Abstract Parathyroid carcinoma is a rare but life-threatening malignancy. Limited treatment options are available for the more than 50% of parathyroid carcinoma patients who develop recurrent disease following surgery with curative intent. Many such patients will become refractory to medical management, eventually succumbing to the sustained, progressive hypercalcemia caused by parathyroid hormone-secreting tumors. Identification of genetic/molecular aberrations that might serve as “actionable targets” for pharmacologic intervention in parathyroid carcinoma have uncovered frequent inactivation of the CDC73 tumor suppressor gene, amplification of the cyclin D1 oncogene and activating mutations of the PI3K/MTOR pathway. We sought to combine these frequent tumor driving mutations to develop a novel preclinical model that could mimic the clinical course of recurrent/metastatic parathyroid carcinoma, for study of disease progression and to eventually develop/test novel preventative/therapeutic interventions. To this end, we crossed existing mouse models with genetic alterations known to occur in human parathyroid carcinomas, namely PTH-cyclinD1 transgenic mice (PCD), oncogenic Pik3ca mutant mice (Pik3caH1047R) and parathyroid-specific Cdc73 knockout mice. PCD mice develop biochemical hyperparathyroidism and parathyroid tumors with known kinetics. Mice harboring double and triple genetic modifications were evaluated independently and compared to PCD and wild type littermate controls. Assessment of progression of biochemical hyperparathyroidism was evaluated by monthly blood collection followed by measurement of serum calcium and parathyroid hormone (PTH). We observed that activation of Pik3ca in the absence of cyclin D1 overexpression was insufficient to drive parathyroid tumorigenesis and that loss of Cdc73 did not appear to enhance parathyroid tumorigenesis beyond the contribution of cyclin D1 overexpression. However, crossed mice harboring activating mutations of both cyclin D1 and Pik3ca developed more severe biochemical primary hyperparathyroidism than those with activated cyclin D1 alone. Some PCDxPik3caH1047R mice developed atypical histologic features consistent with aggressive parathyroid tumors. Thus, the combination of activated Pik3ca and cyclin D1 overexpression appears to yield a more aggressive parathyroid tumor phenotype. Our findings carry important potential therapeutic implications: the combination of cdk4/6 inhibitors with PI3K/MTOR inhibitors is under active investigation in other tumor types and the efficacy of such combination therapy merits investigation in parathyroid carcinomas which may harbor both cyclin D1 amplifications and activating mutations in PI3K/MTOR. Citation Format: Maia Jakubowski, Callie Burke, Justin Bellizzi, Andrew Arnold, Jessica Costa-Guda. Oncogenic PIK3CA mutation enhances cyclin D1-driven parathyroid tumorigenesis in transgenic mice [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 5016.

Evocalcet is an allosteric modulator of the calcium-sensing receptor (CaSR) that effectively suppresses parathyroid hormone (PTH) secretion in both primary and secondary hyperparathyroidism. However, its effects on the PTH-calcium setpoint and parathyroid cell proliferation remain unclear. We investigated these effects using the PC mouse model of primary hyperparathyroidism, which is characterized by parathyroid-targeted cyclin D1 overexpression . Evocalcet was administered orally at a dose of 0.025 mg/g diet. The PTH-calcium setpoint was evaluated, and the antiproliferative effect of evocalcet on parathyroid cells was assessed using 5-bromo-2′-deoxyuridine (BrdU) incorporation assays. The effects of evocalcet were compared with those of cinacalcet. Expression levels of the vitamin D receptor (VDR) and CaSR in parathyroid glands were also examined. Evocalcet significantly reduced the PTH-calcium setpoint in PC mice, restoring it to levels comparable to those observed in wild-type controls. Evocalcet treatment markedly decreased the proportion of BrdU-positive parathyroid cells, indicating suppression of parathyroid cell proliferation. This antiproliferative effect was comparable to that observed with cinacalcet. Neither evocalcet nor cinacalcet altered VDR or CaSR expression in the parathyroid glands. Evocalcet, similar to cinacalcet, lowers the PTH-calcium setpoint and inhibits parathyroid cell proliferation in a mouse model of primary hyperparathyroidism. These findings suggest that evocalcet may not only reduce PTH secretion but also attenuate disease progression in hyperparathyroidism.

