Thyroid Cancer Genetics: How Close are we to Personalizing Clinical Management?

KEYWORDS::

• Cowden syndrome
• personalized medicine
• PTEN mutation
• thyroid cancer

“We celebrate the revelation of the first draft of the human book of life … it is humbling for me and awe inspiring to realize that we have caught the first glimpse of our own instruction book.”

– Francis Collins, NIH (MD, USA)

Over two millennia ago, Hippocrates (460–377 BC) emphasized the importance of individualizing medical care, proclaiming, “It is more important to know what sort of person has a disease than to know what sort of disease a person has” [101]. Even more poignantly, the alleged first Chinese medical text N‘ai Ching (ca 2600 BC) linked prevention with superior doctors and reactive treatment of disease to inferior doctors. Today, advances in genome technologies and the ensuing outpouring of genomic information related to cancer have led to the convergence of the discovery of science and clinical medicine, allowing clinicians to tailor the management for individual patients. Discoveries in the field of endocrine cancer have contributed much to the early successes we are seeing in personalized healthcare and heralded the way forward for personalizing management, not only for the patient, but also for their families.

While thyroid cancer accounts for approximately 1% of all newly diagnosed cancer cases, it is the most common malignancy of endocrine organs. Its incidence has increased significantly in the USA and in other countries over the last several decades [1]. This progressive increase cannot be explained simply by improved diagnostic histopathology or preclinical detection through neck ultrasounds; other factors such as environmental influences and molecular alterations must be taken into account [2]. Well differentiated nonmedullary thyroid cancer (NMTC) accounts for 95% of thyroid malignancies and 5% of these will have familial disease. This contrasts with 10–25% of patients with medullary thyroid cancer (MTC) having a familial etiology [3,4]. Molecular studies have led to increased appreciation of the heterogeneity of thyroid cancers, with hereditary predisposition, somatic mutation and epigenetic modulation all contributing to tumor behavior [5]. For example, the discovery of the RET proto-oncogene as the causative gene for multiple endocrine neoplasia type-2 allowed for gene-based molecular diagnosis, predictive testing and genotype-specific management for affected individuals. This continues to serve as the paradigm for personalizing management based on germline mutation status and precise genotype [4,6]. Somatic RET mutations are also found in at least 50% of sporadic MTCs. Importantly, this link between MTC and RET has recently led to the US FDA-approval of vandetanib in 2011, an orally bioavailable inhibitor of RET, VEGFR-2 and EGFR for use in advanced MTC, even those arising in the sporadic setting, following dramatic improvements in progression-free survival over placebo [7].

Germline mutations in familial NMTC syndromes have not been defined as well as those in MTC. Familial NMTC syndromes include a number of disorders, the commonest of which is Cowden syndrome (CS). As data from the follow-up of CS patients accumulate, we now know that the lifetime risk of thyroid cancer in CS patients with PTEN mutations is approximately 35% [8]. We demonstrated that the presence of a germline PTEN mutation significantly elevated the risk of epithelial thyroid cancer by approximately 70-fold when compared with the general population [9]. We also saw an increased thyroid cancer risk in CS and CS-like patients with germline SDHx alterations and KLLN promoter hypermethylation, two novel genes linked with CS [10,11]. Importantly, our study revealed gene-specific differences in clinical presentations between the three genes, which would suggest the personalization of management in a gene-specific manner. CS and CS-like patients with germline PTEN mutations were more likely to be younger – some presenting as early as 7 years of age – and tended to be associated with benign thyroid diseases, such as thyroid adenoma and thyroiditis. There was an over-representation of follicular thyroid cancer (FTC) compared with papillary thyroid cancer (PTC) in patients with germline PTEN mutations, these were distinct from those who had germline SDHx or KLLN alterations. Finding a germline PTEN mutation in an individual would mean increased clinical screening for other associated cancers, such as breast and endometrial cancers [8]. Knowing that the renal cancers associated with germline PTEN mutations are mainly papillary in histology informs the type of clinical surveillance, namely that the standard ultrasound is insufficient; instead, computed tomography or MRI would be the platform of choice [12]. Individuals with germline SDH mutations have an increased risk of breast and thyroid cancers compared with those with PTEN mutations [13], while individuals with germline KLLN promoter hypermethylation have increased risks of breast and renal cancers [10]. This implies that surveillance of CS patients will need to be personalized depending on the underlying germline alteration involved.

Hereditary thyroid cancer, as illustrated above, provides a model for how personalized healthcare can successfully be applied to the individual and his/her family. Nonetheless, the vast majority of thyroid cancers are sporadic. Sporadic NMTC is characterized by two distinct somatic molecular mechanisms: point mutations or chromosomal rearrangements, which are, in turn, associated with specific etiologic factors involved in thyroid carcinogenesis [5]. There are distinct differences in tumor-mutation profiles between different histological subtypes. Activating somatic mutations in BRAF (in particular the V600E mutation) have been identified in approximately 50% of PTCs and correlate with poor outcome [14,15]. Somatic RAS mutations, particularly those that involve NRAS and HRAS, are found in 10–20% of PTCs (usually in the follicular variant) and in 40–50% of FTCs [16]. Translocations that result in a fusion between PAX8 and the peroxisome proliferator-activated receptor γ gene are found in 30–40% of FTCs [17]. Although these molecular profiles have not as yet significantly affected the diagnostic practice of surgical pathology, there is much hope and hype that it can be used to significantly improve the sensitivity of cytological fine-needle aspiration diagnoses of thyroid nodules. Postsurgical results of nodules with ambiguous fine-needle aspiration yields show that only 20–30% of these patients have cancer. We and others have developed possible gene signatures which show promise in being able to accurately predict thyroid cancer from benign thyroid nodules [18,19].

