Introduction

Pituitary adenomas are common primary intracranial tumors, and while most are benign, these tumors may still cause notable morbidity for patients [1, 2]. As with other neuroendocrine tumors, pituitary adenomas are classified as either “functional” or “nonfunctional”. Functional tumors secrete excess levels of specific hormones, for example growth hormone (GH), adrenocorticotropic hormone (ACTH), prolactin (PRL), and follicle stimulating hormone/luteinizing hormone (FSH/LH), or thyroid stimulating hormone (TSH). In some tumor subsets, continual overproduction of any of these hormones can result in debilitating multisystem dysfunction and ultimately an increased risk of death [3]. In contrast, nonfunctional adenomas do not secrete abnormally excessive hormones. However, these tumors often compress adjacent neurovascular structures, leading to mass effect and visual disturbances secondary to optic chiasm compression. On the genomic level, large cohorts of pituitary adenomas have recently been comprehensively profiled. These studies have revealed, in addition to limited somatic mutations, that a subset of pituitary adenomas are characterized by the presence of somatic copy-number alterations affecting whole chromosome arms that comprise a large fraction (~50%) of the genome [4,5,6].

In most human cancers, neoplastic cells achieve the ability to proliferate unlimitedly by maintaining their telomeres. Telomeres, the repetitive hexanucleotide sequences (TTAGGGn) located at the terminal ends of eukaryotic chromosomes, are pivotal for genome integrity. Critical telomere shortening is a common abnormality observed early in tumorigenesis, where it promotes malignant transformation and tumor progression via telomere destabilization and concomitant chromosomal instability [7]. Consequently, alterations in cancer cell telomere lengths have been identified as prognostic factors in a variety of cancer subtypes [8, 9]. Cancer cells require activation of a telomere maintenance mechanism, predominantly through activating the enzyme telomerase. In contrast, a subset of cancers, including a substantial proportion of pancreatic neuroendocrine tumors (PanNETs) and gliomas, utilize a telomerase-independent mechanism which is mediated by homology directed repair, termed alternative lengthening of telomeres (ALT) [10].

ALT-positive cancers often have lost function, through mutation or deletion, in the alpha thalassemia/mental retardation syndrome X-linked (ATRX) or death domain-associated protein (DAXX) genes. ATRX and DAXX are chromatin modifying proteins, and in concert, this ATRX/DAXX complex deposits histone variant H3.3 in heterochromatic regions of chromosomes containing highly repetitive sequence elements, such as retrotransposons, pericentromeric regions, and telomeres. These repetitive sequences are inherently unstable and are prone to increased replication errors and aberrant homologous recombination. As a result, disruption of ATRX/DAXX chromatin remodeling function at telomeres presumably allows for the development of ALT. In archived tissue specimens, ALT may be identified by the presence of unique ALT-associated features, including ultra-bright telomeric foci and dramatic cell-to-cell telomere heterogeneity. Using this approach, ALT and ATRX mutations have been associated with aggressive clinical behaviors in a variety of tumor types.

In the current study, we evaluated a discovery cohort of 106 pituitary adenomas (88 primary, 18 recurrent tumors), as well as an additional cohort of recurrent tumors (32 tumors from 22 patients). Using these cohorts, we evaluated the presence of telomere alterations, including the presence of ALT, as well as alterations in ATRX and DAXX protein expression and correlated these findings with clinicopathologic factors.

Materials and methods

Case selection and tissue microarray construction

This study was reviewed and approved by the human subjects Institutional Review Boards of the Dana-Farber/Brigham and Women’s Cancer Center, the Brigham and Women’s Hospital, Johns Hopkins Hospital, and the Broad Institute. Written informed consent or a waiver of consent was obtained for all participants. In the discovery cohort, histopathologic diagnosis based on WHO criteria, and tumor purity >80% was confirmed in all samples selected for study by two board-certified neuropathologists (S.S. and S.C.). Tumors were classified by their immunohistochemical expression of hormones (null, PRL, ACTH, GH, FSH/LH, and TSH) as well as by the associated serum levels of the corresponding hormone. Corticotroph, somatotroph (including plurihormonal cases), and lactotroph adenomas were considered functional in our analyses, whereas gonadotroph and null-cell adenomas were considered nonfunctional. We classified tumors as nonfunctional in cases where there was immunohistologic staining for one or more hormones, but serum levels remained normal and there were no clinical signs of endocrinopathy. In the validation cohort of recurrent disease, we included cases of pituitary adenoma that required at least more than one surgery. Classification was performed similarly as above, based on immunohistochemical evaluation of at least GH, ACTH and PRL, and serum levels of the corresponding hormone.