Abstract Somatic alterations in ZFX, a conserved zinc-finger transcription factor involved in stem cell maintenance and proliferation, have been reported across various malignancies including endometrioid carcinoma, melanoma, acute lymphoblastic leukemia, and diffuse large B cell lymphoma. Emerging evidence also links ZFX variants in zinc finger domains 12 and 13 to parathyroid tumorigenesis. Parathyroid gland tumors cause dysregulated calcium homeostasis and increased risk for bone fractures and kidney stones. While most people with parathyroid tumors have only one single gland affected (∼85%), the remaining 15% have multigland disease (MGD). This condition, previously known as parathyroid hyperplasia, has distinct clinical implications including higher rates of recurrence and association with genetic syndromes, such as multiple endocrine neoplasia type 1, that may entail risk for family members. A few studies have reported germline variants in ZFX associated with MGD; however, the contribution of ZFX variants in MGD remains poorly understood.To assess the contribution of ZFX variants in MGD, we screened parathyroid tumor DNA from 26 patients with MGD for variants in zinc finger domains 12 and 13. We identified the variant p.Arg764Trp in one sample and confirmed it to be germline. This variant has been previously reported as both germline and somatic in parathyroid adenomas and as a somatic variant in other malignancies, but has not been reported in MGD. We subsequently screened germline DNA from an additional cohort of 47 patients with MGD and found no variants in the coding sequence of zinc finger domains 12 and 13. The finding of the p.Arg764Trp variant in this cohort further supports the contribution of ZFX variants to MGD. The mechanisms by which such variants predispose affected patients for MGD as opposed to solitary parathyroid tumors merit further investigation. Additionally, while we were unable to access family history in the patient with the p.Arg764Trp variant, a previously reported three-generation family with a different ZFX variant (p.Tyr774Cys) included four females diagnosed with hyperparathyroidism, with one proband having MGD. This familial pattern raises important considerations for genetic testing and counseling of family members of patients with ZFX variants. Citation Format: Sara Halili, Lucas Iommazzo, Callie Burke, Justin Bellizzi, Jessica Costa-Guda, Andrew Arnold. Role of ZFX variants in parathyroid tumorigenesis [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 5021.

HYPOTHESIS:Mice with chronic hypercalcemia have smaller otoconia and altered surface morphology when compared with controls. BACKGROUND:Serum calcium dysregulation may contribute to otoconial degeneration and has been linked to benign paroxysmal positional vertigo (BPPV), but direct evidence is scarce. METHODS:Four female transgenic mice modeling hypercalcemic primary hyperparathyroidism (PHPT) and 4 female age-matched wild-type mice were studied. Utricular otoconia were extracted under sodium cacodylate solution, gold-coated, and imaged by Zeiss Sigma scanning electron microscope. Otoconia were hand-measured using ImageJ to quantify major axis, minor axis, and area. Statistical analysis included linear mixed-effects modeling and Mann-Whitney testing across 3 size categories (small, medium, and large). Fractured otoconia were quantified to assess structural stability. RESULTS:Transgenic PHPT mice exhibited smaller otoconia than control mice across the major axis (5.15±0.05 µm vs. 9.09±0.16 µm), minor axis (2.79±0.02 µm vs. 4.55±0.07 µm), and area (13.24±0.28 µm2 vs. 39.98±1.55 µm2). The linear mixed-effects model showed a significant interaction between group (control vs. transgenic mice) and otoconial size category. Subgroup analysis showed the greatest between-group differences among large otoconia for major axis (P=0.039), minor axis (P=0.038), and area (P=0.032). Otoconia fracture frequency was low but there was a 4.4-fold increase in cracks within the hypercalcemic mice. CONCLUSIONS:Mice with chronic hypercalcemia and PHPT have smaller otoconia when compared with normal mice, with increased frequency of fractured otoconia, consistent with a potential role of chronic calcium dysregulation in otoconial degeneration and potentially BPPV.

