HABP2in Familial Non-medullary Thyroid Cancer: Will the Real Mutation Please Stand Up?

“Veritas filia temporis” (“Truth is the daugher of time”) --Aulus Gellius; Noctes Atticae 12.11.7 Epithelial thyroid cancer is common, and its incidence has continued to rise over the last 15 years. It is believed that approximately 90% of thyroid cancer is sporadic because of somatic genetic alterations (1). Only 3% to 9% of all thyroid cancers are familial nonmedullary thyroid cancer (FNMTC) cases, defined by the presence of thyroid cancer in two or more first-degree relatives, in the absence of obvious predisposing environmental factors. Papillary thyroid cancer (PTC) is the most common (85%-91%) histological subtype among FNMTCs. Out of all FNMTCs, only 5% in the syndromic form have well-defined driver germline mutations. This includes Cowden syndrome, familial adenomatous polyposis/Gardner syndrome, Carney complex type 1, Werner syndrome, and DICER1 syndrome. Even though these are uncommon, it is critical to recognize these disease entities to enable timely surveillance for associated thyroidal and extrathyroidal malignancies to facilitate early detection and successful treatment for the patients and their families. The majority (95%) of FNMTCs are nonsyndromic with less well-defined genetic susceptibility. Understanding the genetic basis of this heterogenous disease entity, and the identification of molecular predictors of disease aggressiveness, could help better risk-stratify, allow predictive testing of family members to identify family members at risk for FNMTC, and guide high-risk surveillance and management planning. Attempts to identify susceptibility genes have been met with variable success. With the utilization of exome sequencing of a large nonsyndromic FNMTC kindred, Gara et al. reported a novel germline mutation in HABP2 G534E (2). Hard on its heels, Zhang and Xing in this issue of the Journal (3) report 14% of 29 unrelated FNMTC families with this same mutation. The Hyaluronan-Binding Protein 2 (HABP2) gene is located on chromosome sub-band10q25.3. Using exome sequencing of peripheral blood DNA in seven affected members of an FNMTC kindred (PTC and follicular adenoma) and unaffected spouses as control subjects, Gara et al. found that a germline G534E variant was identified in the HABP2 gene and found high levels of HABP2 protein in the FNMTC tumors but none in normal thyroid tissue. Overexpressing the G534E variant enhanced colony formation, increased cell migration, and promoted the formation of foci, all hallmarks of cancer. Expressing equal amounts of normal HABP2 and the G534E variant in the same cells still led to increased numbers of foci, indicating that the variant has a dominant-negative effect on the tumor suppressor function of normal HABP2 (2). Of note, the filtering criterion used to identify the mutation in the kindred was an allele frequency of 1% or less for variant prioritization. The two databases that were utilized were the 1000 Genomes Project (phase 3, version 5 [20130502]) and the HapMap3 (release3) databases that included the general population. In comparison, the HABP2 G534E variant was reported in 4.7% of apparently sporadic PTC patient cases from The Cancer Genome Atlas (TCGA) dataset, compared with 0.7% in population databases where disease status is not known. These were further highlighted by subsequent letters to the editor (addressing the Gara et al. report) by Zhou et al. (4), Sponziello et al. (5), and Tomsic et al. (6). The latter commented on higher reported frequencies of the HABP2 G534E variant from other public databases including the Exome Aggregation Consortium database, with an average frequency of 2.2% and 3.3% in the non-Finnish European population (www.exac.broadinstitute.org) and approaching 4% among Europeans and European Americans from public datasets, suggesting that the exact role of this variant in pathogenesis remains unclear (6). A study from China also found no germline HABP2 G534E mutations in probands of 12 FNMTC families and 217 sporadic PTC patient cases for somatic mutation, suggesting population differences and that this particular variant that is not seen in the general Chinese population may not be associated with PTC patient cases in the Chinese population (7), suggesting that there are substantial ethnic differences in prevalence and importance of this variant, if any, in FNMTC tumorigenesis. This highlights the challenges in interpreting population data as many include a mixed population of diseased and normal individuals. Indeed, the ExAC database cited includes data from 60 706 persons, 7601 (12.5%) of whom were patients with cancer who were also included in TCGA database. It is important to emphasize that in vitro functional interrogation will continue to play an important role and that, unlike some advocate, animal models are limited. The radical few believe that only if a variant shows a phenotype in a mouse model should it be considered pathogenic. This view is both impractical and dangerous: there are many examples of mouse models where there is no phenotype although the germline Mendelian gene mutation has been clearly documented to cause disease in humans. Therefore, to understand the pathogenicity of variants identified through exome and genome sequencing will still require “old school” familial studies. It is in this context that the article by Zhang and Xing is very timely. They performed targeted DNA sequencing of HABP2 in 64 members from 29 unrelated kindreds with clinically defined FNMTCs, including 43 PTC patient cases, five benign thyroid neoplasm patient cases, and 16 control subjects. Six (14%, 6/43) individuals from four (14%, 4/29) unrelated kindreds tested positive for the germline HABP2 G534E genetic variant. All patient cases had PTC and were heterozygous for the variant while none of the benign thyroid neoplasm cases or control subjects harbored the variant (3), supporting HABP2 G534E as a susceptibility gene in a subgroup of FNMTCs. It was unclear in the article if all 29 kindreds and those with the HABP2 G534E variant were of European descent. That information would have been useful. The authors also highlight that in some kindreds there was only one individual affected with cancer available for genotyping. That HABP2 G534E segregates with affected members does not exclude the possibility that this variant, even if not pathogenic, could be in linkage with the “real” disease-causing mutation involved in these cases. It is therefore important that clinicians and research labs make the added collaborative effort of testing this variant in as many affected and unaffected FNMTC families to further clarify its role and help inform future clinical decision-making. It is evident that the majority of FNMTCs remain unresolved because either they are likely to be private mutations or additional susceptibility genes have yet to be identified. What many in the community suspect is that multiple variants interacting with one another, even within a single individual or a single family, might be what lends susceptibility to many cases of FNMTC. Gellius’ maxim, “Veritas filia temporis,” or “Truth is the daughter of time,” has been interpreted differently since its formulation. Some readers of his notebook “Noctes Atticae” (“Athenian Nights”), understood it to mean that time leads truth to light. In this perspective, the professional historian can sit passively back and wait. In postmodern scientific practice with clinical implications, truth often demands opposition, controversy, and even conflict for resolution. Our understanding of the role that the HABP2 G534E variant plays in FNMTC has evolved and will continue to do so with time and conflict. Scientific conflict, as illustrated by HABP2 in FNMTC, should be objective and constructive, never ad hominem, thus challenging groups to work together and come up with better experiments, which will lead to ultimate resolution. “Now that we have met with paradox, we have some hope of making progress.” -- Niels Bohr CE is a member of the Strategic Advisory Board of N-of-One. Otherwise, no author had any financial or personal relationships that could inappropriately influence or bias this work. JN is awarded the NMRC (Singapore) Transition Award. CE is the Sondra J. and Stephen R. Hardis Chair of Cancer Genomic Medicine at the Cleveland Clinic and is an American Cancer Society Clinical Research Professor.