Characterization of these genetic events in thyroid cancer may have diagnostic and prognostic implications, but most excitingly, it also provides an opportunity to develop therapies that are targeted at these potential molecular drivers. For patients with MTC treated with vandetanib, a similar response rate was seen in patients with hereditary MTC and in RET mutation-positive sporadic MTC [7]. The M918T mutation (a tyrosine kinase domain mutation) was the most frequent somatic mutation found in patients with RET mutation-positive sporadic MTC. The response rate in patients with sporadic MTC and M918T mutations who were treated with vandetanib was significantly higher (54 vs 32%) than in patients who were negative for mutations or for whom the mutation status was unknown. This places MTC alongside melanoma and gastrointestinal stromal tumors, whereby a woeful lack of efficacy with conventional chemotherapy has now been replaced with renewed hope using targeted therapy.

The concept of personalized medicine is not new, but what is new is our ability to utilize an individual‘s and the tumor‘s genetic make-up to determine clinical management with molecular testing. Over the past two decades, we were able to capitalize on the discovery of RET mutations in multiple endocrine neoplasia type-2 related and sporadic MTC into successful strategies for predictive testing of affected individuals and hence, genotype-specific clinical surveillance and prophylactic surgery [6], and along the way, revolutionized treatment for MTC. It is hoped that similar progress can be achieved by a deepening understanding of how germline PTEN mutations, and SDHx and KLLN alterations impact on NMTC tumorigenesis. We await with interest the outcomes of ongoing trials focused on downstream targets of PTEN, such as mTOR and PIK3CA inhibitors that are underway in both CS patients (NCT00971789) as well as in sporadic thyroid cancer with somatic PTEN loss (NCT01430572) [102,103]. We now know that germline SDHx variants may affect mitochondrial metabolite dysregulation and subsequent tumorigenesis [13]. While the mechanism(s) of disruption of mitochondrial function leading to neoplasia remain unclear, further research may unravel how we can harness drugs targeting ‘energetics‘ for prevention or treatment.

As we learn more about the heterogeneity of tumor formation, we are learning the increasing importance of crosstalk between different pathways leading to differential organ-specific carcinogenesis. It is certain that further progress in our understanding of thyroid cancer molecular genetics will lead to the discovery of novel mutations and other genetic and epigenetic alterations that will help clarify the biology of MTC and NMTC and foster the development of new targeted therapeutic approaches. However, we are going to need to increasingly rethink how we integrate the molecular data that are obtained. In the past, researchers have been approaching the problem in ‘germline‘ versus ‘somatic‘ silos; however, going forward, it is going to be important that we embrace that tumor biology is likely to be dictated by somatic changes in the cancer multi-‘ome‘, the microenvironment, as well as germline ‘omic make-up. It is the natural tendency to focus on one aspect. The most successful manner to personalize clinical management is a multi- and inter-disciplinary one, whereby all ‘omic networks (genomic, metabolomics and others) are integrated with clinical phenotypes (phenomic), including family health history. While many of us know what must be done to reach the nirvana of personalized healthcare, there are powerful forces beyond our control that may derail these noble efforts: lack of research funding, the lack of foresight to invest in this type of bridging research between good ‘omics data and clinical implementation (and we do not mean clinical drug trials), and extraordinary regulatory issues, all of which have already stifled our research efforts, at least in the USA.

“In expanding the field of knowledge we but increase the horizon of ignorance.”

– Henry Miller (1891–1980)

Financial & competing interests disclosure

Some of the primary research described in this editorial is supported, in part, by grants P01CA124570 and R01CA118980 from the National Cancer Institute, MD, USA and the Breast Cancer Research Foundation (all to C Eng). J Ngeow is the National Medical Research Council (Singapore) Fellow and an Ambrose Monell Foundation Cancer Genomic Medicine Clinical Fellow at the Cleveland Clinic Genomic Medicine Institute. C Eng 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, generously funded, in part, by the Fred Morgan Kirby Foundation. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.

Additional information

Funding

Financial & competing interests disclosure Some of the primary research described in this editorial is supported, in part, by grants P01CA124570 and R01CA118980 from the National Cancer Institute, MD, USA and the Breast Cancer Research Foundation (all to C Eng). J Ngeow is the National Medical Research Council (Singapore) Fellow and an Ambrose Monell Foundation Cancer Genomic Medicine Clinical Fellow at the Cleveland Clinic Genomic Medicine Institute. C Eng 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, generously funded, in part, by the Fred Morgan Kirby Foundation. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.