Abstract Loss-of-function mutations in the CDC73 tumor suppressor gene, encoding parafibromin, are the most common genetic alteration in parathyroid carcinoma. Germline CDC73 mutation causes Hyperparathyroidism-Jaw Tumor Syndrome (HPT-JT), a rare tumor predisposition syndrome including parathyroid carcinoma, and both germline and somatic CDC73 mutations are seen in patients with sporadically presenting parathyroid cancer. Oncogenic amplification of CCND1, which encodes cyclin D1, is another frequent finding in parathyroid carcinoma. Parathyroid-targeted overexpression, via a PTH-cyclin D1 (PCD) transgene, leads to parathyroid tumorigenesis, but not parathyroid malignancy, in transgenic mice. The absence of appropriate preclinical models has hindered the development of preventative strategies for patients with germline CDC73 mutation and of non-surgical treatments for parathyroid cancer in general. Thus, we sought to develop a model system that could mimic the clinical course of CDC73/parafibromin-deficient parathyroid carcinoma, including robust parathyroid hormone (PTH)-dependent hypercalcemia and invasion/metastasis.To this end, we crossed parathyroid-targeted CDC73 null mice with PCD transgenic mice, producing offspring with both CDC73 deficiency and cyclin D1 overexpression in parathyroid cells. Double mutant mice were compared to PCD mice and wildtype littermate controls. Mice were monitored for the progression of parathyroid tumorigenesis by monthly blood collection and evaluation of serum calcium and PTH. Mice were euthanized at 18 months of age. The parathyroid glands and surrounding tissue were dissected en bloc and evaluated histologically for general morphology, local invasion, and parathyroid cell proliferation via Ki-67 immunostaining. While PCD-positive mice developed hypercalcemia, a hallmark of parathyroid tumorigenesis, as expected, no significant differences in calcium levels were observed between mice with both CDC73 deficiency and cyclin D1 overexpression as compared to mice with cyclin D1 overexpression alone. Similarly, general parathyroid gland morphology and parathyroid cell proliferation were comparable between the two groups; local tumor invasion was not observed in either group. While the combined disruption of Cdc73 and the overexpression of cyclin D1 did not result in the development of parathyroid carcinoma in our mouse model, this model system may provide valuable insight into genetic modifiers of parathyroid tumorigenesis. Background strain is known to affect the phenotype of many genetically modified mouse models, including parathyroid-targeted Cdc73 deletion; however, the responsible genetic modifiers remain unknown. Our model system could serve as a key reagent in identifying genetic modifiers with relevance to parathyroid tumor development. Citation Format: Jen-Yuan C. Kao, Callie Burke, Justin Bellizzi, Andrew Arnold, Jessica Costa-Guda. CDC73 loss does not augment cyclin D1-driven parathyroid tumorigenesis in transgenic mice [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 5014.

A lack of targeted therapies make parathyroid carcinoma, a diagnostically challenging malignancy, difficult to treat. The rarity of this tumor type necessitates international collaboration to collect a sizable sample set for study. Prior studies have revealed the importance of driver mutations in the CDC73 gene and identified several putative drivers/aberrant pathways including PI3K/mTOR activation and CCND1 (cyclin D1) amplification. In this study, we sought to better understand the prevalence of putative oncogenic drivers in parathyroid carcinoma. We subjected an expanded cohort of 71 sporadic parathyroid carcinomas, fulfilling stringent WHO criteria, to next-generation DNA sequencing on a custom 16-gene targeted panel. One or more variant was detected in 44 tumors (62

Disclosure: M. Jakubowski: None. S. Yaqoob-Krzystofiak: None. C. Burke: None. J. Bellizzi: None. J. Costa: None. A. Arnold: None. Familial isolated primary hyperparathyroidism (FIHP) is a form of primary hyperparathyroidism (PHPT) in which there is familial clustering of individuals with tumors in the parathyroid glands, which secrete inappropriately elevated amounts of parathyroid hormone (PTH) leading to dysregulated calcium homeostasis. The genetic cause of FIHP in most kindreds is unknown. In vitro activating germline variants in GCM2, encoding an eponymous transcription factor critical for parathyroid gland development, were reported in a subset of FIHP kindreds, but their pathogenicity, penetrance, and implications for clinical management are uncertain. One such variant, I383M, was associated with a severe phenotype, including parathyroid carcinoma, in one kindred. We sought to evaluate the effects of GCM2 I383M on PHPT and parathyroid tumorigenesis in a mouse model. Genetically engineered mice with a Gcm2 I383M mutation were compared to wildtype littermate controls. Gcm2 I383M mutant mice were viable, developed normally and appeared healthy. Mice were followed to a terminal timepoint of 18 months, at which parathyroid tissue was excised and blood was collected. Mice were evaluated for biochemical hyperparathyroidism and parathyroid tumorigenesis. Serum calcium, PTH, parathyroid gland volume and parathyroid cell proliferation, as measured by Ki67 immunostaining, were indistinguishable between Gcm2 I383M mutant and wildtype littermates. Thus, our results demonstrate that Gcm2 I383M mutant mice do not develop PHPT or parathyroid tumors. However, as our laboratory has recently demonstrated, another Gcm2 activating variant (Y392S, equivalent to human Y394S) was similarly insufficient to drive PHPT on its own but could cooperatively enhance parathyroid tumorigenesis in a mouse model of PHPT driven by parathyroid-specific overexpression of cyclin D1, suggesting this variant might function as a mild-to-moderate predisposition allele, rather than a tumor driver. Whether GCM2 I383M might similarly serve as a predisposition allele for PHPT merits further investigation. Presentation: Saturday, July 12, 2025

