutation–negative adenomas had a high degree of genomic disruption. We detected chromothripsis-related CNA profiles in two adenoma samples from an AIP mutation–positive patient with acromegaly and a patient with sporadic gigantism. RNA sequencing of these two samples identified 17 fusion transcripts, most of which resulted from chromothripsis-related chromosomal rearrangements. DNA FISH analysis of these samples demonstrated a subclonal architecture with up to six distinct cell populations in each tumor. Conclusion Somatotroph pituitary adenomas display substantial intertumor and intratumor DNA copy-number heterogeneity, as revealed by variable CNA profiles and complex subclonal architecture. The extensive cytogenetic burden in a subgroup of GNAS mutation–negative somatotroph adenomas points to an alternative tumorigenic pathway linked to genomic instability. Growth hormone (GH)-secreting pituitary adenomas are responsible for chronic GH excess leading to gigantism during childhood and acromegaly in adulthood. Pituitary adenomas appear to have a monoclonal origin, arising from either early progenitor cells or fully differentiated hormone-expressing cells (1, 2), but their genetic background is not fully documented. Activating somatic mutations of the GNAS gene, which codes for the G protein–activating α subunit, promote tumor development and GH hypersecretion in ∼30% of somatotroph adenomas (3). No tumor-causing mutations in GH-secreting adenomas without GNAS mutations have been identified, apart from the rare cases due to germline mutations in AIP, MEN1, CDKN1B, or PRKR1A (4–11). Cytogenetic or epigenetic alterations could serve as alternative driver events in GNAS mutation–negative pituitary tumors (12). Recently, germline or somatic microduplications of chromosome region Xq26.3 containing the GPR101 gene were implicated in X-linked acrogigantism (13, 14). This genomic gain in Xq26.3 demonstrates that focal copy-number alterations (CNAs) can promote pituitary tumorigenesis. Several studies using oligonucleotide-based array comparative genomic hybridization (array-CGH) (15–18) and, more recently, next-generation sequencing techniques (9, 10, 19) have shown various profiles of somatic CNAs in different subtypes of pituitary tumors. Given the large number of molecular alterations, it is important to distinguish between changes contributing to tumorigenesis and those that are nonfunctional byproducts. Large-scale sequencing suggests a higher degree of genomic disruption in hormonally active adenomas (19). Only one recent study focused on genomic alterations specifically in 12 samples of GH-secreting adenomas (9). Here, we report the results of cytogenetic profiling based on array-CGH analysis in a large and well-characterized collection of 43 GH-secreting pituitary adenomas derived from patients with gigantism or acromegaly. We aimed at identifying specific patterns of chromosomal alterations that might reflect biological and molecular differences among these tumors, especially as a function of GNAS mutation status. We distinguished two groups of tumors based on the degree of genome disruption. We further studied copy-number oscillations compatible with chromothripsis in two adenomas, one from an AIP mutation–positive patient with acromegaly and one from an AIP mutation–negative patient with gigantism. Using DNA fluorescence in situ hybridization (FISH), we explored intratumor DNA copy-number heterogeneity and obtained insights into the complex subclonal architecture of these tumors. Materials and Methods Patients We analyzed pituitary adenoma samples derived from 43 patients who had undergone transphenoidal surgery for acromegaly or gigantism between 2009 and 2016. The diagnosis of acromegaly was based on the classical symptoms and clinical signs plus elevated age-adjusted IGF-1 concentrations and unsuppressed GH after an oral glucose tolerance test (OGTT) (20). Gigantism was defined by excessively rapid growth for age in children or adolescents (>97th percentile), or a final height >2 standard deviations above normal for the French population, together with abnormal IGF-1 values for age and unsuppressed GH after an OGTT. GH immunoreactivity was confirmed histologically in all samples. Tumors were considered plurihormonal if they were immunopositive for one or more hormones besides GH. Clinical, biological, and morphological data (e.g., sex, age at diagnosis, GH and IGF-1 concentrations, and tumor size and invasion) were systematically collected. Based on pituitary magnetic resonance imaging, the adenomas were classified into microadenomas (maximum diameter, <10 mm), macroadenomas (maximum diameter, 10 to 39 mm), and giant adenomas (maximum diameter, >40 mm). We also analyzed three control pituitary adenoma samples from AIP mutation–positive patients: two patients had ACTH-secreting pituitary adenomas, and one had a nonfunctioning pituitary adenoma. All patients gave written informed consent for genetic analyses. DNA extraction Genomic DNA was extracted from pituitary tumor samples by using the phenol-chloroform-isoamyl alcohol extraction method (21). The concentration and quality of the extracted DNA were determined with a spectrophotometer (ND-1000; NanoDrop Technologies, Wilmington, DE) and evaluated by electrophoresis on 0.7% agarose gel. DNA sequencing All patients were screened for germline AIP and GPR101 mutations as previously described (22). A genetic search for somatic activating mutations of GNAS was carried out in all pituitary tumor samples. The forward primer 5′-CAAATTGATGTGAGCGCTGT-3′ and the reverse primer 5′-AGCATCCTACCGTTGAAGCA-3′ were used to amplify a region of 480 bp that includes the hotspot mutations in codons 201 and 227 of GNAS exons 7 to 9. The following polymerase chain reaction thermocycling conditions were used: initial denaturation at 94°C for 5 minutes, followed by 35 cycles of denaturation at 94°C for 30 seconds, annealing at 58°C for 1 minute and extension at 72°C for 1 minute, followed by a final extension step at 72°C for 7 minutes. Gene mutations were evaluated by conventional Sanger sequencing with an automated capillary sequencer (ABI PRISM 3730S Genetic Analyzer; Applied Biosystems, Foster City, CA). Oligonucleotide-based array-CGH Genomic CNAs were analyzed in all tumor samples by means of array-CGH with 180K oligonucleotide arrays (Agilent Technologies, Massy, France). Tumor DNAs were compared with blood-donor genomic DNA. Hybridization was performed according to the manufacturer’s protocol. Image processing and data analysis used CytoGenomics software (4.0.3.12) (Agilent Technologies). The genomic positions were determined with Build 37 of the human genome. The ADM2 algorithm was used for statistical analysis. CNAs were considered significant if they were defined by four or more consecutive oligonucleotides and spanned at least 13 kb. CNAs reported to be polymorphic copy-number variations in the Database of Genomic Variants (http://dgv.tcag.ca/gb2/gbrowse/dgv2_hg19/) were excluded from subsequent analysis. Total RNA library preparation and sequencing RNA sequencing was performed by IntegraGen (Evry, France) on the two tumor samples affected by chromothripsis. Libraries were prepared with NEBNext Ultra II Directional RNA Library Prep Kit for Illumina according to supplier’s recommendations (NEB). Sequencing was then carried out in paired-end 75b mode with an Illumina HiSeq 4000. Base calling was performed using the Real-Time Analysis software sequence pipeline (version 2.7.7) from Illumina with default parameters. A subset of 500,000 reads from each Fastq file was aligned to the reference human genome hg19/GRCh37 with tophat2 (-p 24 -r 150 -g 2–library-type fr-firststrand) (23). Fusions initially detected by TopHat2 were filtered using the TopHatFusion-post algorithm. Only the most reliable fusions were kept for further analysis (i.e., fusions that were validated by BLAST and with at least 10 pairs of read spanning and validating the fusion event). DNA FISH DNA FISH was performed on interphase nuclei from tumors 33 and 43, which had a chromothripsis profile. The following probes were used in accordance with the manufacturer’s recommendations: SNRPN 15q11.2 (coupled with the 15q subtelomeric region (Cytocell; Amplitech, Compiègne, France) and subtelomeric probes specific for chromosomes 2p, 6p, 6q, 10q, and 11p (Vysis; Abbott Laboratories, Chicago, IL). Bacterial artificial chromosome clones RP11-91N9 (2q37.1), RP11-192G6 (10q25.1), RP11-482O1 (11q13.2, AIP gene), RP11-368H5 (11q25), and RP11-783O1 (15q21.3) were used in accordance with the manufacturer’s recommendations (RainbowFISH; Amplitech). Statistical methods Continuous variables are expressed as mean ± standard deviation if normally distributed or as the median (interquartile range). Normal distribution was assessed with the Shapiro-Wilk test. Categorical variables are expressed as counts and percentages. Quantitative data were analyzed with the nonparametric Mann-Whitney U test or with the Kruskal-Wallis test. Qualitative comparisons used the χ2 test or Fisher's exact test as appropriate. Statistical significance was assumed at P < 0.05. All statistical analyses used Prism 6 (GraphPad Software Inc.). Results Clinical features The characteristics of the patients are summarized in Supplemental Table 1. There were 29 men and 14 women. Mean age at diagnosis was 42 ± 14 years. Four patients had nonsyndromic gigantism, and 39 had adult-onset acromegaly due to pure GH-secreting or plurihormonal adenomas. The mean age-adjusted IGF-1 level was 340 ± 115% of the upper limit of normal. Genetic mutations Acromegalic patient 33 harbored a missense germline c.2T>C mutation in the AIP start codon that abrogated protein expression (24). Acromegalic patient 36 harbored a germline c.924G>C mutation in GPR101 (22). Somatic activating GNAS mutations were found in 13 pituitary tumors (30.2%). These heterozygous missense mutations were p.Arg201Cys in nine patients (69.2%) and p.Gln227Leu in four patients (30.7%). Somatic CNA profiles Somatic CNAs are described in detail in Supplemental Tables 2 and 3. The most common abnormalities were arm-level losses of chromosomes 1, 15q, and 16 and extensive gains of chromosomes 5, 7, 10, 19, 20, and X (Fig. 1A and Supplemental Fig. 1). Most chromosome arms showed strong evidence of preferential gain or loss but rarely showed both. Among the focal CNAs, recurrent microduplication in 10p11.22 was found in 20 tumors, in addition to genomic gain of the entire arm of 10p in three patients (Fig. 1A). The smallest regions of overlap (hg19 chr10: 33101459-33137499) encompassed one gene (CCDC7). Quantitative reverse transcription polymerase chain reaction quantification of CCDC7 transcripts showed no overexpression in samples with genomic gain in 10p11.22. Figure 1. View largeDownload slide CNAs in 43 GH-secreting adenomas. (A) Whole-genome view derived from array-CGH data, with a summary of chromosomal imbalances (significant CNAs). Chromosomal gains (blue bars) and losses (red bars) across the genome are indicated. Each bar represents one CNA. The adenomas were classified into those with a low (CNAs <12%), and high (CNAs ≥24%) level of genomic disruption. (B) Percentages of genome disruption in the two groups of GH-secreting adenomas. Error bars and central values represent medians and interquartile ranges. (C and D) Proportion of significant CNAs in each group according (C) to gains and losses and (D) to size. Figure 1. View largeDownload slide CNAs in 43 GH-secreting adenomas. (A) Whole-genome view derived from array-CGH data, with a summary of chromosomal imbalances (significant CNAs). Chromosomal gains (blue bars) and losses (red bars) across the genome are indicated. Each bar represents one CNA. The adenomas were classified into those with a low (CNAs <12%), and high (CNAs ≥24%) level of genomic disruption. (B) Percentages of genome disruption in the two groups of GH-secreting adenomas. Error bars and central values represent medians and interquartile ranges. (C and D) Proportion of significant CNAs in each group according (C) to gains and losses and (D) to size. Array-CGH identified two distinct groups of tumors, based on the fraction of genome disruption, with significant quantitative and qualitative differences in the CNA profiles (Fig. 1A and 2B). Twenty-seven tumors (63%) had a low level of genomic disruption with total CNAs affecting <12% of the genome, and 16 tumors (37%) had a high level (total CNAs ≥24%) of genomic disruption (10, 19). The number of CNAs ranged from 0 to 51 per tumor in the low-CNA group and from 15 to 78 per tumor in the high CNA group. Losses predominated in the low-CNA group, whereas gains predominated in the high-CNA group (Fig. 1C). Tumor 2 showed only genomic gains. Extensive arm-level chromosome alterations were more frequent in the high-CNA group (Fig. 1D). No differences across the two groups were observed in terms of clinical features, pretreatment IGF-1 concentration, tumor size, invasive behavior, and Ki76 proliferation index. However, patients who displayed a paradoxical response of GH to oral glucose load during an OGTT harbored adenomas that belonged in the majority to the high-CNA group. In addition, plurihormonal tumors were more frequent in the groups with low numbers of somatic CNAs, and a densely granulated histotype was more prevalent in the high CNA