Low risk papillary thyroid cancerBMJ 2014; 348 doi: https://doi.org/10.1136/bmj.g3045 (Published 16 June 2014) Cite this as: BMJ 2014;348:g3045
- Juan P Brito, assistant professor, endocrine fellow, and healthcare delivery scholar12,
- Ian D Hay, professor of medicine and thyroidologist1,
- John C Morris, professor of medicine, thyroidologist1
- 1Division of Endocrinology, Diabetes, Metabolism, and Nutrition, Mayo Clinic, Rochester, MN, USA
- 2Knowledge and Evaluation Research Unit, Mayo Clinic, Rochester, MN 55905, USA
- Correspondence to: J C Morris
Thyroid cancer is one of the fastest growing diagnoses; more cases of thyroid cancer are found every year than all leukemias and cancers of the liver, pancreas, and stomach. Most of these incident cases are papillary in origin and are both small and localized. Patients with these small localized papillary thyroid cancers have a 99% survival rate at 20 years. In view of the excellent prognosis of these tumors, they have been denoted as low risk. The incidence of these low risk thyroid cancers is growing, probably because of the use of imaging technologies capable of exposing a large reservoir of subclinical disease. Despite their excellent prognosis, these subclinical low risk cancers are often treated aggressively. Although surgery is traditionally viewed as the cornerstone treatment for these tumors, there is less agreement about the extent of surgery (lobectomy v near total thyroidectomy) and whether prophylactic central neck dissection for removal of lymph nodes is needed. Many of these tumors are treated with radioactive iodine ablation and thyrotropin suppressive therapy, which—although effective for more aggressive forms of thyroid cancer—have not been shown to be of benefit in the management of these lesions. This review offers an evidence based approach to managing low risk papillary thyroid cancer. It also looks at the future of promising alternative surgical techniques, non-surgical minimally localized invasive therapies (ethanol ablation and laser ablation), and active surveillance, all of which form part of a more individualized treatment approach for low risk papillary thyroid tumors.
Thyroid nodules are common. Depending on the population studied and the method of detection used, the prevalence of thyroid nodules varies from 5% by palpation to 30-67% by ultrasound evaluation.1 2 3 Although most of these thyroid nodules are benign, 5-20% are malignant.4 Therefore, thyroid cancer could be common in the population. This notion is supported by cadaveric studies from Finland, where a third of patients who died from non-thyroid related conditions were found to have thyroid cancer.5
The large reservoir of subclinical thyroid cancer has become more evident with the use of imaging technology. Thyroid cancer is now one of the fastest growing diagnoses; more cases of thyroid cancer are found every year in United States than all leukemias and cancers of the liver, pancreas, and stomach.6 Despite its high prevalence, thyroid cancer is an uncommon cause of death. Most patients with these lesions have an excellent prognosis and, because these tumors follow a highly indolent course they have been denoted as “low risk thyroid cancer.”7
Sources and selection criteria
We conducted a comprehensive search of Medline, Embase, the Cochrane Central Register of Controlled Trials, the Cochrane Database of Systematic Reviews, and Scopus from each database’s inception to January 2014. The search strategy was designed and conducted by one of the study investigators (JPB). We used controlled vocabulary supplemented with keywords to search for “low risk thyroid cancer “and “low risk papillary thyroid cancer”. We included observational and randomized studies published in English that included patients with low risk thyroid cancer as well as reviews, meta-analyses, and clinical guidelines for the treatment and management of thyroid cancer. The reference lists from primary studies and narrative reviews were searched, and we consulted with experts in the field to obtain any additional references of importance.
Several organizations and experts have provided guidance on these low risk tumors for both clinicians and patients.8 9 10 11 Owing to uncertainty about the definition, epidemiology, and management of these cancers, many patients receive similar care to that for more aggressive thyroid cancers. New evidence has led to a better understanding of this condition and may herald a revolution in its management. Here, we review the available evidence and current challenges, and we provide a future perspective on the diagnosis and management of low risk thyroid cancer.