Parathyroid carcinoma (PC) and atypical parathyroid tumors (APT) are incompletely understood and pose challenges in definitive diagnosis. FLCN sequence variants have recently been linked to PC and APT. Inactivating mutations in the ubiquitously expressed FLCN tumor suppressor gene, encoding folliculin, cause Birt-Hogg-Dubé syndrome (BHD), a rare tumor predisposition syndrome. Germline inactivating FLCN variants, accompanied by somatic allelic loss, were reported in 2 unrelated patents with PC, both with clinical features, but no diagnosis, of BHD. Somatic frameshift variants of likely pathogenicity were reported in 1 patient with PC and 1 with APT. On the other hand, neither PC nor APT has been reported in sizeable BHD series. To better understand the frequency of FLCN variants in PC and APT, we analyzed a series of 10 patients with sporadic PC and 14 with APT by direct Sanger DNA sequencing. We identified no inactivating FLCN mutations in any of the PC or APT samples examined. A germline missense variant (p.Gly325Val), predicted as benign/tolerated, was seen in 1 PC and a synonymous variant in 1 APT. The absence of pathogenic mutations detected in our series of PC and APT further suggests that FLCN variants are rare in these tumors. Nevertheless, the potential roles of FLCN in the pathogenesis of PC and APT merits further consideration and study.

Hyperparathyroidism jaw-tumor syndrome is an autosomal dominant disorder caused by mutations in the CDC73/HRPT2 tumor suppressor gene, encoding parafibromin, and manifesting benign or malignant parathyroid tumors, ossifying jaw fibromas, uterine tumors, and kidney lesions. Sporadic parathyroid carcinomas also frequently exhibit inactivating CDC73 mutations and loss of parafibromin. To study the role of CDC73 in parathyroid cell proliferation in vivo, we generated mice with a parathyroid-specific deletion of Cdc73. Homozygous knockout mice on a mixed B6/129/CD1 background had decreased serum calcium and PTH and smaller parathyroid glands compared with heterozygous or wild-type littermates, whereas homozygous Cdc73-null mice on other backgrounds exhibited no abnormalities in parathyroid gland function or development. No hypercalcemia or parathyroid hypercellularity was observed in mice of any background examined at any age. Thus, although postnatally acquired complete loss of CDC73 causes parathyroid cell proliferation and hyperparathyroidism, such as seen in human hyperparathyroidism jaw-tumor syndrome, our results suggest that earlier, developmentally imposed complete loss of Cdc73 can cause a primary defect in parathyroid gland structure/function in a strain-dependent manner. This striking disparity in parathyroid phenotype related to genetic background offers a unique opportunity in an in vivo model system to precisely dissect and identify the responsible molecular mechanisms.

