Genotype-Phenotype Correlations in the Hyperparathyroidism-Jaw Tumor Syndrome

Abstract

Establishing genotype-phenotype correlations in disorders of hereditary endocrine neoplasia is important for clinical screening, genetic counseling, prognostication, surveillance, and surgical strategy, and may also provide clues about disease pathogenesis. Important genotype-phenotype correlations are recognized, for example, in pheochromocytoma/paraganglioma and multiple endocrine neoplasia type 2A. The presence of such correlations has been less clear in other familial endocrine disorders associated with primary hyperparathyroidism including multiple endocrine neoplasia type 1, and the hyperparathyroidism-jaw tumor syndrome (HPT-JT). Characteristic features of HPT-JT, apart from fibro-osseous jaw tumors and uterine lesions, include renal neoplasms, such as Wilms tumor and mixed epithelial and stromal tumor ("renal hamartomas"), and a high incidence of parathyroid cancer. Emerging evidence suggests two different genotype-phenotype correlations in HPT-JT based on the type of variant in the CDC73 tumor suppressor gene. Although multiple CDC73 genotypes can give rise to the Wilms tumor phenotype in HPT-JT, the development of mixed epithelial and stromal tumor of the kidney specifically correlates with the presence of a start-loss variant affecting the initiator methionine codon of parafibromin, the protein product encoded by CDC73. Furthermore, the risk of parathyroid cancer in HPT-JT also appears to correlate with genotype: CDC73 frameshift indel, splice-site, and stop-gain genotypes are associated with a greatly increased risk of parathyroid carcinoma compared to carriers of CDC73 missense and nonframeshift indel variants. The recognition of such genotype-phenotype correlations in HPT-JT may impact genetic counseling, patient care and disease surveillance.

Keywords: parathyroid cancer; primary hyperparathyroidism; tumor suppressor gene.

Figure 1.

Manifestations of the hyperparathyroidism-jaw tumor syndrome (HPT-JT). Manifestations of HPT-JT may include cemento-ossifying tumors of the maxilla or mandible (that are odontogenic in origin and entirely distinct from “brown” tumors that can be a manifestation of osteitis fibrosa cystica resulting from the metabolic disease of severe primary hyperparathyroidism); parathyroid disease (∼20% with parathyroid cancer, 60% with benign adenomas that often include “atypical parathyroid adenomas,” ∼20% with no parathyroid disease); kidney lesions including mixed epithelial and stromal tumors (MEST) and Wilms tumor; and uterine abnormalities that frequently manifest as menorrhagia requiring early hysterectomy, a high incidence of miscarriages, and uterine lesions such as adenofibromas, leiomyomas, and adenomyosis. Renal cysts have been reported as a manifestation of HPT-JT; however, see text for caveats and further discussion on this point. Because of the incomplete and variable penetrance associated with germline CDC73 loss-of-function variant, the manifestations of HPT-JT can differ greatly between affected individuals and within kindreds.

Figure 2.

Distribution of germline CDC73 variants in the hyperparathyroidism-jaw tumor syndrome (HPT-JT). The distribution of types of germline CDC73 variants identified in subjects with HPT-JT is shown as a donut chart, with the percentage of the total for each category of genetic variant indicated. The legend indicates color-coding of the category of genetic variant in the donut chart: blue, frameshift indel or stop-gain (“nonsense”) (72% of total); purple, gross deletion, gross insertion, or splice-site (15% of total); yellow, missense or nonframeshift indel (12% of total); orange, other (variants in intronic or regulatory regions) (1% of total). Data taken from 419 patients with germline CDC73 variants from the discovery and validation cohorts of Li et al (19).

Figure 3.

Genotype of CDC73 germline variant determines risk of parathyroid cancer. Kaplan-Meier plots of event-free survival of parathyroid cancer (left) and primary hyperparathyroidism (PHPT) (right) are shown following categorization by predicted impact of the CDC73 germline variant. Patients in the discovery and validation cohorts with germline CDC73 variant from Li et al were pooled (19). Subsequent Kaplan-Meier analysis revealed that subjects with high-impact variants had a significantly higher risk of developing parathyroid carcinoma compared to subjects harboring low-impact variants (left plot; hazard ratio, 13; P < .001), despite having a similar risk of developing PHPT (right plot; hazard ratio, 1.22; P = .25). Legend: Red, high-impact pathogenic CDC73 variant category (gross deletions, frameshift indels, splice-site, stop-gain variants); blue, low-impact pathogenic CDC73 variant category (missense and nonframeshift indel variants). Insets in red font show hazard ratio, P-value, and confidence intervals (CI) of the respective analyses. Data modified from Fig. 3 of reference (19).