Telomere-specific FISH

Telomere-specific FISH was performed as previously outlined [11, 12]. In brief, tissue slides were deparaffinized, hydrated, and steamed for 25 min in citrate buffer (Vector Laboratories). This was followed by dehydration and hybridization with a Cy3-labeled peptide nucleic acid (PNA) probe complementary to the mammalian telomere repeat sequence [(N-terminus to C-terminus) CCCTAACCCTAACCCTAA]. An Alexa Fluor-488–labeled PNA probe specific to human centromeric DNA repeats was also included as a control to assess the validity of the hybridization. Following post-hybridization washes, the slides were counterstained with DAPI following post-hybridization washes.

Microscopy

Slides were imaged with a Nikon 50i epifluorescence microscope equipped with X-Cite series 120 illuminator (EXFO Photonics Solutions, Ontario, CA) using a ×40/×0.95 NA PlanApo lens with correction collar. For each color channel, separate grayscale images were captured using Nikon NIS-Elements software (NIS-Elements BR 3.2 64-bit) and an attached Photometrics CoolsnapEZ digital cooled charged coupled device camera and saved as 12-bit uncompressed TIFF files. Exposure times were set to avoid fluorescence signal saturation. For telomere length analysis, integration times were 500  ms for Cy3 (telomere) and 100 ms for the DAPI nuclear counterstain. In contrast, for ALT analysis, integration times were 500 ms for Cy3 (telomere) and 100 ms for the DAPI nuclear counterstain.

ALT and telomere length assessment

ALT status was interpreted using previously published criteria and was characterized by the presence of distinct large telomeric FISH DNA signals [10, 13, 14]. In the ALT-negative cases, telomere lengths were qualitatively scored by direct visual assessment of the stained slides, comparing the intensity of telomere signals from cancer cells to the intensity of telomere signals of entrapped non-neoplastic cells. Thus, telomere lengths were evaluated as being either normal, or abnormally short or long.

Immunohistochemistry

Immunohistochemical studies were systematically performed using an ATRX antibody (rabbit polyclonal, 1:200 dilution, catalog# HPA001906 Sigma-Aldrich). Immunostaining was performed on automated instruments (BenchMark, Ventana Medical Systems, Tucson, AZ, USA). The immunohistochemical protocol included deparaffinization, hydration, antigen retrieval, primary antibody incubation, and detection and visualization as per manufacturer’s instructions. Immunohistochemistry for DAXX (rabbit polyclonal, 1:100 dilution, catalog# HPA008736, Atlas Antibodies) was performed manually. Sections were incubated with primary antibody for 2 h at room temperature followed by secondary antibody (Leica Microsystems) for 30 min and detected with 3,30-diaminobenzidine (Sigma-Aldrich) after 10 min.

Next generation sequencing

To identify possible somatic genetic alterations associated with ALT, next generation sequencing was performed on the three ALT-positive cases. While one of the cases was previously assessed by whole exome sequencing [4], the other cases were assessed using Oncopanel, an established hybrid-capture and massively parallel sequencing assay [15].

Statistical analysis

Categories of telomere length (short, normal, and long) were correlated with the clinicopathological characteristics. Variables were described using proportions, ranges, means, medians, and standard deviations as appropriate. Proportions were compared using Chi-Square or Fisher’s exact tests as appropriate. Statistical analyses were performed using GraphPad Prism version 8.0 (San Diego, CA).

Results

Clinicopathological characteristics

The clinicopathological characteristics for the entire study population of the discovery set (n = 106) are shown in Table 1 and some details have been previously published [4]. Briefly, the mean age of patients was 53.4 years (range, 16–86) and 61.3% were male. In all, 88 tumors were newly diagnosed and 18 were recurrent tumors. The majority of the tumors evaluated were nonfunctional (78.3%); while 21.7% were functional, including 15 HGH expressing tumors (somatotroph adenomas) associated with acromegaly, 2 HGH and PRL expressing tumors (somatomammotroph adenomas) associated with acromegaly, 3 PRL expressing tumors (prolactinomas) associated with amenorrhea, galactorrhea, or low testosterone, 1 ACTH expressing tumor (corticotroph adenoma) associated with Cushing’s syndrome, and 1 adenoma with no immunohistochemically detectable hormone expression associated with Cushing’s syndrome. Finally, while most tumors did not display the previously defined disrupted genotype (61.3%), 21.7% did present with this disrupted genotype.