Abstract Parathyroid carcinoma (PC) is rare, poses challenges in definitive diagnosis and treatment, and carries a high rate of recurrence and mortality. Atypical parathyroid tumors (APT) have histopathologic features of malignancy but lack unequivocal evidence of invasion and are of uncertain malignant potential. Analysis of genetic alterations that may play a role in PC development, particularly those that might serve as markers of malignant potential, is important for further understanding this disease. FLCN is a gene found on chromosome 17 that encodes folliculin, a ubiquitously expressed protein with roles in multiple cellular processes such as apoptosis and cell signaling. Inactivating FLCN mutations cause Birt-Hogg-Dubé syndrome (BHD), an autosomal dominant condition characterized by benign fibrofolliculomas, pulmonary cysts and spontaneous pneumothorax, and increased risk of renal cancer. Benign parathyroid tumors had previously been reported in a few BHD patients and recently one study has reported a link between FLCN mutations and parathyroid carcinoma. Germline frameshift mutations in FLCN, accompanied by somatic loss of the normal allele, were reported in two unrelated PC patients, both with clinical features suggestive of BHD, but no established diagnosis of BHD. An identical somatic frameshift FLCN mutation was reported in one PC and one APT. Interestingly, the authors noted low coverage of FLCN on whole exome sequencing, resulting in the failure to detect variants identified by Sanger sequencing, thus raising the possibility that prior studies may have missed FLCN variants. To better understand the frequency of germline or somatic FLCN mutations in PC and APT, we performed Sanger sequence analysis of the entire coding region of the FLCN gene on available tumor DNA from 11 PCs and 15 APTs. We identified no inactivating FLCN mutations in any of the PC or APT samples examined. A germline missense variant, resulting in a p.Gly325Val change, predicted to be benign/tolerated by in silico analyses, was seen in one PC and a synonymous variant (c.1233G>A) was seen in one APT. The absence of pathogenic mutations in our series parathyroid carcinoma and atypical parathyroid tumors suggests that FLCN mutations are rare in these parathyroid tumors. Nevertheless, germline FLCN testing and/or additional screening for BHD-related lesions in parathyroid carcinoma patients merits further consideration and study. Citation Format: Callie Burke, Justin Bellizzi, Jessica Costa-Guda, Andrew Arnold. FLCN variants and their role in parathyroid cancer and atypical parathyroid tumors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 6265.

Primary hyperparathyroidism is a common endocrine disorder that is most often caused by a sporadic single-gland parathyroid adenoma. Currently, the only known and experimentally validated oncoprotein for parathyroid adenomas is cyclin D1. Overexpression of cyclin D1 has been reported in 20-40% of cases; about 8 percent of cases harbor a DNA rearrangement involving the cyclin D1 (CCND1) locus. Other molecular mechanisms underlying this cyclin D1 overexpression have yet to be identified. Here, we explored a potential parathyroid tumorigenic mechanism that could increase cyclin D1 stability through a defect in the AMBRA1 E3 ligase adaptor, a molecule responsible for its degradation. Evidence that AMBRA1 may be a tumor suppressor gene includes: accumulation of cyclin D with hyperproliferation in AMBRA1-deficient cells, context-dependent growth of tumors in Ambra1-deficient mice, and the correlation between low levels of AMBRA1 and poor clinical outcomes in The Cancer Genome Atlas. Therefore, we examined AMBRA1 for evidence of tumor suppressor-type inactivation in a cohort of 98 typical parathyroid adenomas. Tumor genomic DNA was subjected to PCR-amplification of all 18 coding exons of AMBRA1, followed by Sanger sequencing. Sequences were analyzed by comparison to the normal reference sequence (ENST00000683756.1), seeking to assess any observed intragenic or splice-site variants as potential or likely loss-of-function mutations. Thus far, with 97 percent of the coding region for this 1298-amino acid protein fully sequenced, we have identified an inactivating mutation in 1 of 98 tumors (1%): c.126G>A, resulting in an early stop codon p.Trp42. This mutation showed loss of heterozygosity and was confirmed to be somatic by sequencing of the patient’s matched germline DNA. We also identified likely non-pathogenic variants in 5 tumors: missense variant c.3385G>T (p.Ala1129Ser), 3’ UTR variants c.*576A>T and c.*491T>C and synonymous variants c.579C>T and c.2776C>T, each in a single tumor. 4 of 5 were found as germline variants in the normal population and the one missense scored as benign by in silico criteria. To conclude, our observations suggest that AMBRA1 may function as a classical tumor suppressor gene in sporadic parathyroid adenomas at very low frequency. Further study may include protein expression analysis of AMBRA1 to investigate influences on gene expression. To further elucidate the mechanisms behind cyclin D1 overexpression in parathyroid adenomas, other molecules that participate in its degradation should be explored as potential tumor suppressors. Citation Format: Stephanie Chinwo, Justin Bellizzi, Jessica Costa-Guda, Andrew Arnold. Molecular analysis of AMBRA1 as a candidate tumor suppressor in sporadic parathyroid adenomas. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 5279.