The most important predictor of prognosis for thyroid cancer is the histology of the primary tumor. Papillary and follicular thyroid cancers are differentiated thyroid cancers derived from follicular cells and represent 90% of all thyroid cancers.8 Papillary thyroid cancer has a favorable prognosis, with a mortality of 1-2% at 20 years.12 By contrast, follicular thyroid cancer is associated with a mortality of 10-20% at 20 years.8 Other thyroid cancers—medullary, anaplastic, and poorly differentiated thyroid cancer—have an even worse prognosis. Patients with medullary thyroid cancer have a 25-50% mortality at 10 years,8 and most patients with poorly differentiated and anaplastic thyroid cancer die within a few years (five year mortality of 90%).13 Therefore, by definition, low risk thyroid cancer refers to papillary thyroid cancer only.
Low risk papillary thyroid cancers are generally not associated with well recognized predictors of mortality, such as higher grade and aggressive phenotype, local invasion, or distant metastatic disease.14 15 Several classification systems have been developed that include these features and are used to stratify patients into risk categories.16
Figure 1⇓ shows the features that denote low risk papillary thyroid cancer for each of the classification systems. These prognostic scoring and staging systems require the final histological interpretation and, in the case of MACIS, the assessment of residual disease after primary surgical resection.12 Using these systems, 80-85% of papillary cancers are classified as low risk.12 8 16 Although these scores consistently report an excellent prognosis for those classified as low risk (99% at 20 years), they are not designed to predict tumor recurrence.
In its clinical practice guidelines for differentiated thyroid carcinoma, the American Thyroid Association (ATA) suggested a three level classification scheme for predicting recurrence, which requires a dynamic assessment of the risk of recurrence and death over the clinical course of the thyroid cancer.8 The ATA panel characterized low risk thyroid cancers as:
Lesions with no regional or distant metastasis or extra-thyroidal tumor invasion
Absence of histology that is associated with aggressive papillary thyroid cancers, such as tall cell, insular, or columnar cell carcinoma
Resection of all macroscopic tumor (assessed from surgical report or, if conducted, from whole body radioactive iodine scan).
The guidelines also suggested that, if radioactive iodine is given, low risk thyroid cancer should have no uptake of iodine-131 outside the thyroid bed on the first whole body radioactive iodine scan after treatment.
Similar recurrence stratification systems have been proposed by other thyroid societies, such as the Latin American Thyroid Society.20
Delayed risk stratification
Finally, many have argued that the low risk categorization should also consider the effect of initial treatment and require patients to be re-stratified according to the results of the first medical visit (8-12 months). This strategy, called delayed risk stratification, aims to classify patients who are falsely staged intermediate or high risk more accurately. A retrospective analysis that evaluated the predictive value of this approach found that about 50% of intermediate-high risk patients were re-categorised as low risk after the first visit.21
This approach was further validated by a recent retrospective analysis, where delayed risk stratification predicted recurrence in a cohort of patients in whom the presence of antibodies against thyroglobulin could interfere with accurate assessment of this important tumor marker. This is an important finding, because 25% of patients with well differentiated thyroid cancer have anti-thyroglobulin antibodies.22 Although these staging strategies are useful for clinicians and patients to decide postoperative management (adjunctive therapy, frequency, and intensity of follow-up) they are based on conventional clinicopathologic criteria. Ideally, low risk thyroid cancer should be identified before moving to treatment options, particularly before definitive surgery.
Molecular based markers have the potential to improve the diagnosis of thyroid nodules and the risk stratification of thyroid cancers. Two intracellular pathways have been identified that play a role in thyroid cancer—the MAPK (mitogen activated protein kinase) and PI3K-AKT-MTOR (phosphatidylinositide 3-kinase-protein kinase B-mammalian target of rapamycin) pathways. Aberrant activation of the MAPK pathway results in tumor promotion, whereas mutations in the PI3K-AKT-MTOR pathway decrease expression of tumor suppressor genes.23
One mutation, the T1799A BRAF mutation in the MAPK pathway, has been studied as a prognostic molecular marker for aggressive clinicopathological outcomes. A recent meta-analysis of 2470 patients with papillary thyroid cancer found that this mutation was associated with an increased risk of tumor recurrence (relative risk 1.93, 95% confidence interval 1.61 to 2.32), lymph node metastasis (1.32, 1.20 to 1.45), extrathyroidal extension (1.71, 1.50 to 1.94), and advanced stage thyroid cancer (1.70, 1.45 to 1.99).24 The meta-analysis also found a null effect on distant metastasis (0.95, 0.63 to 1.44).24
In a large retrospective multicenter study of 1890 patients,25 the BRAF mutation was associated with a significantly increased cancer related mortality (5.3%, 3.9% to 7.1% in BRAF positive patients v 1.1%, 0.5% to 2.0% in BRAF negative patients (P<0.01), with a median follow-up of 33 months (interquartile range 13-67).