group. All the GNAS mutation–positive tumors belonged to the low-CNA group, and no GNAS mutations were identified in the high-CNA group (Table 1 and Supplemental Table 1). Figure 2. View largeDownload slide Chromothripsis on chromosome 10 in the tumor of patient 33. (A) Whole-genome view derived from array-CGH data showing chromothripsis on chromosome 10 (surrounding region). Chromosomal gains (blue bars) and losses (red bars) across the genome are indicated. Each bar represents one CNA. (B) Detailed view of whole-chromosome 10 ratio plots depicting a complex pattern of alternating copy-number gains and losses. The asterisk represents the physical mapping position of the 10q25.1 probe. (C) Representative examples of FISH results for the adenoma of patient 33. Two loci (green pinpoints) in the terminal region of the short arm of chromosome 6 with no CNAs were visible in 100% of interphase nuclei serving as a positive control. Fifty percent of adenoma cells displayed monosomy in the terminal region of the long arm of chromosome 6 (red pinpoints), consistent with the array-CGH results. Monosomy of the AIP locus 11q13 (red pinpoints) and of the chromothripsis-affected region 10q25.1 (green pinpoints) were seen in 58% and 26% of adenoma cells, respectively. As for the terminal region of the long arm of chromosome 10 (red pinpoints) neighboring the chromothripsis-affected region, 71% of cells showed disomy, whereas monosomy and trisomy were seen in 19% and 10%, respectively, of interphase nuclei. Finally, FISH analysis of the terminal region of the short arm of chromosome 2 (green pinpoint), whose array-CGH ratios suggested mosaicism, showed one locus in 62% of cells, two loci in 21%, three loci in 5%, and no 2p ter loci in 12%. Figure 2. View largeDownload slide Chromothripsis on chromosome 10 in the tumor of patient 33. (A) Whole-genome view derived from array-CGH data showing chromothripsis on chromosome 10 (surrounding region). Chromosomal gains (blue bars) and losses (red bars) across the genome are indicated. Each bar represents one CNA. (B) Detailed view of whole-chromosome 10 ratio plots depicting a complex pattern of alternating copy-number gains and losses. The asterisk represents the physical mapping position of the 10q25.1 probe. (C) Representative examples of FISH results for the adenoma of patient 33. Two loci (green pinpoints) in the terminal region of the short arm of chromosome 6 with no CNAs were visible in 100% of interphase nuclei serving as a positive control. Fifty percent of adenoma cells displayed monosomy in the terminal region of the long arm of chromosome 6 (red pinpoints), consistent with the array-CGH results. Monosomy of the AIP locus 11q13 (red pinpoints) and of the chromothripsis-affected region 10q25.1 (green pinpoints) were seen in 58% and 26% of adenoma cells, respectively. As for the terminal region of the long arm of chromosome 10 (red pinpoints) neighboring the chromothripsis-affected region, 71% of cells showed disomy, whereas monosomy and trisomy were seen in 19% and 10%, respectively, of interphase nuclei. Finally, FISH analysis of the terminal region of the short arm of chromosome 2 (green pinpoint), whose array-CGH ratios suggested mosaicism, showed one locus in 62% of cells, two loci in 21%, three loci in 5%, and no 2p ter loci in 12%. Table 1. Clinical and Histopathologic Comparisons of GH-Secreting Adenomas Classified by Levels of Genomic Disruptiona   Low-CNA (n = 27)  High-CNA (n = 16)  P Value  Clinical parameters         Age, y  38 (28–52)  40 (32–51)  0.43   Male sex, n (%)  16 (59)  13 (81)  0.19   Remission after surgery, n (%)  11 (41)  6 (38)  1.00  Biological parameters         IGF-1% ULN, ng/mL  338 (239–544)  378 (308–379)  0.43   Basal GH, mUI/L  42 (22–64)  17 (11–61)  0.09   Nadir GH, mUI/L  34 (17–55)  14 (9–45)  0.07   Paradoxical response of GH to OGTTb  3/26 (12)  8/14 (57)  0.0070  Radiological parameters         Invasive tumor, n (%)  16 (59)  10 (56)  1.00   Maximum tumor diameter at diagnosis, mm  17 (11–23)  18 (12–26)  0.51   Macroadenoma, n (%)  24 (89)  15 (94)  1.00  Molecular parameters         GNAS mutations, n (%)  13 (48)  0  0.0006   % genome disruption  1.5 (0.005–4.500)  39 (33–44)  <0.0001  Histological parametersc         Mixed/plurihormonal tumor, n (%)  10/24 (42)  2 (13)  0.07   Somatotroph densely granulated, n (%)  6/24 (25)  10 (63)  0.0245   Somatotroph sparsely granulated, n (%)  6/24 (25)  2 (13)  0.43   Somatotroph not otherwise specified, n (%)  2/24 (8)  2 (13)  1.00   Ki67 ≥3%  9/24 (4)  8 (50)  0.53    Low-CNA (n = 27)  High-CNA (n = 16)  P Value  Clinical parameters         Age, y  38 (28–52)  40 (32–51)  0.43   Male sex, n (%)  16 (59)  13 (81)  0.19   Remission after surgery, n (%)  11 (41)  6 (38)  1.00  Biological parameters         IGF-1% ULN, ng/mL  338 (239–544)  378 (308–379)  0.43   Basal GH, mUI/L  42 (22–64)  17 (11–61)  0.09   Nadir GH, mUI/L  34 (17–55)  14 (9–45)  0.07   Paradoxical response of GH to OGTTb  3/26 (12)  8/14 (57)  0.0070  Radiological parameters         Invasive tumor, n (%)  16 (59)  10 (56)  1.00   Maximum tumor diameter at diagnosis, mm  17 (11–23)  18 (12–26)  0.51   Macroadenoma, n (%)  24 (89)  15 (94)  1.00  Molecular parameters         GNAS mutations, n (%)  13 (48)  0  0.0006   % genome disruption  1.5 (0.005–4.500)  39 (33–44)  <0.0001  Histological parametersc         Mixed/plurihormonal tumor, n (%)  10/24 (42)  2 (13)  0.07   Somatotroph densely granulated, n (%)  6/24 (25)  10 (63)  0.0245   Somatotroph sparsely granulated, n (%)  6/24 (25)  2 (13)  0.43   Somatotroph not otherwise specified, n (%)  2/24 (8)  2 (13)  1.00   Ki67 ≥3%  9/24 (4)  8 (50)  0.53  Abbreviation: ULN, upper limit of normal. a Data are presented as median (interquartile range) for continuous variables and as n (percentage) for categorical variables. b OGTT data are not available for one patient in the low-CNA group and for two patients in the high-CNA group. c Histological data are missing for three patients in the low-CNA group. View Large Table 1. Clinical and Histopathologic Comparisons of GH-Secreting Adenomas Classified by Levels of Genomic Disruptiona   Low-CNA (n = 27)  High-CNA (n = 16)  P Value  Clinical parameters         Age, y  38 (28–52)  40 (32–51)  0.43   Male sex, n (%)  16 (59)  13 (81)  0.19   Remission after surgery, n (%)  11 (41)  6 (38)  1.00  Biological parameters         IGF-1% ULN, ng/mL  338 (239–544)  378 (308–379)  0.43   Basal GH, mUI/L  42 (22–64)  17 (11–61)  0.09   Nadir GH, mUI/L  34 (17–55)  14 (9–45)  0.07   Paradoxical response of GH to OGTTb  3/26 (12)  8/14 (57)  0.0070  Radiological parameters         Invasive tumor, n (%)  16 (59)  10 (56)  1.00   Maximum tumor diameter at diagnosis, mm  17 (11–23)  18 (12–26)  0.51   Macroadenoma, n (%)  24 (89)  15 (94)  1.00  Molecular parameters         GNAS mutations, n (%)  13 (48)  0  0.0006   % genome disruption  1.5 (0.005–4.500)  39 (33–44)  <0.0001  Histological parametersc         Mixed/plurihormonal tumor, n (%)  10/24 (42)  2 (13)  0.07   Somatotroph densely granulated, n (%)  6/24 (25)  10 (63)  0.0245   Somatotroph sparsely granulated, n (%)  6/24 (25)  2 (13)  0.43   Somatotroph not otherwise specified, n (%)  2/24 (8)  2 (13)  1.00   Ki67 ≥3%  9/24 (4)  8 (50)  0.53    Low-CNA (n = 27)  High-CNA (n = 16)  P Value  Clinical parameters         Age, y  38 (28–52)  40 (32–51)  0.43   Male sex, n (%)  16 (59)  13 (81)  0.19   Remission after surgery, n (%)  11 (41)  6 (38)  1.00  Biological parameters         IGF-1% ULN, ng/mL  338 (239–544)  378 (308–379)  0.43   Basal GH, mUI/L  42 (22–64)  17 (11–61)  0.09   Nadir GH, mUI/L  34 (17–55)  14 (9–45)  0.07   Paradoxical response of GH to OGTTb  3/26 (12)  8/14 (57)  0.0070  Radiological parameters         Invasive tumor, n (%)  16 (59)  10 (56)  1.00   Maximum tumor diameter at diagnosis, mm  17 (11–23)  18 (12–26)  0.51   Macroadenoma, n (%)  24 (89)  15 (94)  1.00  Molecular parameters         GNAS mutations, n (%)  13 (48)  0  0.0006   % genome disruption  1.5 (0.005–4.500)  39 (33–44)  <0.0001  Histological parametersc         Mixed/plurihormonal tumor, n (%)  10/24 (42)  2 (13)  0.07   Somatotroph densely granulated, n (%)  6/24 (25)  10 (63)  0.0245   Somatotroph sparsely granulated, n (%)  6/24 (25)  2 (13)  0.43   Somatotroph not otherwise specified, n (%)  2/24 (8)  2 (13)  1.00   Ki67 ≥3%  9/24 (4)  8 (50)  0.53  Abbreviation: ULN, upper limit of normal. a Data are presented as median (interquartile range) for continuous variables and as n (percentage) for categorical variables. b OGTT data are not available for one patient in the low-CNA group and for two patients in the high-CNA group. c Histological data are missing for three patients in the low-CNA group. View Large Identification and characterization of chromothripsis in two pituitary tumors Adenoma samples from patients 33 (Fig. 2) and 42 (Fig. 3) showed multiple noncontiguous CNAs compatible with chromothripsis. Chromothripsis is a phenomenon whereby tens to hundreds of clustered chromosomal rearrangements (gains and deletions) involving localized genomic regions generating frequent oscillations between two copy-number states can be acquired in a one-off cellular catastrophe (25). Neither of these patients had received radiotherapy prior to surgery. Figure 3. View largeDownload slide Chromothripsis on chromosomes 2, 11, and 15 in the tumor of patient 42. (A) Whole-genome view derived from array-CGH data, with chromothripsis on chromosomes 2, 11, and 15 (surrounding regions). Chromosomal gains (blue bars) and losses (red bars) across the genome are indicated. Each bar represents one CNA. (B) Detailed view of whole-chromosome 2, 11, and 15 ratio plots depicting a complex pattern of alternating copy-number gains and losses. The physical mapping positions of probes 2q37.1, 11q25, 15q11.2, and 15q21.3 used for FISH analysis are indicated by asterisks. (C) Representative FISH results for the adenoma of patient 42. Two loci in the terminal region of the short arm of chromosome 2 with no CNAs (green pinpoints) and from the terminal region of the long arm of chromosome 2 (red pinpoints) affected by chromothripsis, were visible in 100% (positive control) and 75% of adenoma cells, respectively. We analyzed the terminal region of the short arm of chromosome 11 (green pinpoints) and the 11q25 locus (red pinpoints), both affected by chromothripsis. The former presented a complex pattern, with 0, 1, 2, and 3 visible loci in 12%, 32%, 14%, and 42% of adenoma cells, respectively, whereas the latter showed monosomy in 18% of cells and disomy in 82% of cells. FISH analysis of the 15q21.3 locus (red pinpoints) and the terminal region of the long arm of chromosome 15 (green pinpoints), both affected by chromothripsis, showed monosomy in 45% and 80% of cells, respectively. The 2pter locus (green pinpoints) again served as a control, in addition to the 15q11.2 locus (red pinpoints), both regions being devoid of CNAs. Both showed two pinpoints in all interphase nuclei. Figure 3. View largeDownload slide Chromothripsis on chromosomes 2, 11, and 15 in the tumor of patient 42. (A) Whole-genome view derived from array-CGH data, with chromothripsis on chromosomes 2, 11, and 15 (surrounding regions). Chromosomal gains (blue bars) and losses (red bars) across the genome are indicated. Each bar represents one CNA. (B) Detailed view of whole-chromosome 2, 11, and 15 ratio plots depicting a complex pattern of alternating copy-number gains and losses. The physical mapping positions of probes 2q37.1, 11q25, 15q11.2, and 15q21.3 used for FISH analysis are indicated by asterisks. (C) Representative FISH results for the adenoma of patient 42. Two loci in the terminal region of the short arm of chromosome 2 with no CNAs (green pinpoints) and from the terminal region of the long arm of chromosome 2 (red pinpoints) affected by chromothripsis, were visible in 100% (positive control) and 75% of adenoma cells, respectively. We analyzed the terminal region of the short arm of chromosome 11 (green pinpoints) and the 11q25 locus (red pinpoints), both affected by chromothripsis. The former presented a complex pattern, with 0, 1, 2, and 3 visible loci in 12%, 32%, 14%, and 42% of adenoma cells, respectively, whereas the latter showed monosomy in 18% of cells and disomy in 82% of cells. FISH analysis of the 15q21.3 locus (red pinpoints) and the terminal region of the long arm of chromosome 15 (green pinpoints), both affected by chromothripsis, showed monosomy in 45% and 80% of cells, respectively. The 2pter locus (green pinpoints) again served as a control, in addition to the 15q11.2 locus (red pinpoints), both regions being devoid of CNAs. Both showed two pinpoints in all interphase nuclei. Patient 33 is a 42-year-old woman carrying a germline c.2T>C AIP mutation and a mixed GH- and PRL-secreting adenoma. Chromothripsis was observed on the long arm of chromosome 10 in its distal region (10q23.33q26.3) and included 39 CNAs (21 gains and 18 losses) and at least 78 breakpoints (Fig. 2A and 2B). Ratio plots suggested that chromothripsis was present in a mosaic state. A deletion of 40.5 Mb was also detected in the terminal region 10q23.33q26.3, suggesting instability in this region. No hallmarks of chromothripsis were identified in the patient's germline DNA. In addition, no chromothripsis-related pattern was identified in pituitary adenoma samples from