Abstract Disclosure: S. Chinwo: None. J. Costa: None. A. Arnold: None. J. Bellizzi: None. Primary hyperparathyroidism is a common endocrine disorder that is most often caused by a sporadic single-gland parathyroid adenoma. Currently, the only known and experimentally validated oncoprotein for parathyroid adenomas is cyclin D1. The overexpression of cyclin D1 protein has been reported in 20-40% of cases and about 8 percent of cases harbor a DNA rearrangement involving the cyclin D1 (CCND1) locus. Otherwise, the molecular mechanisms underlying this cyclin D1 overexpression have yet to be identified. Here, we explored a potential parathyroid tumorigenic mechanism that could increase cyclin D1 stability through a defect in the AMBRA1 E3 ligase adaptor, a molecule responsible for its degradation. Evidence that AMBRA1 may be a tumor suppressor gene includes the accumulation of cyclin D with hyperproliferation in AMBRA1-deficient cells, the context-dependent growth of tumors in Ambra1-deficient mice, and the finding that low levels of AMBRA1 correlated with poor clinical outcomes in The Cancer Genome Atlas. Therefore, we proceeded to examine AMBRA1 for evidence of tumor suppressor-type inactivation within a cohort of 98 typical parathyroid adenomas. Genomic DNA samples from these tumors were subjected to PCR-amplification of all 18 coding exons of AMBRA1, followed by Sanger sequencing. Sequences were analyzed by comparison to the normal reference sequence: ENST00000683756.1, seeking to assess any observed intragenic or splice-site variants as potential or likely loss-of-function mutations. Thus far, with 97 percent of the coding region for this 1298-amino acid protein fully sequenced, we have identified an inactivating mutation in 1 of the 98 tumors (1%): c.126G>A in exon 2, resulting in an early stop codon p.Trp42*. Sequencing the patient’s matched germline DNA confirmed the mutation was somatic and showed loss of heterozygosity. We also identified likely-nonpathogenic variants in 5 parathyroid adenomas: missense variant c.3385G>T (p.Ala1129Ser), 3’ UTR variants c.*576A>T and c.*491T>C and synonymous variants c.579C>T and c.2776C>T, each in a single tumor. 4 of 5 were found as germline variants in the normal population and the one missense scored benign by in silico criteria. To conclude, our observations suggest that AMBRA1 may function as a classical tumor suppressor gene in sporadic parathyroid adenomas at very low frequency. Further study may include protein expression analysis of AMBRA1 to investigate influences on gene expression. To further elucidate the mechanisms behind cyclin D1 overexpression in parathyroid adenomas, other molecules that participate in its degradation should be explored as potential tumor suppressors. Presentation: Saturday, June 17, 2023

Abstract Disclosure: J. Costa-Guda: None. J. Bellizzi: None. S. Gorka: None. C. Titarenko: None. A. Arnold: None. Parathyroid carcinoma (PC) is a rare, life-threatening malignancy. Despite relatively slow growth, recurrent or metastatic disease is typically incurable and often fatal due to cumulative complications of sustained elevation of parathyroid hormone (PTH) and accompanying hypercalcemia. Medical therapy aids in managing sequelae of hyperparathyroidism (HPT) but non-surgical interventions to reduce tumor burden have been largely ineffective in PC. Cyclin D1 amplification is seen in nearly 30% of PC, making it a key potentially actionable therapeutic target. Importantly, in contrast to other tissue contexts, cyclin D1-driven parathyroid neoplasia has been experimentally demonstrated to depend on cyclin D1’s ability to bind and activate its partner cyclin dependent kinases (cdk). This suggests that parathyroid tumor cells may be more sensitive to alterations in cdk-mediated proliferation control, and thus more responsive to pharmacologic cdk inhibition, compared with other tumor types. We sought to test the efficacy of cdk4/6 inhibitor therapy in cyclin D1-overexpressing parathyroid tumors in a pre-clinical system. We utilized a mouse model, harboring a PTH-CCND1 (PCD) transgene to drive parathyroid-specific cyclin D1 overexpression, which develops hormonally-active parathyroid tumors with known kinetics. Transgenic mice with established hyperparathyroidism and wild type littermates were treated with the cdk4/6 inhibitors palbociclib or abemaciclib for three weeks. While vehicle-treated PCD mice showed a 7% increase in serum calcium (Ca) levels over the course of treatment, Ca decreased by 1.8% and 0.5% in palbociclib- and abemaciclib-treated PCD mice, respectively, from pre-treatment levels. This treatment effect persisted for at least 60 days after discontinuation of treatment. Decreased parathyroid proliferation, as evidenced by 4.7-fold and 17.3-fold fewer Ki-67-positive cells in palbociclib- and abemaciclib-treated PCD mice, respectively, compared with vehicle-treated PCD mice, was also seen. Cdk inhibitor treatment had no effect on Ca or parathyroid cell proliferation in wild type mice. These results demonstrate that palbociclib and abemaciclib can significantly slow the progression of hypercalcemia and parathyroid proliferation in a cyclin D1-driven parathyroid tumor model, suggesting that cyclin D1 overexpressing parathyroid cells may be particularly sensitive to cdk4/6 inhibition. Our findings provide the first evidence of a direct pharmacologic effect in a relevant preclinical system to support the hypothesis that selected HPT patients, eg those suffering from advanced PC with activated cyclin D1, could substantially benefit from cdk4/6 inhibitor therapy. This evidence may aid clinicians in selecting therapeutic interventions, including when considering such drugs for off-label use or inclusion of PC patients in “basket” clinical trials. Presentation: Thursday, June 15, 2023

