Low-risk thyroid carcinoma: when is the treatment worse than the disease?

The routine use of imaging modalities has led to the discovery of numerous incidental abnormalities, from pulmonary nodules to adrenal adenomas. Faced with the challenge of unexpected findings, clinicians often rely on epidemiologic studies, expert opinions and professional society-sponsored clinical practice guidelines, in the absence of sufficiently strong interventional evidence, when deciding whether or not to act.

Increased use of neck imaging modalities, including laryngoscopy, computed tomography and ultrasound, has led to a tenfold increase in the discovery of thyroid nodules that are nonpalpable by manual neck examination [1]. These nodules are often biopsied, which, in turn, leads to higher rates of malignancy diagnosis. The incidence of thyroid cancer has increased from 1973 to 2000 by 2.4-fold, with papillary thyroid microcarcinoma (PTMC; <1 cm) being the predominant new cancer discovered [2]. The increase in incidence of these cancers is believed to be a direct consequence of increased imaging and not an actual increase in the number of cancers in the population.

The highest prevalence of incidentally found thyroid cancer in the literature is reported in a retrospective study of 1475 patients undergoing thyroid ultrasound [3]. In total, 13% of these patients had incidentally found nodules and 28.8% of these nodules were cancerous, which is surprisingly high. The majority of other studies show rates of thyroid cancer in nodules ranging from 5 to 15% [4]. The incidence of thyroid nodules increases dramatically with age, approaching 50–60% by the age of 70 years [5]. Autopsy studies and studies of thyroidectomies performed for non-cancer indications reveal a prevalence of PTMC of 4–6% [6]. The high prevalence of these small cancers poses difficult clinical questions: Is this a clinically significant disease? How aggressively should clinicians manage it?

Studies looking at the natural history of thyroid cancers have yielded definitive tumor characteristics that classify them as ‘low risk’, including size less than 1 cm, lack of capsular invasion, lack of distant metastasis and well-differentiated histology [7]. Patient characteristics also factor into the classification of cancer as low-risk, including negative family history, lack of radiation exposure, female gender and age under 45 years [7]. The 10-year disease-specific mortality of all papillary cancers is 7% [8], while mortality from PTMC is less than 1% [9]. Given the low mortality rate of PTMC, the decision to treat it by total thyroidectomy and radioactive iodine (RAI) therapy, as one would a larger papillary cancer, must be weighed against the risks of these aggressive treatments.

The American Thyroid Association guidelines state that these low-risk PTMCs can be managed by hemithyroidectomy alone, with completion thyroidectomy reserved for those patients with any features of local invasion, lymph node involvement or unfavorable histology [10,11]. Controversy surrounding this guideline comes from arguments that these low-risk PTMCs are still clinically significant papillary carcinomas when their natural history is observed and should, therefore, be treated more aggressively.

One study comparing PTMCs with papillary thyroid macrocarcinomas (PTCs) shows equal rates of nodal metastasis, vascular invasion and multifocality and unfavorable cytology regardless of size, suggesting that size below 1 cm does not predict a benign disease course [12]. Another study of 484 patients with papillary thyroid cancer (21.1% with PTMC and 79% with PTC) observed the response to treatment by total thyroidectomy and RAI over 15 years [13]. Over that time period, only one death in the PTC group occurred and 12.2% of the PTMC group had a positive thyroglobulin (a marker of regrowth of thyroid tissue indicating a possible return of malignancy) at the end of the study compared with 24.0% of the PTC group, which was not statistically significant. The researchers noted that, similar to the previous study, the subcentimeter size did not seem to predict the disease course in papillary cancer, as similar rates of vascular invasion and nodal metastasis were seen in both groups. They even went on to recommend total thyroidectomy followed by RAI in all papillary thyroid cancer patients, even if there is a low risk based on the size, pointing to the successful low rates of disease recurrence found in the study. An accompanying editorial to this study by Michael Tuttle from The Memorial Sloan Kettering Hospital (NY, USA) pointed out that the survival for PTMC is over 99%, and the aforementioned study did not offer evidence that aggressive treatment can increase that survival rate [6].