Table 1 Clinicopathologic and molecular features of pituitary adenomas.
Full size table

In situ telomere analysis in pituitary adenomas

Telomere lengths were qualitatively scored through direct visual assessment by comparing the intensity of telomere signals in the cancer cells to the intensity of telomere signals of entrapped non-neoplastic cells within the same case. Based on the cancer cell telomere length, the cases were grouped into either short (59.4% cases; n = 64), normal (36.8% cases; n = 39), or long (0% cases; n = 0) categories. In Fig. 1, representative pituitary adenomas demonstrating cancer cells with short telomeres and normal telomeres are shown. The cases with short telomeres did not significantly differ from the cases with normal telomeres in any of the clinicopathological characteristics, including gender, clinical endocrinologic status, recurrent status, or displaying the disrupted genotype.

Fig. 1: Telomere length analysis by FISH in pituitary adenomas.
figure 1

Two representative examples of cases showing either short or normal telomere lengths in tumor cells are shown. a This case shows strikingly diminished telomere signals in tumor cells (asterisks) as compared to the entrapped non-neoplastic cells (arrows). b This case displays comparable telomere intensities in tumor cells (asterisks) with those observed in the entrapped non-neoplastic cells (arrows). In both images, the DNA is stained with DAPI (blue) and telomere DNA is stained with the Cy3-labeled telomere-specific peptide nucleic acid probe (red). It is noteworthy that the centromere DNA, stained with the FITC-labeled centromere-specific peptide nucleic acid probe, has been omitted from the image to emphasize the differences in telomere lengths. Original magnification, ×400.

Full size image

Characterization of ALT-positive pituitary adenomas

ALT was identified in 3 of 106 (2.8%) pituitary tumors (cases 15, 95, and 111). Two were male and one was female, with ages ranging from 39 to 73. Two of these cases were nonfunctional, whereas one case was HGH expressing (densely or sparsely granulated, associated with acromegaly). Two of the three cases (66%) were recurrent, as compared to only 16 of 103 (15.5%) of the ALT-negative pituitary adenomas (chi-square test; p = 0.01). As shown in Fig. 2, ultra-bright telomeric foci indicative of ALT were present in all three cases. ATRX loss was observed in case 95 and DAXX loss observed in case 15. In contrast, case 111 retained nuclear protein positivity for both ATRX and DAXX. As expected, ATRX and DAXX protein expression was retained in all of the ALT-negative pituitary adenomas, although four cases did display partial loss of ATRX and one case displayed partial loss of DAXX. Finally, we evaluated the somatic mutational spectrum in these three ALT-positive cases. Case 15 was previously assessed by whole exome sequencing [4] and 117 mutations were identified, including in ARID1B, ATP2B2, CHD4, PI3KR1, and RUNX1. The other two ALT-positive cases were assessed using an established hybrid-capture and massively parallel sequencing assay (OncoPanel) [15]. Case 95 had mutations in AR, CHEK2, CYLD, FGFR3, KAT6A, and TRAF7; whereas, case 111 only had one somatic mutation (EP300). Interestingly, mutations in ATRX or DAXX were not identified in any of the three cases.

Fig. 2: ALT-positive pituitary adenomas and ATRX and DAXX immunostaining.
figure 2

Representative regions of the three cases (15, 95, and 111) are shown. a, d, g Representative images of ALT-positive cases displaying ultrabright telomeric FISH signals as indicative of ALT. The DNA is stained with DAPI (blue) and telomere DNA is stained with the Cy3-labeled telomere-specific peptide nucleic acid probe (red). b, e, h Case 95 shows tumor cell-specific loss of nuclear ATRX expression, whereas cases 15 and 111 display intact ATRX nuclear expression. (Original magnification ×400). c, f, i Case 15 shows tumor cell-specific loss of nuclear DAXX expression, whereas cases 95 and 111 display intact DAXX nuclear expression. For all images presented—original magnification, ×400.

Full size image

Assessment of additional recurrent pituitary adenomas

Because we identified a small number of ALT-positive pituitary adenomas that appeared to be enriched in recurrent disease, we next assessed tumors from a different cohort of patients that had all developed recurrent disease. In total, we assessed 32 recurrent pituitary adenomas from 22 patients. Of the 22 patients, ALT was identified in 8 (36%) patients using the telomere-specific FISH assay. Of the ten patients for whom tumor tissue was available to be studied from two different surgeries, we found that both tumors were ALT-negative in seven patients and both tumors were ALT-positive in two patients. We did identify one case in which the original pituitary adenoma was ALT-negative, but 2 years later, the subsequent recurrent adenoma was ALT-positive with partial ATRX loss, thereby suggesting that ALT activation can occur upon progression.