three patients harboring germline AIP mutations. Patient 42 is a 17-year-old man with acromegalogigantism due to an invasive GH-secreting macroadenoma. The genetic search for AIP and GPR101 mutations and Xq26 microduplication was negative. Chromothripsis was detected on the distal long arm of chromosome 2 (region 2q35q37.3), the distal short arm of chromosome 11 (region 11p15.5p15.4), the long arm of chromosome 11 (region 11q13.3q25), and the whole long arm of chromosome 15 (Fig. 3A and 3B). It included eight CNAs (two gains and six losses) and 16 breakpoints in 2q35q37.3, four CNAs (four losses) and eight breakpoints in 11p15.5p15.4, 17 CNAs (6 gains and 11 losses) and 33 breakpoints in 11q13.3q25, and 18 CNAs (one gain and 17 losses) and 34 breakpoints in chromosome 15. Ratio plots also indicated mosaicism. A deletion of 25.5 Mb was detected in the terminal region 2q34q37.3, indicating instability of this region. Chromothripsis was absent in the patient's germline DNA. Gene cartography of chromothripsis-related CNAs in both these tumors included a number of genes described in http://cancer.sanger.ac.uk/census/. These changes include amplification of oncogenes such as TLX1, FGFR2, and TCF7L2; loss of tumor-suppressor genes such as SUFU and VENTX; and deletions of genes involved in chromosome segregation and DNA crosslink repair such as BARD1, DCLRE1A, HDAC4, PAX3, SMC3, and XRCC5. Chromothripsis at chromosome 15q led to loss of the USP8 gene (26). To explore the consequences of chromothripsis, we performed RNA sequencing on the two concerned adenoma samples. A total of 17 fusion transcripts were identified and summarized in Supplemental Table 4. None of the transcripts was previously reported in the Catalogue of Somatic Mutations in Cancer (http://cancer.sanger.ac.uk/cosmic/fusion) or in The Atlas of Genetics and Cytogenetics in Oncology and Haematology (http://atlasgeneticsoncology.org/index.html). In tumor sample 33, seven fusion transcripts were identified. All were intrachromosomal, and five involved gene partners on chromosomes affected by chromothripsis. Two fusion transcripts were predicted to be in-frame, with both involving the TCERG1L gene as a partner gene. TCERG1L has not been previously reported as a fusion gene partner with any other gene. Tumor 42 displayed 10 fusion transcripts. Four were intrachromosomal, and six were interchromosomal. Seven fusion transcripts were generated by chromothripsis. None of these fusions was predicted to be in frame. Further characterization of chromothripsis and DNA copy-number heterogeneity To further study the chromothripsis-related loci, we used DNA FISH to visualize chromosome segments at a single-cell level. In the tumor from patient 33, we analyzed two regions of chromosome 6 as controls. DNA FISH visualized two loci of the terminal segment of chromosome 6p in 100% of adenoma cells and one locus of the terminal segment of chromosome 6p in 50% of adenoma cells, consistent with the arm-level 6q deletion detected with array-CGH. Because this tumor was from an AIP mutation carrier, we further analyzed the 11q13 locus and identified loss of heterozygosity (LOH) at 11q13 in 58% of adenoma cells. Monosomy of 10q25.1 locus, affected by chromothripsis, was detected in 26% of cells. Four different cell populations coexist in this tumor, with either monosomy or disomy at one or both of the examined loci (Fig. 2C). This cellular heterogeneity was further confirmed when studying the 2p and 10q terminal regions. Six different cell populations were distinguished, with various combinations of CNAs ranging from nullosomy to trisomy (Fig. 2C). In the tumor from patient 42, in accordance with the array-CGH findings, DNA FISH showed two loci of the terminal segment of chromosome 2p in all adenoma cells. Chromothripsis resulted in loss of the 2q37.1 locus in 25% of adenoma cells. Loss of the 15q21.3 locus and of the terminal segment of chromosome arm 15q was detected in 45% and 80% of cells, respectively, whereas two loci of region 15q11.2, which was unaffected by chromothripsis, were detected in all adenoma cells. Regarding chromosome 11, DNA FISH revealed the coexistence of six different cell populations, with CNAs ranging from nullosomy to trisomy, in the 11p terminal segment, together with monosomy or disomy at the 11q25 locus, both of which were affected by chromothripsis (Fig. 3C). DNA copy numbers determined by DNA FISH were in agreement with the array-CGH log ratios, which reflect the sum of CNAs occurring at a concerned locus. Discussion Several recent studies using next-generation sequencing methods have failed to discover new tumor-causing gene mutations in GNAS mutation–negative pituitary adenomas (9–11, 19), thus pointing to a role of alternative oncogenic driver events, including cytogenetic alterations. Here, we describe the results of a combined cytogenetic search using array-CGH and DNA FISH in a large and well-characterized collection of GH-secreting pituitary adenomas from 43 patients with gigantism or acromegaly. Based on the copy-number profiles, we found that 63% had CNAs involving <12% of the genome, and the remaining 37% had disruption affecting 24% to 45% of the genome. None of the tumors presented disruptions affecting 12% to 24% of their genome. These findings are in agreement with Bi et al. (19) and Song et al. (10), who similarly identified two groups of pituitary adenomas (including all histotypes) based on the degree of genomic disruption. As previously reported (19, 27), arm-level CNAs were the most common abnormalities. Such chromosome arm-level aberrations usually result from missegregation during anaphase at the level of the centromere (28). This high rate of extensive chromosome alterations among genomically disrupted pituitary adenomas (which are invariably benign) exceeds that observed in several types of cancer (27). Gains and losses were equally prevalent in the low-CNA group, whereas gains predominated with the highly disrupted group. Unlike in malignancies (29), the frequency of arm-level CNAs was not related to the length of the chromosome arm. On the contrary, focal amplifications or deletions were rare, with the exception of 10p11.22 microduplication that was found in 20 adenomas, including four adenomas from the low-CNA group. This duplication encompassing CCDC7 gene was not associated with increased gene expression. The presence of this microduplication in almost half of the tumors suggests that it might be a benign, previously unreported copy-number variation. These two groups of tumors classified according to the degree of genome disruption did not significantly differ with respect to tumor size, invasiveness, proliferative index, or surgical outcome. However, a paradoxical response of GH to oral glucose load and a densely granulated histotype were more frequent in the high-CNA group, highlighting that particular phenotype of these tumors. In addition, all 13 tumors harboring GNAS mutations belonged to the low-CNA group. Thus, two distinct molecular subclasses could be distinguished on the basis of GNAS mutation status and the degree of genome disruption. The GNAS mutation–positive adenomas had few CNAs, highlighting the oncogenic potency of constitutive GNAS activation (3), which by itself drives tumor growth and GH hypersecretion. GNAS-mutation–negative adenomas with a high degree of genomic disruption have a specific molecular pathogenesis. Given the large number of extensive genome disruptions in these tumors, it remains to be established whether (some) alterations are driving tumorigenesis or are just nonfunctional byproducts that accumulate during tumor development (29). Amplification of chromosome arm 20q, which contains the GNAS locus, was observed in 12 adenomas. As previously reported, none of these tumors with 20q duplication harbored a GNAS mutation, suggesting that genomic gain in 20q could be an alternative mechanism of GNAS activation in mutation-negative GH-secreting pituitary adenomas (27). Analysis of somatic structural variants revealed two tumors with complex chromosomal rearrangements compatible with chromothripsis. A key feature of chromothripsis is the formation of ten to hundreds of locally clustered DNA rearrangements through a single, cataclysmic event. Chromosomes affected by chromothripsis show a characteristic pattern of copy-number oscillations, whereby typically only two or occasionally three copy-number states are detectable along the chromosome in the context of a large number of rearrangements (25, 30). Several mechanisms, including ionizing radiation, premature chromosome compaction, DNA replication stress, telomere shortening, aborting apoptosis, TP53 mutations, micronuclei collapse, and hyperploidy formation, have been implicated (30–32). Although chromothripsis has mainly been described in malignant tumors and linked to aggressive behavior and poor patient survival, similar massive DNA rearrangements have been reported in benign uterine leiomyomas (33). Chromothripsis-like patterns of CNAs were recently described in a GH-secreting adenoma harboring a GNAS mutation (9) and in a TSH-secreting adenoma (34). We observed massive DNA rearrangements resulting from chromothripsis in an adenoma from an AIP mutation–positive patient with acromegaly, localized on the long arm of chromosome 10, and in an adenoma from an AIP mutation–negative patient with gigantism, involving chromosome arms 2q, 11p, 11q, and 15q. The regions affected by chromothripsis contained several simultaneous tumorigenic DNA alterations, including oncogene amplification, loss of tumor suppressor genes, and deletion of genes involved in chromosome segregation and DNA crosslink repair. RNA sequencing and analysis of fusion transcripts did not identify gene fusions with oncogenic potential. Two fusion transcripts were predicted to be in frame and therefore potentially functional if translated, both of which had TCERG1L as a gene fusion partner. TCERG1L is a transcription elongation regulator 1–like protein reported to be downregulated by hypermethylation in colon primary tumors. This gene has not been reported to be partnered with any other genes (35). Because ratio plots indicated that chromothripsis was present in a mosaic state, we applied DNA FISH to these tumor samples to analyze chromosome segments at the single-cell level. We found good concordance between the percentage of subpopulations visualized on DNA FISH and the log-ratios of CNAs observed by CGH. Contrary to some cancer types in which rearrangement outcomes of chromothripsis are detectable in almost all cells (25, 31), we found rearrangements related to chromothripsis solely in tumor subclones of the two pituitary adenomas analyzed here. As also demonstrated in the case of a GNAS mutation–positive GH-secreting adenoma described by Valimaki et al. (9), chromothripsis-related alterations in pituitary tumors are very unlikely to be tumor-initiating events; rather, they confer a growth advantage. Genomic instability is an important source of diversified cell populations, resulting in intratumor heterogeneity (36). DNA copy numbers detected in our FISH experiments distinguished up to six distinct cell subpopulations, demonstrating extensive intratumor genetic diversity in GH-secreting pituitary tumors, that are considered monoclonal in origin (1, 2). This subclonal architecture of GH-secreting pituitary adenomas, previously reconstructed from sequencing data (9), provides certain insights into tumorigenesis. In the tumor from an AIP mutation–positive patient, LOH of 11q13 was detected in 58% of cells, rearrangements derived from chromothripsis on chromosome arm 10q were identified in 26% of cells, and the two alterations coexisted in 20% of cells. The precise dynamics of these chromosomal changes during tumor evolution remain to be determined. However, 11q13 LOH appears to be a late event, suggesting that biallelic AIP inactivation is not a prerequisite for AIP-related tumorigenesis. Because intratumor genetic heterogeneity results in phenotypic diversity, the subclonal architecture of pituitary tumors should be taken into account when studying the pathophysiology and dynamics of tumor evolution, tumor behaviors such as invasiveness and aggressiveness, and the variability of therapeutic responses (36, 37). Our study has some limitations. Cross-contamination from normal tissue of the adenoma samples cannot be completely excluded particularly in the microadenomas. However, 89% of adenomas included in this study were macroadenomas (Table 1), and tumor sampling was performed by highly experienced neurosurgeons, thus minimizing the risk of contamination. Furthermore, because densely granulated adenoma histotype was more prevalent in the high CNA group, better responses to somatostatin analogs would be expected in this group (38). However, responses to somatostatin analogs treatment could not be assessed due the small number of treated patients. In summary, molecular stratification of GH-secreting pituitary adenomas based on the degree of genome disruption can divide them into low- and high-CNA groups. Driving GNAS mutations are associated with lesser genomic disruption, whereas some GNAS mutation–negative tumors exhibit a high degree of predominantly arm-level chromosomal alterations. We identified chromothripsis-related chromosome CNAs in two tumors resulting in fusion transcripts, along with extensive intratumor DNA copy-number heterogeneity, thereby revealing the complex and heterogeneous clonal landscape of GH-secreting pituitary adenomas. Whether the molecular stratification based on CNAs represents a new prognostic tool remains to be determined Abbreviations: Abbreviations: array-CGH array comparative genomic hybridization CNA copy-number alteration FISH fluorescence in situ hybridization GH growth hormone LOH loss of heterozygosity OGTT oral glucose tolerance test Acknowledgments The authors thank Noémie Renault and Sylvain Guibert from GeCo, Integragen, for their technical support. Financial Support: This work was supported by grants from Inserm and Université Paris-Sud. M.H. was the recipient of a fellowship from Fondation pour la Recherche Médicale (FDM20150632800). P.K. was the recipient of a Contrat d’Interface Inserm. Disclosure Summary: The authors have nothing to disclose. References 1. Herman V, Fagin J, Gonsky R, Kovacs K, Melmed S. Clonal origin of pituitary adenomas. J Clin Endocrinol Metab . 