Primary hyperparathyroidism (PHPT) is a common endocrinopathy for which several pathogenic mechanisms, including cyclin D1 overexpression, have been identified. Vitamin D nutritional status may influence parathyroid tumorigenesis, but evidence remains circumstantial. To assess the potential influence of vitamin D insufficiency/deficiency on initiation or progression of parathyroid tumorigenesis, we superimposed vitamin D insufficiency or deficiency on parathyroid tumor-prone parathyroid hormone-cyclin D1 transgenic mice. Mice were placed on diets containing either 2.75 IU/g, 0.25 IU/g, or 0.05 IU/g cholecalciferol, either prior to expected onset of PHPT or after onset of biochemical PHPT. When introduced early, superimposed vitamin D insufficiency/deficiency had no effect on serum calcium or on parathyroid gland growth. However, when introduced after the onset of biochemical PHPT, vitamin D deficiency led to larger parathyroid glands without differences in serum biochemical parameters. Our results suggest that low vitamin D status enhances proliferation of parathyroid cells whose growth is already being tumorigenically driven, in contrast to its apparent lack of direct proliferation-initiating action on normally growing parathyroid cells in this model. These results are consistent with the hypothesis that suboptimal vitamin D status may not increase incidence of de novo parathyroid tumorigenesis but may accelerate growth of a preexisting parathyroid tumor.

Abstract Primary hyperparathyroidism is a common endocrine disorder that is most often caused by a sporadic single-gland parathyroid adenoma. Currently, the only known and experimentally validated oncoprotein for parathyroid adenomas is cyclin D1. The overexpression of cyclin D1 protein has been reported in twenty to forty percent of cases. Eight percent of this overexpression can be explained by DNA rearrangement involving the cyclin D1 locus. The remaining molecular mechanisms behind this overexpression have yet to be identified. In this study, we explored a potential parathyroid tumorigenic mechanism that could increase cyclin D1 stability through a defect in the AMBRA1 E3 ligase adaptor, a molecule responsible for its degradation. Evidence that AMBRA1 may be a tumor suppressor gene includes the accumulation of cyclin D with hyperproliferation in AMBRA1-deficient cells, the context-dependent growth of tumors in Ambra1-deficient mice, and the finding that low levels of AMBRA1 correlated with poor clinical outcomes in The Cancer Genome Atlas (TCGA). We have therefore proceeded examine AMBRA1 for evidence of tumor suppressor-type inactivation within a cohort of parathyroid adenomas. Genomic DNA from twenty-six typical sporadic parathyroid adenomas was subjected to PCR-amplification of all eighteen coding exons of AMBRA1, followed by Sanger sequencing. Sequences were analyzed by comparison to the normal reference sequence, seeking to assess any observed intragenic or splice-site variants as potential or likely loss-of-function mutations. Thus far, with ninety-seven percent of the coding region fully sequenced, no intragenic AMBRA1 inactivating mutations were observed. Three known single nucleotide polymorphisms, considered non-pathogenic, were identified: missense variant c.3028G>T, and 3' UTR variants c.*491T>C and c.*576A>T, each in a single tumor. Future study of increased numbers of parathyroid adenomas is needed to exclude the possibility that AMBRA1 inactivation could play a driver role in a small percentage of cases, and the possibility that AMBRA1 might contribute less directly e.g. via secondarily altered expression also warrants investigation. Nonetheless, our current observations strongly suggest that AMBRA1 does not have a frequent role as a classic 2-hit tumor suppressor gene in sporadic parathyroid adenomas. Presentation: Saturday, June 11, 2022 1:00 p.m. - 3:00 p.m.