However, the association was no longer significant after adjusting for clinical and histopathological features, and the tumors in most patients with BRAF mutations still behaved in a low risk manner—only 95% of patients with a BRAF mutation had a death that was related to papillary thyroid cancer.25 Similarly, recent large retrospective cohorts comprising 429 and 766 patients with papillary thyroid cancer found no association between a BRAF mutation and tumor multicentricity, lymphovascular invasion, extranodal extension, central neck involvement, advanced stage (III-IV), distant metastasis, or cause specific survival.26 27
These results raise the question of whether the presence of this mutation provides any prognostic value over the existing clinical staging systems for low risk thyroid cancer.28 Other markers are being explored. For example, mutations in the gene encoding the telomerase promoter (TERT), which can lead to persistent telomere lengthening, are indicators of clinically aggressive thyroid tumors and correlate with worse disease specific mortality. In a retrospective study of 647 patients with thyroid tumors,29 TERT mutations were associated with a high risk of disease specific mortality in patients with papillary thyroid cancer (hazard ratio 23.8, 1.3 to 415) compared with those without the mutation. No TERT mutations were found in tumors less than 1 cm. Although this and other promising tumors markers (such as micro-RNA markers and epigenetic changes in tumor genes) could potentially be translated into practice to help define and differentiate low risk from high risk thyroid cancer, clinicopathological features are currently the best way to predict mortality and recurrence.
Thyroid cancer is the most common endocrine cancer and accounts for about 2% of all cancers in women and men.30 Over the past three decades, the incidence of thyroid cancer has risen worldwide (fig 2⇓).31 The surge of new cases is growing more rapidly in countries where healthcare expenditure is driven by the private rather than the public sector, which have a low proportion of public health expenditure and a high proportion of private health financing.32 For example, in the US the incidence of well differentiated thyroid cancer increased from 4.9 per 100 000 to 14.3 per 100 000 between 1975 and 2009.33 Similarly, in South Korea, a country that relies heavily on patients’ direct payment supplemented by private insurance,28 32 the incidence of thyroid cancer increased from 10.6 per 100 000 per year in 1996 to 111.3 per 100 000 per year in 2010 in women, making thyroid cancer the most common cause of cancer among South Korean women.34
The rate of increase in the incidence of thyroid cancers varies across countries. In Europe, for instance, the incidence of thyroid cancer in France resembles that seen in the US (3.4/100 000 per year in 1983 to 11.7/100 000 per year in 2002), the incidence in Sweden has not changed as dramatically (2.4/100 000 per year in 1958 to 3.5/100 000 per year in 2002).35
Geographic variation is also seen within countries. A recent report from Belgium showed that, between 2004 and 2006, the incidence of thyroid cancer was lower in the northern than in the southern part of the country (4.1/100 000 v 8.3/100 000).36 Interestingly, the worldwide rise in the incidence of thyroid cancer has not been accompanied by an increase in mortality from this cancer, suggesting that these increasingly common incident cases represent a mild form of the disease that rarely causes death.
In countries where further analysis based on subtype and size of the tumor has been conducted, low risk thyroid cancer accounts for 90% of incident cases.37 It has been suggested that the explosion of new cases of thyroid cancer, especially low risk papillary thyroid cancer, is due to the development and use of imaging technologies capable of exposing a large reservoir of subclinical disease.38 In support of this hypothesis, many autopsy studies have shown that thyroid cancer is common among people who die from non-thyroid related causes. Two studies, one in Spain and one in Finland,5 analyzed the entire gland histologically.5 39 They concluded that the frequency of thyroid cancer was 24% and 36%, respectively, suggesting the presence of a large reservoir of undiagnosed, and arguably clinically unimportant, disease.