1990; 71( 6): 1427– 1433. Google Scholar CrossRef Search ADS PubMed  2. Melmed S. Acromegaly pathogenesis and treatment. J Clin Invest . 2009; 119( 11): 3189– 3202. Google Scholar CrossRef Search ADS PubMed  3. Vallar L, Spada A, Giannattasio G. Altered Gs and adenylate cyclase activity in human GH-secreting pituitary adenomas. Nature . 1987; 330( 6148): 566– 568. Google Scholar CrossRef Search ADS PubMed  4. Lecoq AL, Kamenický P, Guiochon-Mantel A, Chanson P. Genetic mutations in sporadic pituitary adenomas--what to screen for? Nat Rev Endocrinol . 2015; 11( 1): 43– 54. Google Scholar CrossRef Search ADS PubMed  5. Vergès B, Boureille F, Goudet P, Murat A, Beckers A, Sassolas G, Cougard P, Chambe B, Montvernay C, Calender A. Pituitary disease in MEN type 1 (MEN1): data from the France-Belgium MEN1 multicenter study. J Clin Endocrinol Metab . 2002; 87( 2): 457– 465. Google Scholar CrossRef Search ADS PubMed  6. 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. Positional cloning of the gene for multiple endocrine neoplasia-type 1. Science . 1997; 276( 5311): 404– 407. Google Scholar CrossRef Search ADS PubMed  7. Vierimaa O, Georgitsi M, Lehtonen R, Vahteristo P, Kokko A, Raitila A, Tuppurainen K, Ebeling TM, Salmela PI, Paschke R, Gündogdu S, De Menis E, Mäkinen MJ, Launonen V, Karhu A, Aaltonen LA. Pituitary adenoma predisposition caused by germline mutations in the AIP gene. Science . 2006; 312( 5777): 1228– 1230. Google Scholar CrossRef Search ADS PubMed  8. Bertherat J, Horvath A, Groussin L, Grabar S, Boikos S, Cazabat L, Libe R, René-Corail F, Stergiopoulos S, Bourdeau I, Bei T, Clauser E, Calender A, Kirschner LS, Bertagna X, Carney JA, Stratakis CA. Mutations in regulatory subunit type 1A of cyclic adenosine 5′-monophosphate-dependent protein kinase (PRKAR1A): phenotype analysis in 353 patients and 80 different genotypes. J Clin Endocrinol Metab . 2009; 94( 6): 2085– 2091. Google Scholar CrossRef Search ADS PubMed  9. Välimäki N, Demir H, Pitkänen E, Kaasinen E, Karppinen A, Kivipelto L, Schalin-Jäntti C, Aaltonen LA, Karhu A. Whole-genome sequencing of growth hormone (GH)-secreting pituitary adenomas. J Clin Endocrinol Metab . 2015; 100( 10): 3918– 3927. Google Scholar CrossRef Search ADS PubMed  10. Song ZJ, Reitman ZJ, Ma ZY, Chen JH, Zhang QL, Shou XF, Huang CX, Wang YF, Li SQ, Mao Y, Zhou LF, Lian BF, Yan H, Shi YY, Zhao Y. The genome-wide mutational landscape of pituitary adenomas. Cell Res . 2016; 26( 11): 1255– 1259. Google Scholar CrossRef Search ADS PubMed  11. Ronchi CL, Peverelli E, Herterich S, Weigand I, Mantovani G, Schwarzmayr T, Sbiera S, Allolio B, Honegger J, Appenzeller S, Lania AG, Reincke M, Calebiro D, Spada A, Buchfelder M, Flitsch J, Strom TM, Fassnacht M. Landscape of somatic mutations in sporadic GH-secreting pituitary adenomas. Eur J Endocrinol . 2016; 174( 3): 363– 372. Google Scholar CrossRef Search ADS PubMed  12. Yacqub-Usman K, Richardson A, Duong CV, Clayton RN, Farrell WE. The pituitary tumour epigenome: aberrations and prospects for targeted therapy. Nat Rev Endocrinol . 2012; 8( 8): 486– 494. Google Scholar CrossRef Search ADS PubMed  13. Trivellin G, Daly AF, Faucz FR, Yuan B, Rostomyan L, Larco DO, Schernthaner-Reiter MH, Szarek E, Leal LF, Caberg JH, Castermans E, Villa C, Dimopoulos A, Chittiboina P, Xekouki P, Shah N, Metzger D, Lysy PA, Ferrante E, Strebkova N, Mazerkina N, Zatelli MC, Lodish M, Horvath A, de Alexandre RB, Manning AD, Levy I, Keil MF, Sierra ML, Palmeira L, Coppieters W, Georges M, Naves LA, Jamar M, Bours V, Wu TJ, Choong CS, Bertherat J, Chanson P, Kamenický P, Farrell WE, Barlier A, Quezado M, Bjelobaba I, Stojilkovic SS, Wess J, Costanzi S, Liu P, Lupski JR, Beckers A, Stratakis CA. Gigantism and acromegaly due to Xq26 microduplications and GPR101 mutation. N Engl J Med . 2014; 371( 25): 2363– 2374. Google Scholar CrossRef Search ADS PubMed  14. Iacovazzo D, Caswell R, Bunce B, Jose S, Yuan B, Hernández-Ramírez LC, Kapur S, Caimari F, Evanson J, Ferraù F, Dang MN, Gabrovska P, Larkin SJ, Ansorge O, Rodd C, Vance ML, Ramírez-Renteria C, Mercado M, Goldstone AP, Buchfelder M, Burren CP, Gurlek A, Dutta P, Choong CS, Cheetham T, Trivellin G, Stratakis CA, Lopes MB, Grossman AB, Trouillas J, Lupski JR, Ellard S, Sampson JR, Roncaroli F, Korbonits M. Germline or somatic GPR101 duplication leads to X-linked acrogigantism: a clinico-pathological and genetic study. Acta Neuropathol Commun . 2016; 4( 1): 56. Google Scholar CrossRef Search ADS PubMed  15. Hui AB, Pang JC, Ko CW, Ng HK. Detection of chromosomal imbalances in growth hormone-secreting pituitary tumors by comparative genomic hybridization. Hum Pathol . 1999; 30( 9): 1019– 1023. Google Scholar CrossRef Search ADS PubMed  16. Szymas J, Schluens K, Liebert W, Petersen I. Genomic instability in pituitary adenomas. Pituitary . 2002; 5( 4): 211– 219. Google Scholar CrossRef Search ADS PubMed  17. Harada K, Nishizaki T, Ozaki S, Kubota H, Harada K, Okamura T, Ito H, Sasaki K. Cytogenetic alterations in pituitary adenomas detected by comparative genomic hybridization. Cancer Genet Cytogenet . 1999; 112( 1): 38– 41. Google Scholar CrossRef Search ADS PubMed  18. Pack SD, Qin LX, Pak E, Wang Y, Ault DO, Mannan P, Jaikumar S, Stratakis CA, Oldfield EH, Zhuang Z, Weil RJ. Common genetic changes in hereditary and sporadic pituitary adenomas detected by comparative genomic hybridization. Genes Chromosomes Cancer . 2005; 43( 1): 72– 82. Google Scholar CrossRef Search ADS PubMed  19. Bi WL, Horowitz P, Greenwald NF, Abedalthagafi M, Agarwalla PK, Gibson WJ, Mei Y, Schumacher SE, Ben-David U, Chevalier A, Carter S, Tiao G, Brastianos PK, Ligon AH, Ducar M, MacConaill L, Laws ER Jr., Santagata S, Beroukhim R, Dunn IF. Landscape of genomic alterations in pituitary adenomas. Clin Cancer Res . 2017; 23: 1841– 1851 Google Scholar CrossRef Search ADS PubMed  20. Giustina A, Chanson P, Kleinberg D, Bronstein MD, Clemmons DR, Klibanski A, van der Lely AJ, Strasburger CJ, Lamberts SW, Ho KK, Casanueva FF, Melmed S; Acromegaly Consensus Group. Expert consensus document: a consensus on the medical treatment of acromegaly. Nat Rev Endocrinol . 2014; 10( 4): 243– 248. Google Scholar CrossRef Search ADS PubMed  21. Maniatis T, Fritsch EF, Sambrook J. Molecular cloning: a laboratory manual . New York, NY: Cold Spring Harbor Laboratory; 1982. 22. Boissot N, Thomas S, Chovelon V, Lecoq H. NBS-LRR-mediated resistance triggered by aphids: viruses do not adapt; aphids adapt via different mechanisms. BMC Plant Biol . 2016; 16( 1): 25. Google Scholar CrossRef Search ADS PubMed  23. Kim D, Pertea G, Trapnell C, Pimentel H, Kelley R, Salzberg SL. TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome Biol . 2013; 14( 4): R36. Google Scholar CrossRef Search ADS PubMed  24. Lecoq AL, Bouligand J, Hage M, Cazabat L, Salenave S, Linglart A, Young J, Guiochon-Mantel A, Chanson P, Kamenický P. Very low frequency of germline GPR101 genetic variation and no biallelic defects with AIP in a large cohort of patients with sporadic pituitary adenomas. Eur J Endocrinol . 2016; 174( 4): 523– 530. Google Scholar CrossRef Search ADS PubMed  25. Stephens PJ, Greenman CD, Fu B, Yang F, Bignell GR, Mudie LJ, Pleasance ED, Lau KW, Beare D, Stebbings LA, McLaren S, Lin ML, McBride DJ, Varela I, Nik-Zainal S, Leroy C, Jia M, Menzies A, Butler AP, Teague JW, Quail MA, Burton J, Swerdlow H, Carter NP, Morsberger LA, Iacobuzio-Donahue C, Follows GA, Green AR, Flanagan AM, Stratton MR, Futreal PA, Campbell PJ. Massive genomic rearrangement acquired in a single catastrophic event during cancer development. Cell . 2011; 144( 1): 27– 40. Google Scholar CrossRef Search ADS PubMed  26. Reincke M, Sbiera S, Hayakawa A, Theodoropoulou M, Osswald A, Beuschlein F, Meitinger T, Mizuno-Yamasaki E, Kawaguchi K, Saeki Y, Tanaka K, Wieland T, Graf E, Saeger W, Ronchi CL, Allolio B, Buchfelder M, Strom TM, Fassnacht M, Komada M. Mutations in the deubiquitinase gene USP8 cause Cushing’s disease. Nat Genet . 2015; 47( 1): 31– 38. Google Scholar CrossRef Search ADS PubMed  27. Bi WL, Greenwald NF, Ramkissoon SH, Abedalthagafi M, Coy SM, Ligon KL, Mei Y, MacConaill L, Ducar M, Min L, Santagata S, Kaiser UB, Beroukhim R, Laws ER Jr, Dunn IF. Clinical Identification of oncogenic drivers and copy-number alterations in pituitary tumors. Endocrinology . 2017; 158( 7): 2284– 2291. Google Scholar CrossRef Search ADS PubMed  28. Ganem NJ, Godinho SA, Pellman D. A mechanism linking extra centrosomes to chromosomal instability. Nature . 2009; 460( 7252): 278– 282. Google Scholar CrossRef Search ADS PubMed  29. Beroukhim R, Mermel CH, Porter D, Wei G, Raychaudhuri S, Donovan J, Barretina J, Boehm JS, Dobson J, Urashima M, Mc Henry KT, Pinchback RM, Ligon AH, Cho YJ, Haery L, Greulich H, Reich M, Winckler W, Lawrence MS, Weir BA, Tanaka KE, Chiang DY, Bass AJ, Loo A, Hoffman C, Prensner J, Liefeld T, Gao Q, Yecies D, Signoretti S, Maher E, Kaye FJ, Sasaki H, Tepper JE, Fletcher JA, Tabernero J, Baselga J, Tsao MS, Demichelis F, Rubin MA, Janne PA, Daly MJ, Nucera C, Levine RL, Ebert BL, Gabriel S, Rustgi AK, Antonescu CR, Ladanyi M, Letai A, Garraway LA, Loda M, Beer DG, True LD, Okamoto A, Pomeroy SL, Singer S, Golub TR, Lander ES, Getz G, Sellers WR, Meyerson M. The landscape of somatic copy-number alteration across human cancers. Nature . 2010; 463( 7283): 899– 905. Google Scholar CrossRef Search ADS PubMed  30. Korbel JO, Campbell PJ. Criteria for inference of chromothripsis in cancer genomes. Cell . 2013; 152( 6): 1226– 1236. Google Scholar CrossRef Search ADS PubMed  31. Rausch T, Jones DT, Zapatka M, Stütz AM, Zichner T, Weischenfeldt J, Jäger N, Remke M, Shih D, Northcott PA, Pfaff E, Tica J, Wang Q, Massimi L, Witt H, Bender S, Pleier S, Cin H, Hawkins C, Beck C, von Deimling A, Hans V, Brors B, Eils R, Scheurlen W, Blake J, Benes V, Kulozik AE, Witt O, Martin D, Zhang C, Porat R, Merino DM, Wasserman J, Jabado N, Fontebasso A, Bullinger L, Rücker FG, Döhner K, Döhner H, Koster J, Molenaar JJ, Versteeg R, Kool M, Tabori U, Malkin D, Korshunov A, Taylor MD, Lichter P, Pfister SM, Korbel JO. Genome sequencing of pediatric medulloblastoma links catastrophic DNA rearrangements with TP53 mutations. Cell . 2012; 148( 1-2): 59– 71. Google Scholar CrossRef Search ADS PubMed  32. Zhang CZ, Spektor A, Cornils H, Francis JM, Jackson EK, Liu S, Meyerson M, Pellman D. Chromothripsis from DNA damage in micronuclei. Nature . 2015; 522( 7555): 179– 184. Google Scholar CrossRef Search ADS PubMed  33. Mehine M, Kaasinen E, Mäkinen N, Katainen R, Kämpjärvi K, Pitkänen E, Heinonen HR, Bützow R, Kilpivaara O, Kuosmanen A, Ristolainen H, Gentile M, Sjöberg J, Vahteristo P, Aaltonen LA. Characterization of uterine leiomyomas by whole-genome sequencing. N Engl J Med . 2013; 369( 1): 43– 53. Google Scholar CrossRef Search ADS PubMed  34. Sapkota S, Horiguchi K, Tosaka M, Yamada S, Yamada M. Whole-exome sequencing study of thyrotropin-secreting pituitary adenomas. J Clin Endocrinol Metab . 2017; 102( 2): 566– 575. Google Scholar PubMed  35. Yi JM, Dhir M, Guzzetta AA, Iacobuzio-Donahue CA, Heo K, Yang KM, Suzuki H, Toyota M, Kim HM, Ahuja N. DNA methylation biomarker candidates for early detection of colon cancer. Tumour Biol . 2012; 33( 2): 363– 372. Google Scholar CrossRef Search ADS PubMed  36. Burrell RA, McGranahan N, Bartek J, Swanton C. The causes and consequences of genetic heterogeneity in cancer evolution. Nature . 2013; 501( 7467): 338– 345. Google Scholar CrossRef Search ADS PubMed  37. Jamal-Hanjani M, Quezada SA, Larkin J, Swanton C. Translational implications of tumor heterogeneity. Clin Cancer Res . 2015; 21: 1258– 1266. Google Scholar CrossRef Search ADS PubMed  38. Larkin S, Reddy R, Karavitaki N, Cudlip S, Wass J, Ansorge O. Granulation pattern, but not GSP or GHR mutation, is associated with clinical characteristics in somatostatin-naive patients with somatotroph adenomas. Eur J Endocrinol . 2013; 168( 4): 491– 499. Google Scholar CrossRef Search ADS PubMed  Copyright © 2018 Endocrine Society http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Clinical Endocrinology and Metabolism Oxford University Press

Loading next page...
 
/lp/ou_press/genomic-alterations-and-complex-subclonal-architecture-in-sporadic-gh-0xZb3w0YGt
Publisher
Endocrine Society
Copyright
Copyright © 2018 Endocrine Society
ISSN
0021-972X
eISSN
1945-7197
D.O.I.