Therefore, what remains unclear from these studies pointing toward the possible aggressive nature of PTMC, is what to do if the patient has a microcarcinoma and low-risk factors from the beginning. Such a case frequently presents as a patient with a subcentimeter papillary carcinoma incidentally discovered with no evidence of nodal metastasis, capsular invasion or vascular invasion. In the aforementioned study, the authors contend that even these cancers can behave aggressively; however, this risk has not been quantified or demonstrated, nor has aggressive therapy demonstrated a benefit.

The use of RAI to treat thyroid cancer has been a common practice for decades. The goal of this therapy is to destroy any residual thyroid tissue after thyroidectomy: both any potential micrometatastasis of malignant tissue as well as normal thyroid remnant tissue, which then allows the use of thyroglobulin protein for surveillance of recurrent disease. After Ernest Mazzaferri’s landmark study in 1977 convinced the world that RAI use in patients with nodal and metastatic disease increased survival [14], the use of RAI markedly increased in all thyroid cancer patients, regardless of evidence of metastatic disease; however, its use at the Mayo clinic in patients with local disease has decreased from 61% in the mid-1980s to 40% in 2000 [15]. This is owing to a lack of strong evidence showing a benefit of RAI in such low-risk patients. In fact, in a recent review and analysis of all observational studies looking at the use of RAI in well-differentiated papillary thyroid carcinomas, no benefit was found in terms of tumor recurrence or mortality [16]. There have been no randomized trials looking at mortality benefit with RAI use in this population, mainly because such a study would be nearly impossible with tens of thousands of patients needing to be followed for decades given the low mortality rates of these cancers [17].

In an attempt to observe the natural course of PTMC and to assess the safety of ‘watchful waiting’, patients with PTMC were offered this option instead of thyroidectomy. In total, 78% of the patients opted for surgery when given the alternative choice of watchful waiting [18]. The markedly high preference for surgery over watchful waiting suggests that most patients, once informed that they have a cancer, are not comfortable with leaving that cancer in place, even if they are quoted 99% survival rates. In those patients that did undergo surgery, 50.5% were discovered to have nodal metastasis. This large number could be construed as worrisome, as they are now classified as having more advanced disease; however, it should be noted that in those patients that chose watchful waiting, 70% had stable tumor size over 5 years and no patients in either group died. It would seem that since nodal metastases were present in so many of the patients who chose surgery, metastases should have been present in an equal number in those who chose watchful waiting, and yet such a presence of metastasis did not impact mortality or disease progression. This fact raises the question of what impact metastatic disease from PTMC has on mortality.

Treatment with RAI is not benign, so routine use as described previously should be evaluated carefully. In addition to commonly quoted side effects, such as sialoadenitis, other consequences of treatment are emerging. One study showed that, in elderly patients, doses of RAI routinely given to treat carcinomas exceeded the maximum tolerable activity in the blood as much as 22–38% of the time [19]. An absolute increased risk of 14.4% for secondary malignancy in those who received RAI has been identified, including soft tissue, colon and salivary gland cancers [20].

Another consideration is the focality of the PTMC within the gland. Multifocal PTMC is associated with increased risk of metastasis and persistent disease, and these multiple small cancers often represent de novo tumors based on clonal genetic analysis [21,22]. A higher number of genetic variants of tumor within a single gland increases the chances of malignant potential as the variety of mutations increase, which draws attention to the importance of genetic markers in determining the disease potential of PTMC.

The future of identifying true low-risk carcinomas may develop from studying tumor gene markers. For example, the BRAF gene mutation leads to an increased MAPK pathway that results in more rapid tumor growth, and its presence is associated with higher malignant potential and poorer patient outcome [23]. Other pathways of tumor suppression and apoptosis are being investigated in PTC and may lead to additional genetic markers that would help resolve the issue of how aggressively to treat small cancers [24].

Currently, we have no solid evidence that PTMC confined within the thyroid poses such a threat to a patient to warrant conventional thyroid cancer therapy of total thyroidectomy followed by RAI. The job of the clinician is not only to treat disease but offer comfort and support to the patient. If the clinician can identify when a thyroid cancer is truly low risk, sparing the patient the anxiety and stress of these treatments, as well as a diagnosis of ‘cancer’, this may be the most important intervention that the clinician can offer that patient.

Financial & competing interests disclosure

The authors have no 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. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

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