Although it is not clear whether this pool of subclinical disease has changed over time, our capacity to detect it has changed. Increased availability of sensitive imaging technology may have increased the diagnosis of “occult” thyroid cancers. A recent retrospective study found that, among incident cases of thyroid cancer, 39% were detected through the use of imaging (mainly neck sonography and computer tomography), 15% were detected by pathological studies, and the remaining 46% were detected by palpation.40 At least half of cases detected incidentally through imaging could have been labeled as low risk cancers. This is supported by a recent analysis of the US Surveillance, Epidemiology, and End Results Program (SEER) database,41 which showed that the incidence of thyroid cancer was positively associated with markers of access to these technologies, such as college education, white collar employment, and higher family income. Incidence was negatively correlated with being uninsured, in poverty, unemployed, of non-white ethnicity, non-English speaking, and lacking high school education.
It has been suggested that the increased frequency of thyroid cancer might also reflect a new risk factor, not yet identified, that is causing the surge of incidence without affecting mortality. Candidates for these risk factors are increased exposure to low dose ionizing radiation from radiographic imaging as well as hormonal or nutritional factors.42 43 44 The association of these risk factors with low risk thyroid cancer is weak and inconsistent, and no causal pathways have been described to link them to the increased incidence of thyroid cancer. Although other novel risk factors may play a role in the surge of new thyroid cancer cases, it is unlikely that they can explain the magnitude and geographic distribution of this surge worldwide.
Treatment options for low risk thyroid cancer
Traditionally, the management of low risk papillary thyroid cancer involved removal of the primary tumor. Patients undergo preoperative assessment including neck ultrasound to map neck lymph nodes and fine needle aspiration biopsy of suspicious nodes. In those patients identified as having localized disease by this preoperative assessment, the most common treatment is surgical thyroidectomy. This permits removal of the primary tumor, facilitates postoperative treatment, and enables accurate staging and follow-up.8 Although it is generally agreed that surgery is fundamental to the management of these tumors, there is less agreement about the extent of the surgery (lobectomy v near total thyroidectomy) and whether prophylactic central neck dissection for removal of lymph nodes is needed.
No randomized trials have investigated the advantages of total thyroidectomy over lobectomy for patients with low risk papillary thyroid cancer. Thyroidectomy may remove the entire thyroid (total thyroidectomy) or may consist of ipsilateral total lobectomy and contralateral subtotal lobectomy, with only a small amount of thyroid tissue left to safeguard parathyroid function (near total thyroidectomy).
The rationale for thyroidectomy in these patients is that there are fewer recurrences with this intervention than with lobectomy. This argument is supported by a population based study from the National Cancer Data Base.45 The study population comprised 52 173 patients undergoing thyroid surgery for papillary thyroid cancer. For tumors 1 cm or more, lobectomy resulted in a higher risk of recurrence (hazard ratio 1.15, 1.02 to 1.3) and non-disease specific mortality (1.31, 1.07 to 1.6). In view of this study and other benefits of thyroidectomy (possible future treatment with radioactive iodine and facilitation of follow-up with thyroglobulin), the ATA recommended that for patients with tumors greater than 1 cm, the initial surgical procedure should be a near total or total thyroidectomy unless there are contraindications to this surgery.8 Other thyroid guidelines provide similar recommendations.8 9 10 11
Prophylactic central node dissection
The inclusion of prophylactic central node dissection (PCND)—resection of level VI, central compartment cervical lymph nodes in a patient with no evidence of lymph node involvement on physical examination, preoperative imaging, or intraoperative imaging8—during initial surgery is controversial.46 This approach differs from the standard practice of performing central compartment dissection for visually or palpable lymph nodes seen at surgery, or seen preoperatively on preoperative ultrasound.