10.1210/jc.2017-02287
Publisher site
See Article on Publisher Site

Abstract

Abstract Purpose The molecular pathogenesis of growth hormone-secreting pituitary adenomas is not fully understood. Cytogenetic alterations might serve as alternative driver events in GNAS mutation–negative somatotroph tumors. Experimental Design We performed cytogenetic profiling of pituitary adenomas obtained from 39 patients with acromegaly and four patients with sporadic gigantism by using array comparative genomic hybridization analysis. We explored intratumor DNA copy-number heterogeneity in two tumor samples by using DNA fluorescence in situ hybridization (FISH). Results Based on copy-number profiles, we found two groups of adenomas: a low–copy-number alteration (CNA) group (<12% of genomic disruption, 63% of tumors) and a high-CNA group (24% to 45% of genomic disruption, 37% of tumors). Arm-level CNAs were the most common abnormalities. GNAS mutation–positive adenomas belonged exclusively to the low-CNA group, whereas a subgroup of GNAS mutation–negative adenomas had a high degree of genomic disruption. We detected chromothripsis-related CNA profiles in two adenoma samples from an AIP mutation–positive patient with acromegaly and a patient with sporadic gigantism. RNA sequencing of these two samples identified 17 fusion transcripts, most of which resulted from chromothripsis-related chromosomal rearrangements. DNA FISH analysis of these samples demonstrated a subclonal architecture with up to six distinct cell populations in each tumor. Conclusion Somatotroph pituitary adenomas display substantial intertumor and intratumor DNA copy-number heterogeneity, as revealed by variable CNA profiles and complex subclonal architecture. The extensive cytogenetic burden in a subgroup of GNAS mutation–negative somatotroph adenomas points to an alternative tumorigenic pathway linked to genomic instability. Growth hormone (GH)-secreting pituitary adenomas are responsible for chronic GH excess leading to gigantism during childhood and acromegaly in adulthood. Pituitary adenomas appear to have a monoclonal origin, arising from either early progenitor cells or fully differentiated hormone-expressing cells (1, 2), but their genetic background is not fully documented. Activating somatic mutations of the GNAS gene, which codes for the G protein–activating α subunit, promote tumor development and GH hypersecretion in ∼30% of somatotroph adenomas (3). No tumor-causing mutations in GH-secreting adenomas without GNAS mutations have been identified, apart from the rare cases due to germline mutations in AIP, MEN1, CDKN1B, or PRKR1A (4–11). Cytogenetic or epigenetic alterations could serve as alternative driver events in GNAS mutation–negative pituitary tumors (12). Recently, germline or somatic microduplications of chromosome region Xq26.3 containing the GPR101 gene were implicated in X-linked acrogigantism (13, 14). This genomic gain in Xq26.3 demonstrates that focal copy-number alterations (CNAs) can promote pituitary tumorigenesis. Several studies using oligonucleotide-based array comparative genomic hybridization (array-CGH) (15–18) and, more recently, next-generation sequencing techniques (9, 10, 19) have shown various profiles of somatic CNAs in different subtypes of pituitary tumors. Given the large number of molecular alterations, it is important to distinguish between changes contributing to tumorigenesis and those that are nonfunctional byproducts. Large-scale sequencing suggests a higher degree of genomic disruption in hormonally active adenomas (19). Only one recent study focused on genomic alterations specifically in 12 samples of GH-secreting adenomas (9). Here, we report the results of cytogenetic profiling based on array-CGH analysis in a large and well-characterized collection of 43 GH-secreting pituitary adenomas derived from patients with gigantism or acromegaly. We aimed at identifying specific patterns of chromosomal alterations that might reflect biological and molecular differences among these tumors, especially as a function of GNAS mutation status. We distinguished two groups of tumors based on the degree of genome disruption. We further studied copy-number oscillations compatible with chromothripsis in two adenomas, one from an AIP mutation–positive patient with acromegaly and one from an AIP mutation–negative patient with gigantism. Using DNA fluorescence in situ hybridization (FISH), we explored intratumor DNA copy-number heterogeneity and obtained insights into the complex subclonal architecture of these tumors. Materials and Methods Patients We analyzed pituitary adenoma samples derived from 43 patients who had undergone transphenoidal surgery for acromegaly or gigantism between 2009 and 2016. The diagnosis of acromegaly was based on the classical symptoms and clinical signs plus elevated age-adjusted IGF-1 concentrations and unsuppressed GH after an oral glucose tolerance test (OGTT) (20). Gigantism was defined by excessively rapid growth for age in children or adolescents (>97th percentile), or a final height >2 standard deviations above normal for the French population, together with abnormal IGF-1 values for age and unsuppressed GH after an OGTT. GH immunoreactivity was confirmed histologically in all samples. Tumors were considered plurihormonal if they were immunopositive for one or more hormones besides GH. Clinical, biological, and morphological data (e.g., sex, age at diagnosis, GH and IGF-1 concentrations, and tumor size and invasion) were systematically collected. Based on pituitary magnetic resonance imaging, the adenomas were classified into microadenomas (maximum diameter, <10 mm), macroadenomas (maximum diameter, 10 to 39 mm), and giant adenomas (maximum diameter, >40 mm). We also analyzed three control pituitary adenoma samples from AIP mutation–positive patients: two patients had ACTH-secreting pituitary adenomas, and one had a nonfunctioning pituitary adenoma. All patients gave written informed consent for genetic analyses. DNA extraction Genomic DNA was extracted from pituitary tumor samples by using the phenol-chloroform-isoamyl alcohol extraction method (21). The concentration and quality of the extracted DNA were determined with a spectrophotometer (ND-1000; NanoDrop Technologies, Wilmington, DE) and evaluated by electrophoresis on 0.7% agarose gel. DNA sequencing All patients were screened for germline AIP and GPR101 mutations as previously described (22). A genetic search for somatic activating mutations of GNAS was carried out in all pituitary tumor samples. The forward primer 5′-CAAATTGATGTGAGCGCTGT-3′ and the reverse primer 5′-AGCATCCTACCGTTGAAGCA-3′ were used to amplify a region of 480 bp that includes the hotspot mutations in codons 201 and 227 of GNAS exons 7 to 9. The following polymerase chain reaction thermocycling conditions were used: initial denaturation at 94°C for 5 minutes, followed by 35 cycles of denaturation at 94°C for 30 seconds, annealing at 58°C for 1 minute and extension at 72°C for 1 minute, followed by a final extension step at 72°C for 7 minutes. Gene mutations were evaluated by conventional Sanger sequencing with an automated capillary sequencer (ABI PRISM 3730S Genetic Analyzer; Applied Biosystems, Foster City, CA). Oligonucleotide-based array-CGH Genomic CNAs were analyzed in all tumor samples by means of array-CGH with 180K oligonucleotide arrays (Agilent Technologies, Massy, France). Tumor DNAs were compared with blood-donor genomic DNA. Hybridization was performed according to the manufacturer’s protocol. Image processing and data analysis used CytoGenomics software (4.0.3.12) (Agilent Technologies). The genomic positions were determined with Build 37 of the human genome. The ADM2 algorithm was used for statistical analysis. CNAs were considered significant if they were defined by four or more consecutive oligonucleotides and spanned at least 13 kb. CNAs reported to be polymorphic copy-number variations in the Database of Genomic Variants (http://dgv.tcag.ca/gb2/gbrowse/dgv2_hg19/) were excluded from subsequent analysis. Total RNA library preparation and sequencing RNA sequencing was performed by IntegraGen (Evry, France) on the two tumor samples affected by chromothripsis. Libraries were prepared with NEBNext Ultra II Directional RNA Library Prep Kit for Illumina according to supplier’s recommendations (NEB). Sequencing was then carried out in paired-end 75b mode with an Illumina HiSeq 4000. Base calling was performed using the Real-Time Analysis software sequence pipeline (version 2.7.7) from Illumina with default parameters. A subset of 500,000 reads from each Fastq file was aligned to the reference human genome hg19/GRCh37 with tophat2 (-p 24 -r 150 -g 2–library-type fr-firststrand) (23). Fusions initially detected by TopHat2 were filtered using the TopHatFusion-post algorithm. Only the most reliable fusions were kept for further analysis (i.e., fusions that were validated by BLAST and with at least 10 pairs of read spanning and validating the fusion event). DNA FISH DNA FISH was performed on interphase nuclei from tumors 33 and 43, which had a chromothripsis profile. The following probes were used in accordance with the manufacturer’s recommendations: SNRPN 15q11.2 (coupled with the 15q subtelomeric region (Cytocell; Amplitech, Compiègne, France) and subtelomeric probes specific for chromosomes 2p, 6p, 6q, 10q, and 11p (Vysis; Abbott Laboratories, Chicago, IL). Bacterial artificial chromosome clones RP11-91N9 (2q37.1), RP11-192G6 (10q25.1), RP11-482O1 (11q13.2, AIP gene), RP11-368H5 (11q25), and RP11-783O1 (15q21.3) were used in accordance with the manufacturer’s recommendations (RainbowFISH; Amplitech). Statistical methods Continuous variables are expressed as mean ± standard deviation if normally distributed or as the median (interquartile range). Normal distribution was assessed with the Shapiro-Wilk test. Categorical variables are expressed as counts and percentages. Quantitative data were analyzed with the nonparametric Mann-Whitney U test or with the Kruskal-Wallis test. Qualitative comparisons used the χ2 test or Fisher's exact test as appropriate. Statistical significance was assumed at P < 0.05. All statistical analyses used Prism 6 (GraphPad Software Inc.). Results Clinical features The characteristics of the patients are summarized in Supplemental Table 1. There were 29 men and 14 women. Mean age at diagnosis was 42 ± 14 years. Four patients had nonsyndromic gigantism, and 39 had adult-onset acromegaly due to pure GH-secreting or plurihormonal adenomas. The mean age-adjusted IGF-1 level was 340 ± 115% of the upper limit of normal. Genetic mutations Acromegalic patient 33 harbored a missense germline c.2T>C mutation in the AIP start codon that abrogated protein expression (24). Acromegalic patient 36 harbored a germline c.924G>C mutation in GPR101 (22). Somatic activating GNAS mutations were found in 13 pituitary tumors (30.2%). These heterozygous missense mutations were p.Arg201Cys in nine patients (69.2%) and p.Gln227Leu in four patients (30.7%). Somatic CNA profiles Somatic CNAs are described in detail in Supplemental Tables 2 and 3. The most common abnormalities were arm-level losses of chromosomes 1, 15q, and 16 and extensive gains of chromosomes 5, 7, 10, 19, 20, and X (Fig. 1A and Supplemental Fig. 1). Most chromosome arms showed strong evidence of preferential gain or loss but rarely showed both. Among the focal CNAs, recurrent microduplication in 10p11.22 was found in 20 tumors, in addition to genomic gain of the entire arm of 10p in three patients (Fig. 1A). The smallest regions of overlap (hg19 chr10: 33101459-33137499) encompassed one gene (CCDC7). Quantitative reverse transcription polymerase chain reaction quantification of CCDC7 transcripts showed no overexpression in samples with genomic gain in 10p11.22. Figure 1. View largeDownload slide CNAs in 43 GH-secreting adenomas. (A) Whole-genome view derived from array-CGH data, with a summary of chromosomal imbalances (significant CNAs). Chromosomal gains (blue bars) and losses (red bars) across the genome are indicated. Each bar represents one CNA. The adenomas were classified into those with a low (CNAs <12%), and high (CNAs ≥24%) level of genomic disruption. (B) Percentages of genome disruption in the two groups of GH-secreting adenomas. Error bars and central values represent medians and interquartile ranges. (C and D) Proportion of significant CNAs in each group according (C) to gains and losses and (D) to size. Figure 1. View largeDownload slide CNAs in 43 GH-secreting adenomas. (A) Whole-genome view derived from array-CGH data, with a summary of chromosomal imbalances (significant CNAs). Chromosomal gains (blue bars) and losses (red bars) across the genome are indicated. Each bar represents one CNA. The adenomas were classified into those with a low (CNAs <12%), and high (CNAs ≥24%) level of genomic disruption. (B) Percentages of genome disruption in the two groups of GH-secreting adenomas. Error bars and central values represent medians and interquartile ranges. (C and D) Proportion of significant CNAs in each group according (C) to gains and losses and (D) to size. Array-CGH identified two distinct groups of tumors, based on the fraction of genome disruption, with significant quantitative and qualitative differences in the CNA profiles (Fig. 1A and 2B). Twenty-seven tumors (63%) had a low level of genomic disruption with total CNAs affecting <12% of the genome, and 16 tumors (37%) had a high level (total CNAs ≥24%) of genomic disruption (10, 19). The number of CNAs ranged from 0 to 51 per