Discussion

Pituitary adenomas represent a heterogeneous group of neoplasms that derive and/or share immunophenotypic and functional properties with endocrine cells of the adenohypophysis. The WHO Classification of the Tumours of the Endocrine Organs recognizes seven major categories of pituitary adenomas based on endocrine cell identity and several rare subtypes [16]. Recent genomic studies have emerged in the past years characterizing alterations associated with several adenoma subtypes. For example, activating mutations involving GNAS leading to cAMP/protein kinase A pathway activation is a recurrent event in somatotroph adenomas [17, 18], and somatic mutations in the deubiquitinase gene USP8, resulting in increased EGF pathway signaling are present in a subset (~40–60%) of functioning corticotrophic adenomas [19, 20]. In this study, we identified that the majority of pituitary adenomas have shortened telomere lengths. In addition, we identify a subset of pituitary adenomas that display the ALT phenotype and are enriched in recurrent disease.

While 10–15% of all cancers are ALT-positive, the prevalence of ALT is enriched in multiple tumor types, including PanNETs, gliomas, neuroblastomas, and sarcomas [10, 13]. The vast majority of ALT-positive cell lines and cancers have lost functional ATRX [21]. However, DAXX alterations have been identified in cell lines [22, 23] and some cancer types [24], and DAXX alterations are twice as prevalent as ATRX alterations in ALT-positive PanNETs [11, 25, 26]. Across multiple studies of primary well-differentiated PanNETs, loss of function of the ATRX/DAXX chromatin remodeling complex, mainly through somatic mutations, and acquisition of ALT are associated with decreased recurrence-free survival [11, 25,26,27,28,29] and a molecular subtype that resembles islet α-cells at the epigenetic and transcriptomic levels [30, 31]. Additional neuroendocrine tumor types that have ATRX mutations in a subset of cases are pheochromocytomas and paragangliomas [32,33,34]. Similar to the PanNETs, pheochromocytomas and paragangliomas with somatic ATRX mutations are strongly associated with ALT and an aggressive clinical behavior. However, in the ALT-positive cases identified in this current study in which sequencing was performed, mutations in ATRX or DAXX were not found. Although, two cases displayed ATRX or DAXX protein loss, suggesting either homozygous loss of the gene or the presence of another silencing mechanism (e.g. promoter methylation). In contrast, case 111 was wild-type for ATRX and DAXX at the gene and protein levels and only one somatic mutation in EP300, a histone acetyltransferase that regulates transcription via chromatin remodeling, was found [35]. While EP300 has not been previously linked to telomere maintenance, these results suggest the possibility of an additional driver of ALT, similar to recent studies that have identified SMARCAL1 mutations in a subset of ALT-positive glioblastomas [36] and SLX4IP mutations in a subset of ALT-positive osteosarcomas [37].

In studies of pituitary adenomas, a published cohort of 42 pituitary adenomas from pediatric and adolescent patients revealed that three cases displayed loss of nuclear ATRX protein expression, although the vast majority of adenomas retained ATRX [38]. Similarly, ATRX and DAXX protein expression were retained in a large series of adenohypophyseal endocrine tumors of different hormonal and clinical types from 246 patients; however, one (or two) corticotroph carcinomas examined displayed ATRX loss [39]. Although only in a single pituitary carcinoma, a rare tumor that is pathologically indistinguishable from adenoma but strictly defined by the presence of leptomeningeal or extracranial metastases, Guo et al. documented the presence of a somatic mutation in ATRX, along with mutations in PTEN and TP53 [40].

There are a number of strengths of our current study. The discovery cohort consisted of a large number of pituitary adenomas with well-annotated clinical and pathological data. In addition, analysis of a second cohort that we collected to characterize recurrent pituitary adenomas in more detail confirmed the presence of ALT in a substantial fraction of recurrent disease. Also, we assessed two different cancer-specific telomere alterations, telomere shortening and presence of ALT, using a robust telomere-specific FISH assay. However, despite these strengths, there are also limitations to our study. This is a retrospective study with imperfect clinical follow-up. In addition, there is a relatively small sample size with limited number of ALT-positive cases and a limited number of cases with available tumor from primary and recurrent disease resections. The clinical significance of the subset of ALT-positive pituitary adenomas remains to be determined.

In summary, short telomeres are prevalent in pituitary adenomas. In addition, a substantial subset of pituitary adenomas, enriched in recurrent tumors, are ALT-positive. Future studies are necessary to validate and extend these findings.