The argument supporting PCND is that nodal metastasis, which may be invisible on preoperative imaging (such as neck ultrasound), may be correlated with the persistence and recurrence of papillary thyroid cancer and that tumors often metastasize to cervical lymph nodes.47 Retrospective studies of patients with tumors less than 1 cm who underwent node dissection showed the presence of microscopic nodal disease in 12-60% of patients.48 49 50 In addition, proponents of PCND argue that this surgical intervention may also facilitate proper staging and postoperative tumor marker follow-up because it improves the assessment and removal of occult metastases.51
The argument against PCND is based on the uncertainty about its benefits for recurrence and mortality. A meta-analysis of retrospective studies,52 comprising 1264 patients undergoing thyroidectomy or PCND, showed no difference in the risk of recurrence of thyroid cancer 1.05 (0.48 to 2.31) between the two groups. These results conflict with the results of a more recent meta-analysis of retrospective studies comprising 3331 patients.53 In this systematic review, PCND was associated with a lower risk of short term (less than five years) locoregional recurrence than thyroidectomy alone (4.7% v 8.6%; pooled incidence rate ratio 0.65, 0.48 to 0.86). However, these results are confounded by the patients in the PCND group being 2.6 times more likely to receive radioactive iodine than those in the thyroidectomy group.
In addition, a retrospective study with a longer follow-up period evaluated disease specific survival for patients who underwent PCND for low risk thyroid cancer against historical cohorts without PCND. It showed that in 13 years of follow-up the rate of disease specific mortality was similar for both groups at 1.9%.54 These findings are supported by a similar retrospective Norwegian study and a prospective Korean study.55 56 This evidence suggests that PCND might not change the incidence of recurrence or disease specific mortality for patients with low risk papillary thyroid cancer.
The most salient argument against routine application of PCND is that its performance may increase the risk of complications of thyroidectomy, such as recurrent laryngeal nerve injury and hypoparathyroidism, especially if the surgeon is relatively inexperienced in the procedure.46
The role of surgery for thyroid cancer is to clear the neck of disease. Some argue that this goal is achievable by a more conservative procedure—lobectomy. This intervention differs from a bilateral lobar resection (near total or total thyroidectomy) in that a hemi-thyroidectomy is performed, which removes only the lobe of the thyroid containing the nodule that harbors the malignant cells (determined by fine needle aspiration biopsy before surgery).
A recent large observational study based on the SEER database, analyzed 22 724 patients who had undergone surgery for papillary thyroid cancer between 1988 and 2001.57 Controlling for tumor size, no survival benefit was seen with more aggressive surgical treatment for those patients with low risk tumors (total thyroidectomy v lobectomy). The lack of difference in survival between these two interventions is backed up by an earlier institutional retrospective study of 1038 patients and a database analysis of 53 856 patients in the National Cancer Data Base from 1985 to 1995.58 59 The ATA recommends that thyroid lobectomy alone may be sufficient treatment for small (<1 cm), low risk, unifocal, intrathyroidal thyroid cancers in the absence of previous head and neck irradiation or radiologically or clinically involved cervical nodal metastasis.8
There is much uncertainty about the benefits of these surgical procedures for low risk papillary thyroid cancer. Total thyroidectomy and PCND could facilitate follow-up and staging of low risk tumors and might decrease the need for further intervention and the inherent anxiety associated with such procedures. However, patients with low risk tumors who undergo more invasive surgical procedures do not live longer than those who undergo lobectomy. It is likely that, because of the indolent course of low risk thyroid cancer, these two procedures are equivalent—it is difficult to show benefit of any intervention when the risk of mortality is close to 1-2% after 20 years of follow-up. However, all patients who have thyroidectomy need thyroid replacement (with its own burden of treatment and follow-up), compared with only half of patients who have lobectomy.60 Finally, the morbidity associated with surgical intervention is directly and inversely correlated with the extent of surgical intervention and the surgeon’s experience (fig 3⇓).
Postoperative radioiodine remnant ablation
Radioiodine remnant ablation (RRA) aims to destroy the residual normal thyroid tissue seen after thyroidectomy. It differs from radioiodine “treatment,” which is meant to destroy known residual thyroid cancer. However, it is often considered adjuvant therapy in cases where the risk of recurrence and mortality is a concern.8
Although radioactive iodine was first used to treat metastatic thyroid cancer in 1940,64 it was not until 1960 that it was used to ablate postoperative remnant tissue.65 By the 1990s, administration of RRA had become the standard of care for patients with thyroid cancer. The goals of RRA were, initially, to destroy microscopic residual disease and, more recently, to allow accurate surveillance with tumor markers (serum thyroglobulin measurements).