tumor in the low-CNA group and from 15 to 78 per tumor in the high CNA group. Losses predominated in the low-CNA group, whereas gains predominated in the high-CNA group (Fig. 1C). Tumor 2 showed only genomic gains. Extensive arm-level chromosome alterations were more frequent in the high-CNA group (Fig. 1D). No differences across the two groups were observed in terms of clinical features, pretreatment IGF-1 concentration, tumor size, invasive behavior, and Ki76 proliferation index. However, patients who displayed a paradoxical response of GH to oral glucose load during an OGTT harbored adenomas that belonged in the majority to the high-CNA group. In addition, plurihormonal tumors were more frequent in the groups with low numbers of somatic CNAs, and a densely granulated histotype was more prevalent in the high CNA group. All the GNAS mutation–positive tumors belonged to the low-CNA group, and no GNAS mutations were identified in the high-CNA group (Table 1 and Supplemental Table 1). Figure 2. View largeDownload slide Chromothripsis on chromosome 10 in the tumor of patient 33. (A) Whole-genome view derived from array-CGH data showing chromothripsis on chromosome 10 (surrounding region). Chromosomal gains (blue bars) and losses (red bars) across the genome are indicated. Each bar represents one CNA. (B) Detailed view of whole-chromosome 10 ratio plots depicting a complex pattern of alternating copy-number gains and losses. The asterisk represents the physical mapping position of the 10q25.1 probe. (C) Representative examples of FISH results for the adenoma of patient 33. Two loci (green pinpoints) in the terminal region of the short arm of chromosome 6 with no CNAs were visible in 100% of interphase nuclei serving as a positive control. Fifty percent of adenoma cells displayed monosomy in the terminal region of the long arm of chromosome 6 (red pinpoints), consistent with the array-CGH results. Monosomy of the AIP locus 11q13 (red pinpoints) and of the chromothripsis-affected region 10q25.1 (green pinpoints) were seen in 58% and 26% of adenoma cells, respectively. As for the terminal region of the long arm of chromosome 10 (red pinpoints) neighboring the chromothripsis-affected region, 71% of cells showed disomy, whereas monosomy and trisomy were seen in 19% and 10%, respectively, of interphase nuclei. Finally, FISH analysis of the terminal region of the short arm of chromosome 2 (green pinpoint), whose array-CGH ratios suggested mosaicism, showed one locus in 62% of cells, two loci in 21%, three loci in 5%, and no 2p ter loci in 12%. Figure 2. View largeDownload slide Chromothripsis on chromosome 10 in the tumor of patient 33. (A) Whole-genome view derived from array-CGH data showing chromothripsis on chromosome 10 (surrounding region). Chromosomal gains (blue bars) and losses (red bars) across the genome are indicated. Each bar represents one CNA. (B) Detailed view of whole-chromosome 10 ratio plots depicting a complex pattern of alternating copy-number gains and losses. The asterisk represents the physical mapping position of the 10q25.1 probe. (C) Representative examples of FISH results for the adenoma of patient 33. Two loci (green pinpoints) in the terminal region of the short arm of chromosome 6 with no CNAs were visible in 100% of interphase nuclei serving as a positive control. Fifty percent of adenoma cells displayed monosomy in the terminal region of the long arm of chromosome 6 (red pinpoints), consistent with the array-CGH results. Monosomy of the AIP locus 11q13 (red pinpoints) and of the chromothripsis-affected region 10q25.1 (green pinpoints) were seen in 58% and 26% of adenoma cells, respectively. As for the terminal region of the long arm of chromosome 10 (red pinpoints) neighboring the chromothripsis-affected region, 71% of cells showed disomy, whereas monosomy and trisomy were seen in 19% and 10%, respectively, of interphase nuclei. Finally, FISH analysis of the terminal region of the short arm of chromosome 2 (green pinpoint), whose array-CGH ratios suggested mosaicism, showed one locus in 62% of cells, two loci in 21%, three loci in 5%, and no 2p ter loci in 12%. Table 1. Clinical and Histopathologic Comparisons of GH-Secreting Adenomas Classified by Levels of Genomic Disruptiona   Low-CNA (n = 27)  High-CNA (n = 16)  P Value  Clinical parameters         Age, y  38 (28–52)  40 (32–51)  0.43   Male sex, n (%)  16 (59)  13 (81)  0.19   Remission after surgery, n (%)  11 (41)  6 (38)  1.00  Biological parameters         IGF-1% ULN, ng/mL  338 (239–544)  378 (308–379)  0.43   Basal GH, mUI/L  42 (22–64)  17 (11–61)  0.09   Nadir GH, mUI/L  34 (17–55)  14 (9–45)  0.07   Paradoxical response of GH to OGTTb  3/26 (12)  8/14 (57)  0.0070  Radiological parameters         Invasive tumor, n (%)  16 (59)  10 (56)  1.00   Maximum tumor diameter at diagnosis, mm  17 (11–23)  18 (12–26)  0.51   Macroadenoma, n (%)  24 (89)  15 (94)  1.00  Molecular parameters         GNAS mutations, n (%)  13 (48)  0  0.0006   % genome disruption  1.5 (0.005–4.500)  39 (33–44)  <0.0001  Histological parametersc         Mixed/plurihormonal tumor, n (%)  10/24 (42)  2 (13)  0.07   Somatotroph densely granulated, n (%)  6/24 (25)  10 (63)  0.0245   Somatotroph sparsely granulated, n (%)  6/24 (25)  2 (13)  0.43   Somatotroph not otherwise specified, n (%)  2/24 (8)  2 (13)  1.00   Ki67 ≥3%  9/24 (4)  8 (50)  0.53    Low-CNA (n = 27)  High-CNA (n = 16)  P Value  Clinical parameters         Age, y  38 (28–52)  40 (32–51)  0.43   Male sex, n (%)  16 (59)  13 (81)  0.19   Remission after surgery, n (%)  11 (41)  6 (38)  1.00  Biological parameters         IGF-1% ULN, ng/mL  338 (239–544)  378 (308–379)  0.43   Basal GH, mUI/L  42 (22–64)  17 (11–61)  0.09   Nadir GH, mUI/L  34 (17–55)  14 (9–45)  0.07   Paradoxical response of GH to OGTTb  3/26 (12)  8/14 (57)  0.0070  Radiological parameters         Invasive tumor, n (%)  16 (59)  10 (56)  1.00   Maximum tumor diameter at diagnosis, mm  17 (11–23)  18 (12–26)  0.51   Macroadenoma, n (%)  24 (89)  15 (94)  1.00  Molecular parameters         GNAS mutations, n (%)  13 (48)  0  0.0006   % genome disruption  1.5 (0.005–4.500)  39 (33–44)  <0.0001  Histological parametersc         Mixed/plurihormonal tumor, n (%)  10/24 (42)  2 (13)  0.07   Somatotroph densely granulated, n (%)  6/24 (25)  10 (63)  0.0245   Somatotroph sparsely granulated, n (%)  6/24 (25)  2 (13)  0.43   Somatotroph not otherwise specified, n (%)  2/24 (8)  2 (13)  1.00   Ki67 ≥3%  9/24 (4)  8 (50)  0.53  Abbreviation: ULN, upper limit of normal. a Data are presented as median (interquartile range) for continuous variables and as n (percentage) for categorical variables. b OGTT data are not available for one patient in the low-CNA group and for two patients in the high-CNA group. c Histological data are missing for three patients in the low-CNA group. View Large Table 1. Clinical and Histopathologic Comparisons of GH-Secreting Adenomas Classified by Levels of Genomic Disruptiona   Low-CNA (n = 27)  High-CNA (n = 16)  P Value  Clinical parameters         Age, y  38 (28–52)  40 (32–51)  0.43   Male sex, n (%)  16 (59)  13 (81)  0.19   Remission after surgery, n (%)  11 (41)  6 (38)  1.00  Biological parameters         IGF-1% ULN, ng/mL  338 (239–544)  378 (308–379)  0.43   Basal GH, mUI/L  42 (22–64)  17 (11–61)  0.09   Nadir GH, mUI/L  34 (17–55)  14 (9–45)  0.07   Paradoxical response of GH to OGTTb  3/26 (12)  8/14 (57)  0.0070  Radiological parameters         Invasive tumor, n (%)  16 (59)  10 (56)  1.00   Maximum tumor diameter at diagnosis, mm  17 (11–23)  18 (12–26)  0.51   Macroadenoma, n (%)  24 (89)  15 (94)  1.00  Molecular parameters         GNAS mutations, n (%)  13 (48)  0  0.0006   % genome disruption  1.5 (0.005–4.500)  39 (33–44)  <0.0001  Histological parametersc         Mixed/plurihormonal tumor, n (%)  10/24 (42)  2 (13)  0.07   Somatotroph densely granulated, n (%)  6/24 (25)  10 (63)  0.0245   Somatotroph sparsely granulated, n (%)  6/24 (25)  2 (13)  0.43   Somatotroph not otherwise specified, n (%)  2/24 (8)  2 (13)  1.00   Ki67 ≥3%  9/24 (4)  8 (50)  0.53    Low-CNA (n = 27)  High-CNA (n = 16)  P Value  Clinical parameters         Age, y  38 (28–52)  40 (32–51)  0.43   Male sex, n (%)  16 (59)  13 (81)  0.19   Remission after surgery, n (%)  11 (41)  6 (38)  1.00  Biological parameters         IGF-1% ULN, ng/mL  338 (239–544)  378 (308–379)  0.43   Basal GH, mUI/L  42 (22–64)  17 (11–61)  0.09   Nadir GH, mUI/L  34 (17–55)  14 (9–45)  0.07   Paradoxical response of GH to OGTTb  3/26 (12)  8/14 (57)  0.0070  Radiological parameters         Invasive tumor, n (%)  16 (59)  10 (56)  1.00   Maximum tumor diameter at diagnosis, mm  17 (11–23)  18 (12–26)  0.51   Macroadenoma, n (%)  24 (89)  15 (94)  1.00  Molecular parameters         GNAS mutations, n (%)  13 (48)  0  0.0006   % genome disruption  1.5 (0.005–4.500)  39 (33–44)  <0.0001  Histological parametersc         Mixed/plurihormonal tumor, n (%)  10/24 (42)  2 (13)  0.07   Somatotroph densely granulated, n (%)  6/24 (25)  10 (63)  0.0245   Somatotroph sparsely granulated, n (%)  6/24 (25)  2 (13)  0.43   Somatotroph not otherwise specified, n (%)  2/24 (8)  2 (13)  1.00   Ki67 ≥3%  9/24 (4)  8 (50)  0.53  Abbreviation: ULN, upper limit of normal. a Data are presented as median (interquartile range) for continuous variables and as n (percentage) for categorical variables. b OGTT data are not available for one patient in the low-CNA group and for two patients in the high-CNA group. c Histological data are missing for three patients in the low-CNA group. View Large Identification and characterization of chromothripsis in two pituitary tumors Adenoma samples from patients 33 (Fig. 2) and 42 (Fig. 3) showed multiple noncontiguous CNAs compatible with chromothripsis. Chromothripsis is a phenomenon whereby tens to hundreds of clustered chromosomal rearrangements (gains and deletions) involving localized genomic regions generating frequent oscillations between two copy-number states can be acquired in a one-off cellular catastrophe (25). Neither of these patients had received radiotherapy prior to surgery. Figure 3. View largeDownload slide Chromothripsis on chromosomes 2, 11, and 15 in the tumor of patient 42. (A) Whole-genome view derived from array-CGH data, with chromothripsis on chromosomes 2, 11, and 15 (surrounding regions). Chromosomal gains (blue bars) and losses (red bars) across the genome are indicated. Each bar represents one CNA. (B) Detailed view of whole-chromosome 2, 11, and 15 ratio plots depicting a complex pattern of alternating copy-number gains and losses. The physical mapping positions of probes 2q37.1, 11q25, 15q11.2, and 15q21.3 used for FISH analysis are indicated by asterisks. (C) Representative FISH results for the adenoma of patient 42. Two loci in the terminal region of the short arm of chromosome 2 with no CNAs (green pinpoints) and from the terminal region of the long arm of chromosome 2 (red pinpoints) affected by chromothripsis, were visible in 100% (positive control) and 75% of adenoma cells, respectively. We analyzed the terminal region of the short arm of chromosome 11 (green pinpoints) and the 11q25 locus (red pinpoints), both affected by chromothripsis. The former presented a complex pattern, with 0, 1, 2, and 3 visible loci in 12%, 32%, 14%, and 42% of adenoma cells, respectively, whereas the latter showed monosomy in 18% of cells and disomy in 82% of cells. FISH analysis of the 15q21.3 locus (red pinpoints) and the terminal region of the long arm of chromosome 15 (green pinpoints), both affected by chromothripsis, showed monosomy in 45% and 80% of cells, respectively. The 2pter locus (green pinpoints) again served as a control, in addition to the 15q11.2 locus (red pinpoints), both regions being devoid of CNAs. Both showed two pinpoints in all interphase nuclei. Figure 3. View largeDownload slide Chromothripsis on chromosomes 2, 11, and 15 in the tumor of patient 42. (A) Whole-genome view derived from array-CGH data, with chromothripsis on chromosomes 2, 11, and 15 (surrounding regions). Chromosomal gains (blue bars) and losses (red bars) across the genome are indicated. Each bar represents one CNA. (B) Detailed view of whole-chromosome 2, 11, and 15 ratio plots depicting a complex pattern of alternating copy-number gains and losses. The physical mapping positions of probes 2q37.1, 11q25, 15q11.2, and 15q21.3 used for FISH analysis are indicated by asterisks. (C) Representative FISH results for the adenoma of patient 42. Two loci in the terminal region of the short arm of chromosome 2 with no CNAs (green pinpoints) and from the terminal region of the long arm of chromosome 2 (red pinpoints) affected by chromothripsis, were visible in 100% (positive control) and 75% of adenoma cells, respectively. We analyzed the terminal region of the short arm of chromosome 11 (green pinpoints) and the 11q25 locus (red pinpoints), both affected by chromothripsis. The former presented a complex pattern, with 0, 1, 2, and 3 visible loci in 12%, 32%, 14%, and 42% of adenoma cells, respectively, whereas the latter showed monosomy in 18% of cells and disomy in 82% of cells. FISH analysis of the 15q21.3 locus (red pinpoints) and the terminal region of the long arm of chromosome 15 (green pinpoints), both affected by chromothripsis, showed monosomy in 45% and 80% of cells, respectively. The 2pter locus (green pinpoints) again served as a control, in addition to the 15q11.2 locus (red pinpoints), both regions being devoid of CNAs. Both showed two pinpoints in all interphase nuclei. Patient 33 is a 42-year-old woman carrying a germline c.2T>C AIP mutation and a mixed GH- and PRL-secreting adenoma. Chromothripsis was observed on the long arm of chromosome 10 in its distal region (10q23.33q26.3) and included 39 CNAs (21 gains and 18 losses) and at least 78 breakpoints (Fig. 2A and 2B). Ratio plots suggested that chromothripsis was present in a mosaic state. A deletion of 40.5 Mb was also detected in the terminal region 10q23.33q26.3, suggesting instability in this region. No hallmarks of chromothripsis were identified in the patient's germline DNA. In addition, no chromothripsis-related pattern was identified in pituitary adenoma samples from three patients harboring germline AIP mutations. Patient 42 is a 17-year-old man with acromegalogigantism due to an invasive GH-secreting macroadenoma. The genetic search for AIP and GPR101 mutations and Xq26 microduplication was negative. Chromothripsis was detected on the distal long arm of chromosome 2 (region 2q35q37.3), the distal short arm of chromosome 11 (region 11p15.5p15.4), the long arm of chromosome 11 (region 11q13.3q25), and the whole long arm of chromosome 15 (Fig. 3A and 3B). It included eight CNAs (two gains and six losses) and 16 breakpoints in 2q35q37.3, four CNAs (four losses) and eight breakpoints in 11p15.5p15.4, 17 CNAs (6 gains and 11 losses) and 33 breakpoints in 11q13.3q25, and 18 CNAs (one gain and 17 losses) and 34 breakpoints in chromosome 15. Ratio plots also indicated mosaicism. A deletion of 25.5 Mb was detected in the terminal region 2q34q37.3, indicating instability of this region. Chromothripsis was absent in the patient's germline DNA. Gene cartography of chromothripsis-related CNAs in both these tumors included a number of genes described in http://cancer.sanger.ac.uk/census/. These changes include amplification of oncogenes such as TLX1, FGFR2, and TCF7L2; loss of tumor-suppressor genes such as SUFU and VENTX; and deletions of genes involved in chromosome segregation and DNA crosslink repair such as BARD1, DCLRE1A, HDAC4, PAX3, SMC3, and XRCC5. Chromothripsis at chromosome 15q led to loss of the USP8 gene (26). To explore the consequences of chromothripsis, we performed RNA sequencing on the two concerned adenoma samples. A total of 17 fusion transcripts were identified and summarized in Supplemental Table 4. None of the transcripts was previously reported in the Catalogue of Somatic Mutations in Cancer (http://cancer.sanger.ac.uk/cosmic/fusion) or in The Atlas of Genetics and Cytogenetics in Oncology and Haematology (http://atlasgeneticsoncology.org/index.html). In tumor sample 33, seven fusion transcripts were identified. All were intrachromosomal, and five involved gene partners on chromosomes affected by chromothripsis. Two fusion transcripts were predicted to be in-frame, with both involving the TCERG1L gene as a partner gene. TCERG1L has not been previously reported as a fusion gene partner with any other gene. Tumor 42 displayed 10 fusion transcripts. Four were intrachromosomal, and six were interchromosomal. Seven fusion transcripts were generated by chromothripsis. None of these fusions was predicted to be in frame. Further characterization of chromothripsis and DNA copy-number heterogeneity To further study the chromothripsis-related loci, we used DNA FISH to visualize chromosome segments at a single-cell level. In the tumor from patient 33, we analyzed two regions of chromosome 6 as controls. DNA FISH visualized two loci of the terminal segment of chromosome 6p in 100% of adenoma cells and one locus of the terminal segment of chromosome 6p in 50% of adenoma cells, consistent with the arm-level 6q deletion detected with array-CGH. Because this tumor was from an AIP mutation carrier, we further analyzed the 11q13 locus and identified loss of heterozygosity (LOH) at 11q13 in 58% of adenoma cells. Monosomy of 10q25.1 locus, affected by chromothripsis, was detected in 26% of cells. Four different cell populations coexist in this tumor, with either monosomy or disomy at one or both of the examined loci (Fig. 2C). This cellular heterogeneity was further confirmed when studying the 2p and 10q terminal regions. Six different cell populations were distinguished, with various combinations of CNAs ranging from nullosomy to trisomy (Fig. 2C). In the tumor from patient 42, in accordance with the array-CGH findings, DNA FISH showed two loci of the terminal segment of chromosome 2p in all adenoma cells. Chromothripsis resulted in loss of the 2q37.1 locus in 25% of adenoma cells. Loss of the 15q21.3 locus and of the terminal segment of chromosome arm 15q was detected in 45% and 80% of cells, respectively, whereas two loci of region 15q11.2, which was unaffected by chromothripsis, were detected in all adenoma cells. Regarding chromosome 11, DNA FISH revealed the coexistence of six different cell populations, with CNAs ranging from nullosomy to trisomy, in the 11p terminal segment, together with monosomy or disomy at the 11q25 locus, both of which were affected by chromothripsis (Fig. 3C). DNA copy numbers determined by DNA FISH were in agreement with the array-CGH log ratios, which reflect the sum of CNAs occurring at a concerned locus. Discussion Several recent studies using next-generation sequencing methods have failed to discover new tumor-causing gene mutations in GNAS mutation–negative pituitary adenomas (9–11, 19), thus pointing to a role of alternative oncogenic driver events, including cytogenetic alterations. Here, we describe the results of a combined cytogenetic search using array-CGH and DNA FISH in a large and well-characterized collection of GH-secreting pituitary adenomas from 43 patients with gigantism or acromegaly. Based on the copy-number profiles, we found that 63% had CNAs involving <12% of the genome, and the remaining 37% had disruption affecting 24% to 45% of the genome. None of the tumors presented disruptions affecting 12% to 24% of their genome. These findings are in agreement with Bi et al. (19) and Song et al. (10), who similarly identified two groups of pituitary adenomas (including all histotypes) based on the degree of genomic disruption. As previously reported (19, 27), arm-level CNAs were the most common abnormalities. Such chromosome arm-level aberrations usually result from missegregation during anaphase at the level of the centromere (28). This high rate of extensive chromosome alterations among genomically disrupted pituitary adenomas (which are invariably benign) exceeds that observed in several types of cancer (27). Gains and losses were equally prevalent in the low-CNA group, whereas gains predominated with the highly disrupted group. Unlike in malignancies (29), the frequency of arm-level CNAs was not related to the length of the chromosome arm. On the contrary, focal amplifications or deletions were rare, with the exception of 10p11.22 microduplication that was found in 20 adenomas, including four adenomas from the low-CNA group. This duplication encompassing CCDC7 gene was not associated with increased gene expression. The presence of this microduplication in almost half of the tumors suggests that it might be a benign, previously unreported copy-number variation. These two groups of tumors classified according to the degree of genome disruption did not significantly differ with respect to tumor size, invasiveness, proliferative index, or surgical outcome. However, a paradoxical response of GH to oral glucose load and a densely granulated histotype were more frequent in the high-CNA group, highlighting that particular phenotype of these tumors. In addition, all 13 tumors harboring GNAS mutations belonged to the low-CNA group. Thus, two distinct molecular subclasses could be distinguished on the basis of GNAS mutation status and the degree of genome disruption. The GNAS mutation–positive adenomas had few CNAs, highlighting the oncogenic potency of constitutive GNAS activation (3), which by itself drives tumor growth and GH hypersecretion. GNAS-mutation–negative adenomas with a high degree of genomic disruption have a specific molecular pathogenesis. Given the large number of extensive genome disruptions in these tumors, it remains to be established whether (some) alterations are driving tumorigenesis or are just nonfunctional byproducts that accumulate during tumor development (29). Amplification of chromosome arm 20q, which contains the GNAS locus, was observed in 12 adenomas. As previously reported, none of these tumors with 20q duplication harbored a GNAS mutation, suggesting that genomic gain in 20q could be an alternative mechanism of GNAS activation in mutation-negative GH-secreting pituitary adenomas (27). Analysis of somatic structural variants revealed two tumors with complex chromosomal rearrangements compatible with chromothripsis. A key feature of chromothripsis is the formation of ten to hundreds of locally clustered DNA rearrangements through a single, cataclysmic event. Chromosomes affected by chromothripsis show a characteristic pattern of copy-number oscillations, whereby typically only two or occasionally three copy-number states are detectable along the chromosome in the context of a large number of rearrangements (25, 30). Several mechanisms, including ionizing radiation, premature chromosome compaction, DNA replication stress, telomere shortening, aborting apoptosis, TP53 mutations, micronuclei collapse, and hyperploidy formation, have been implicated (30–32). Although chromothripsis has mainly been described in malignant tumors and linked to aggressive behavior and poor patient survival, similar massive DNA rearrangements have been reported in benign uterine leiomyomas (33). Chromothripsis-like patterns of CNAs were recently described in a GH-secreting adenoma harboring a GNAS mutation (9) and in a TSH-secreting adenoma (34). We observed massive DNA rearrangements resulting from chromothripsis in an adenoma from an AIP mutation–positive patient with acromegaly, localized on the long arm of chromosome 10, and in an adenoma from an AIP mutation–negative patient with gigantism, involving chromosome arms 2q, 11p, 11q, and 15q. The regions affected by chromothripsis contained several simultaneous tumorigenic DNA alterations, including oncogene amplification, loss of tumor suppressor genes, and deletion of genes involved in chromosome segregation and DNA crosslink repair. RNA sequencing and analysis of fusion transcripts did not identify gene fusions with oncogenic potential. Two fusion transcripts were predicted to be in frame and therefore potentially functional if translated, both of which had TCERG1L as a gene fusion partner. TCERG1L is a transcription elongation regulator 1–like protein reported to be downregulated by hypermethylation in colon primary tumors. This gene has not been reported to be partnered with any other genes (35). Because ratio plots indicated that chromothripsis was present in a mosaic state, we applied DNA FISH to these tumor samples to analyze chromosome segments at the single-cell level. We found good concordance between the percentage of subpopulations visualized on DNA FISH and the log-ratios of CNAs observed by CGH. Contrary to some cancer types in which rearrangement outcomes of chromothripsis are detectable in almost all cells (25, 31), we found rearrangements related to chromothripsis solely in tumor subclones of the two pituitary adenomas analyzed here. As also demonstrated in the case of a GNAS mutation–positive GH-secreting adenoma described by Valimaki et al. (9), chromothripsis-related alterations in pituitary tumors are very unlikely to be tumor-initiating events; rather, they confer a growth advantage. Genomic instability is an important source of diversified cell populations, resulting in intratumor heterogeneity (36). DNA copy numbers detected in our FISH experiments distinguished up to six distinct cell subpopulations, demonstrating extensive intratumor genetic diversity in GH-secreting pituitary tumors, that are considered monoclonal in origin (1, 2). This subclonal architecture of GH-secreting pituitary adenomas, previously reconstructed from sequencing data (9), provides certain insights into tumorigenesis. In the tumor from an AIP mutation–positive patient, LOH of 11q13 was detected in 58% of cells, rearrangements derived from chromothripsis on chromosome arm 10q were identified in 26% of cells, and the two alterations coexisted in 20% of cells. The precise dynamics of these chromosomal changes during tumor evolution remain to be determined. However, 11q13 LOH appears to be a late event, suggesting that biallelic AIP inactivation is not a prerequisite for AIP-related tumorigenesis. Because intratumor genetic heterogeneity results in phenotypic diversity, the subclonal architecture of pituitary tumors should be taken into account when studying the pathophysiology and dynamics of tumor evolution, tumor behaviors such as invasiveness and aggressiveness, and the variability of therapeutic responses (36, 37). Our study has some limitations. Cross-contamination from normal tissue of the adenoma samples cannot be completely excluded particularly in the microadenomas. However, 89% of adenomas included in this study were macroadenomas (Table 1), and tumor sampling was performed by highly experienced neurosurgeons, thus minimizing the risk of contamination. Furthermore, because densely granulated adenoma histotype was more prevalent in the high CNA group, better responses to somatostatin analogs would be expected in this group (38). However, responses to somatostatin analogs treatment could not be assessed due the small number of treated patients. In summary, molecular stratification of GH-secreting pituitary adenomas based on the degree of genome disruption can divide them into low- and high-CNA groups. Driving GNAS mutations are associated with lesser genomic disruption, whereas some GNAS mutation–negative tumors exhibit a high degree of predominantly arm-level chromosomal alterations. We identified chromothripsis-related chromosome CNAs in two tumors resulting in fusion transcripts, along with extensive intratumor DNA copy-number heterogeneity, thereby revealing the complex and heterogeneous clonal landscape of GH-secreting pituitary adenomas. Whether the molecular stratification based on CNAs represents a new prognostic tool remains to be determined Abbreviations: Abbreviations: array-CGH array comparative genomic hybridization CNA copy-number alteration FISH fluorescence in situ hybridization GH growth hormone LOH loss of heterozygosity OGTT oral glucose tolerance test Acknowledgments The authors thank Noémie Renault and Sylvain Guibert from GeCo, Integragen, for their technical support. Financial Support: This work was supported by grants from Inserm and Université Paris-Sud. M.H. was the recipient of a fellowship from Fondation pour la Recherche Médicale (FDM20150632800). P.K. was the recipient of a Contrat d’Interface Inserm. Disclosure Summary: The authors have nothing to disclose. References 1. Herman V, Fagin J, Gonsky R, Kovacs K, Melmed S. Clonal origin of pituitary adenomas. J Clin Endocrinol Metab . 1990; 71( 6): 1427– 1433. Google Scholar CrossRef Search ADS PubMed  2. Melmed S. Acromegaly pathogenesis and treatment. J Clin Invest . 2009; 119( 11): 3189– 3202. Google Scholar CrossRef Search ADS PubMed  3. Vallar L, Spada A, Giannattasio G. Altered Gs and adenylate cyclase activity in human GH-secreting pituitary adenomas. Nature . 1987; 330( 6148): 566– 568. Google Scholar CrossRef Search ADS PubMed  4. Lecoq AL, Kamenický P, Guiochon-Mantel A, Chanson P. Genetic mutations in sporadic pituitary adenomas--what to screen for? Nat Rev Endocrinol . 2015; 11( 1): 43– 54. Google Scholar CrossRef Search ADS PubMed  5. Vergès B, Boureille F, Goudet P, Murat A, Beckers A, Sassolas G, Cougard P, Chambe B, Montvernay C, Calender A. Pituitary disease in MEN type 1 (MEN1): data from the France-Belgium MEN1 multicenter study. J Clin Endocrinol Metab . 2002; 87( 2): 457– 465. Google Scholar CrossRef Search ADS PubMed  6. 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. Positional cloning of the gene for multiple endocrine neoplasia-type 1. Science . 1997; 276( 5311): 404– 407. Google Scholar CrossRef Search ADS PubMed  7. Vierimaa O, Georgitsi M, Lehtonen R, Vahteristo P, Kokko A, Raitila A, Tuppurainen K, Ebeling TM, Salmela PI, Paschke R, Gündogdu S, De Menis E, Mäkinen MJ, Launonen V, Karhu A, Aaltonen LA. Pituitary adenoma predisposition caused by germline mutations in the AIP gene. Science . 2006; 312( 5777): 1228– 1230. Google Scholar CrossRef Search ADS PubMed  8. Bertherat J, Horvath A, Groussin L, Grabar S, Boikos S, Cazabat L, Libe R, René-Corail F, Stergiopoulos S, Bourdeau I, Bei T, Clauser E, Calender A, Kirschner LS, Bertagna X, Carney JA, Stratakis CA. Mutations in regulatory subunit type 1A of cyclic adenosine 5′-monophosphate-dependent protein kinase (PRKAR1A): phenotype analysis in 353 patients and 80 different genotypes. J Clin Endocrinol Metab . 2009; 94( 6): 2085– 2091. Google Scholar CrossRef Search ADS PubMed  9. Välimäki N, Demir H, Pitkänen E, Kaasinen E, Karppinen A, Kivipelto L, Schalin-Jäntti C, Aaltonen LA, Karhu A. Whole-genome sequencing of growth hormone (GH)-secreting pituitary adenomas. J Clin Endocrinol Metab . 2015; 100( 10): 3918– 3927. Google Scholar CrossRef Search ADS PubMed  10. Song ZJ, Reitman ZJ, Ma ZY, Chen JH, Zhang QL, Shou XF, Huang CX, Wang YF, Li SQ, Mao Y, Zhou LF, Lian BF, Yan H, Shi YY, Zhao Y. The genome-wide mutational landscape of pituitary adenomas. Cell Res . 2016; 26( 11): 1255– 1259. Google Scholar CrossRef Search ADS PubMed  11. Ronchi CL, Peverelli E, Herterich S, Weigand I, Mantovani G, Schwarzmayr T, Sbiera S, Allolio B, Honegger J, Appenzeller S, Lania AG, Reincke M, Calebiro D, Spada A, Buchfelder M, Flitsch J, Strom TM, Fassnacht M. Landscape of somatic mutations in sporadic GH-secreting pituitary adenomas. Eur J Endocrinol . 2016; 174( 3): 363– 372. Google Scholar CrossRef Search ADS PubMed  12. Yacqub-Usman K, Richardson A, Duong CV, Clayton RN, Farrell WE. The pituitary tumour epigenome: aberrations and prospects for targeted therapy. Nat Rev Endocrinol . 2012; 8( 8): 486– 494. Google Scholar CrossRef Search ADS PubMed  13. Trivellin G, Daly AF, Faucz FR, Yuan B, Rostomyan L, Larco DO, Schernthaner-Reiter MH, Szarek E, Leal LF, Caberg JH, Castermans E, Villa C, Dimopoulos A, Chittiboina P, Xekouki P, Shah N, Metzger D, Lysy PA, Ferrante E, Strebkova N, Mazerkina N, Zatelli MC, Lodish M, Horvath A, de Alexandre RB, Manning AD, Levy I, Keil MF, Sierra ML, Palmeira L, Coppieters W, Georges M, Naves LA, Jamar M, Bours V, Wu TJ, Choong CS, Bertherat J, Chanson P, Kamenický P, Farrell WE, Barlier A, Quezado M, Bjelobaba I, Stojilkovic SS, Wess J, Costanzi S, Liu P, Lupski JR, Beckers A, Stratakis CA. Gigantism and acromegaly due to Xq26 microduplications and GPR101 mutation. N Engl J Med . 2014; 371( 25): 2363– 2374. Google Scholar CrossRef Search ADS PubMed  14. Iacovazzo D, Caswell R, Bunce B, Jose S, Yuan B, Hernández-Ramírez LC, Kapur S, Caimari F, Evanson J, Ferraù F, Dang MN, Gabrovska P, Larkin SJ, Ansorge O, Rodd C, Vance ML, Ramírez-Renteria C, Mercado M, Goldstone AP, Buchfelder M, Burren CP, Gurlek A, Dutta P, Choong CS, Cheetham T, Trivellin G, Stratakis CA, Lopes MB, Grossman AB, Trouillas J, Lupski JR, Ellard S, Sampson JR, Roncaroli F, Korbonits M. Germline or somatic GPR101 duplication leads to X-linked acrogigantism: a clinico-pathological and genetic study. Acta Neuropathol Commun . 2016; 4( 1): 56. Google Scholar CrossRef Search ADS PubMed  15. Hui AB, Pang JC, Ko CW, Ng HK. Detection of chromosomal imbalances in growth hormone-secreting pituitary tumors by comparative genomic hybridization. Hum Pathol . 1999; 30( 9): 1019– 1023. Google Scholar CrossRef Search ADS PubMed  16. Szymas J, Schluens K, Liebert W, Petersen I. Genomic instability in pituitary adenomas. Pituitary . 2002; 5( 4): 211– 219. Google Scholar CrossRef Search ADS PubMed  17. Harada K, Nishizaki T, Ozaki S, Kubota H, Harada K, Okamura T, Ito H, Sasaki K. Cytogenetic alterations in pituitary adenomas detected by comparative genomic hybridization. Cancer Genet Cytogenet . 1999; 112( 1): 38– 41. Google Scholar CrossRef Search ADS PubMed  18. Pack SD, Qin LX, Pak E, Wang Y, Ault DO, Mannan P, Jaikumar S, Stratakis CA, Oldfield EH, Zhuang Z, Weil RJ. Common genetic changes in hereditary and sporadic pituitary adenomas detected by comparative genomic hybridization. Genes Chromosomes Cancer . 2005; 43( 1): 72– 82. Google Scholar CrossRef Search ADS PubMed  19. Bi WL, Horowitz P, Greenwald NF, Abedalthagafi M, Agarwalla PK, Gibson WJ, Mei Y, Schumacher SE, Ben-David U, Chevalier A, Carter S, Tiao G, Brastianos PK, Ligon AH, Ducar M, MacConaill L, Laws ER Jr., Santagata S, Beroukhim R, Dunn IF. Landscape of genomic alterations in pituitary adenomas. Clin Cancer Res . 2017; 23: 1841– 1851 Google Scholar CrossRef Search ADS PubMed  20. Giustina A, Chanson P, Kleinberg D, Bronstein MD, Clemmons DR, Klibanski A, van der Lely AJ, Strasburger CJ, Lamberts SW, Ho KK, Casanueva FF, Melmed S; Acromegaly Consensus Group. Expert consensus document: a consensus on the medical treatment of acromegaly. Nat Rev Endocrinol . 2014; 10( 4): 243– 248. Google Scholar CrossRef Search ADS PubMed  21. Maniatis T, Fritsch EF, Sambrook J. Molecular cloning: a laboratory manual . New York, NY: Cold Spring Harbor Laboratory; 1982. 22. Boissot N, Thomas S, Chovelon V, Lecoq H. NBS-LRR-mediated resistance triggered by aphids: viruses do not adapt; aphids adapt via different mechanisms. BMC Plant Biol . 2016; 16( 1): 25. Google Scholar CrossRef Search ADS PubMed  23. Kim D, Pertea G, Trapnell C, Pimentel H, Kelley R, Salzberg SL. TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome Biol . 2013; 14( 4): R36. Google Scholar CrossRef Search ADS PubMed  24. Lecoq AL, Bouligand J, Hage M, Cazabat L, Salenave S, Linglart A, Young J, Guiochon-Mantel A, Chanson P, Kamenický P. Very low frequency of germline GPR101 genetic variation and no biallelic defects with AIP in a large cohort of patients with sporadic pituitary adenomas. Eur J Endocrinol . 2016; 174( 4): 523– 530. Google Scholar CrossRef Search ADS PubMed  25. Stephens PJ, Greenman CD, Fu B, Yang F, Bignell GR, Mudie LJ, Pleasance ED, Lau KW, Beare D, Stebbings LA, McLaren S, Lin ML, McBride DJ, Varela I, Nik-Zainal S, Leroy C, Jia M, Menzies A, Butler AP, Teague JW, Quail MA, Burton J, Swerdlow H, Carter NP, Morsberger LA, Iacobuzio-Donahue C, Follows GA, Green AR, Flanagan AM, Stratton MR, Futreal PA, Campbell PJ. Massive genomic rearrangement acquired in a single catastrophic event during cancer development. Cell . 2011; 144( 1): 27– 40. Google Scholar CrossRef Search ADS PubMed  26. Reincke M, Sbiera S, Hayakawa A, Theodoropoulou M, Osswald A, Beuschlein F, Meitinger T, Mizuno-Yamasaki E, Kawaguchi K, Saeki Y, Tanaka K, Wieland T, Graf E, Saeger W, Ronchi CL, Allolio B, Buchfelder M, Strom TM, Fassnacht M, Komada M. Mutations in the deubiquitinase gene USP8 cause Cushing’s disease. Nat Genet . 2015; 47( 1): 31– 38. Google Scholar CrossRef Search ADS PubMed  27. Bi WL, Greenwald NF, Ramkissoon SH, Abedalthagafi M, Coy SM, Ligon KL, Mei Y, MacConaill L, Ducar M, Min L, Santagata S, Kaiser UB, Beroukhim R, Laws ER Jr, Dunn IF. Clinical Identification of oncogenic drivers and copy-number alterations in pituitary tumors. Endocrinology . 2017; 158( 7): 2284– 2291. Google Scholar CrossRef Search ADS PubMed  28. Ganem NJ, Godinho SA, Pellman D. A mechanism linking extra centrosomes to chromosomal instability. Nature . 2009; 460( 7252): 278– 282. Google Scholar CrossRef Search ADS PubMed  29. Beroukhim R, Mermel CH, Porter D, Wei G, Raychaudhuri S, Donovan J, Barretina J, Boehm JS, Dobson J, Urashima M, Mc Henry KT, Pinchback RM, Ligon AH, Cho YJ, Haery L, Greulich H, Reich M, Winckler W, Lawrence MS, Weir BA, Tanaka KE, Chiang DY, Bass AJ, Loo A, Hoffman C, Prensner J, Liefeld T, Gao Q, Yecies D, Signoretti S, Maher E, Kaye FJ, Sasaki H, Tepper JE, Fletcher JA, Tabernero J, Baselga J, Tsao MS, Demichelis F, Rubin MA, Janne PA, Daly MJ, Nucera C, Levine RL, Ebert BL, Gabriel S, Rustgi AK, Antonescu CR, Ladanyi M, Letai A, Garraway LA, Loda M, Beer DG, True LD, Okamoto A, Pomeroy SL, Singer S, Golub TR, Lander ES, Getz G, Sellers WR, Meyerson M. The landscape of somatic copy-number alteration across human cancers. Nature . 2010; 463( 7283): 899– 905. Google Scholar CrossRef Search ADS PubMed  30. Korbel JO, Campbell PJ. Criteria for inference of chromothripsis in cancer genomes. Cell . 2013; 152( 6): 1226– 1236. Google Scholar CrossRef Search ADS PubMed  31. Rausch T, Jones DT, Zapatka M, Stütz AM, Zichner T, Weischenfeldt J, Jäger N, Remke M, Shih D, Northcott PA, Pfaff E, Tica J, Wang Q, Massimi L, Witt H, Bender S, Pleier S, Cin H, Hawkins C, Beck C, von Deimling A, Hans V, Brors B, Eils R, Scheurlen W, Blake J, Benes V, Kulozik AE, Witt O, Martin D, Zhang C, Porat R, Merino DM, Wasserman J, Jabado N, Fontebasso A, Bullinger L, Rücker FG, Döhner K, Döhner H, Koster J, Molenaar JJ, Versteeg R, Kool M, Tabori U, Malkin D, Korshunov A, Taylor MD, Lichter P, Pfister SM, Korbel JO. Genome sequencing of pediatric medulloblastoma links catastrophic DNA rearrangements with TP53 mutations. Cell . 2012; 148( 1-2): 59– 71. Google Scholar CrossRef Search ADS PubMed  32. Zhang CZ, Spektor A, Cornils H, Francis JM, Jackson EK, Liu S, Meyerson M, Pellman D. Chromothripsis from DNA damage in micronuclei. Nature . 2015; 522( 7555): 179– 184. Google Scholar CrossRef Search ADS PubMed  33. Mehine M, Kaasinen E, Mäkinen N, Katainen R, Kämpjärvi K, Pitkänen E, Heinonen HR, Bützow R, Kilpivaara O, Kuosmanen A, Ristolainen H, Gentile M, Sjöberg J, Vahteristo P, Aaltonen LA. Characterization of uterine leiomyomas by whole-genome sequencing. N Engl J Med . 2013; 369( 1): 43– 53. Google Scholar CrossRef Search ADS PubMed  34. Sapkota S, Horiguchi K, Tosaka M, Yamada S, Yamada M. Whole-exome sequencing study of thyrotropin-secreting pituitary adenomas. J Clin Endocrinol Metab . 2017; 102( 2): 566– 575. Google Scholar PubMed  35. Yi JM, Dhir M, Guzzetta AA, Iacobuzio-Donahue CA, Heo K, Yang KM, Suzuki H, Toyota M, Kim HM, Ahuja N. DNA methylation biomarker candidates for early detection of colon cancer. Tumour Biol . 2012; 33( 2): 363– 372. Google Scholar CrossRef Search ADS PubMed  36. Burrell RA, McGranahan N, Bartek J, Swanton C. The causes and consequences of genetic heterogeneity in cancer evolution. Nature . 2013; 501( 7467): 338– 345. Google Scholar CrossRef Search ADS PubMed  37. Jamal-Hanjani M, Quezada SA, Larkin J, Swanton C. Translational implications of tumor heterogeneity. Clin Cancer Res . 2015; 21: 1258– 1266. Google Scholar CrossRef Search ADS PubMed  38. Larkin S, Reddy R, Karavitaki N, Cudlip S, Wass J, Ansorge O. Granulation pattern, but not GSP or GHR mutation, is associated with clinical characteristics in somatostatin-naive patients with somatotroph adenomas. Eur J Endocrinol . 2013; 168( 4): 491– 499. Google Scholar CrossRef Search ADS PubMed  Copyright © 2018 Endocrine Society

Journal

Journal of Clinical Endocrinology and MetabolismOxford University Press

Published: Feb 21, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off