Nevertheless, over the past two decades evidence has emerged of a lack of any significant benefit of RRA for low risk papillary thyroid cancer. One of the first studies to suggest this lack of benefit was a large retrospective cohort of 1163 patients with a MACIS score less than 6 (low risk papillary thyroid cancer). These patients were treated between 1970 and 2000 and underwent near total or total thyroidectomy performed by a small group of specialized surgeons. After a follow-up of 20 years, in both the 527 node positive and the 636 node negative cases, cause specific mortality and tumor recurrence were the same in those who had surgery without postoperative RRA and in those who received postoperative RRA.66 67 Furthermore, two systematic reviews investigating the effectiveness of radioactive iodine for treatment of patients with thyroid cancer found unclear benefits of RRA for low risk patients with papillary thyroid cancer and no statistically valid improvement in mortality or disease specific survival.68 69
Finally, RRA in low risk papillary thyroid cancer did not reduce recurrence in multifocal node positive disease.70 In addition, a recent large prospective multicenter study with a follow-up of 10.3 years found similar overall survival (95.8% v 94.6%) and disease-free survival (hazard ratio 0.73, 0.43 to 1.25) in patients with low risk thyroid cancer who received RRA after surgery versus those who had surgery alone.71
Despite this lack of benefit for RRA in patients with low risk thyroid cancer, its use has increased. In the US, the use of this technique in these patients increased from four in 100 patients to 38 in 100 patients between 1973 and 2007.72 This finding is supported by a time trend analysis of 89 219 patients with well differentiated thyroid cancer treated at 981 hospitals associated with the US National Cancer Data Base between 1990 and 2008.73 In this analysis the increased use of RRA for low risk thyroid cancer varied between hospitals, and this was attributed to unexplained hospital characteristics, suggesting considerable variation in practice for these patients. This use of RRA for low risk patients was perhaps driven by the belief that this treatment improves the specificity of postoperative thyroglobulin assays, which may detect persistent or recurrent disease.
Enthusiasm for the use of RRA to facilitate follow-up was tempered by a recent retrospective study of 290 consecutive patients with low risk thyroid cancer treated with thyroidectomy alone who underwent yearly follow-up with serum thyroglobulin assays.74 It found that after about five years, thyroglobulin was undetectable in 95% of patients not treated with RRA compared with 99% in a matched cohort of patients treated with RRA. Only one patient who was not treated with RRA had a recurrence, as detected by an increase in thyroglobulin. This indicates that thyroglobulin continues to be a reliable marker of follow-up for patients with low risk thyroid cancer who did not receive RRA.
The administration of radioactive iodine is not risk free—it is associated with short term side effects (dry eyes, altered taste, and inflammation of salivary glands) and perhaps an increase in salivary gland cancers and leukemia.72 In view of this evidence, recent guidelines do not recommend the use of RRA in patients with low risk thyroid cancer with lesions less than 1 cm who do not have high risk features (such as aggressive variants and vascular invasion). They recommend that it is used only in those with documented lymph node metastasis or other high risk features.8 10
Thyrotropin suppressive therapy
Similar to normal follicular cells, thyroid cancer cells express thyrotropin receptors and respond to this hormone by increasing cell growth.75 Therefore, postoperative thyrotropin suppressive therapy has played a major role in the treatment of well differentiated thyroid cancer since the first report of regression of papillary thyroid cancer in two patients given thyroid extracts in 1937.76
In support of this practice, a meta-analysis of 4174 patients with well differentiated thyroid cancer showed that thyrotropin suppressive therapy reduced the risk of disease progression or recurrence and death (relative risk 0.73, 0.6 to 0.88).77 Although it is generally agreed that patients with high risk thyroid cancers should receive suppressive therapy to maintain thyrotropin concentrations below 0.1 mU/L,78 79 such treatment did not improve the rate of recurrence or disease specific survival in patients with low risk papillary thyroid cancer.80
Therefore, the goal of thyroid hormone treatment should be different for patients with low risk papillary thyroid cancer than for those with indeterminate or high risk disease. For low risk patients, post-thyroidectomy thyrotropin suppressive therapy should provide adequate replacement to avoid symptoms of hypothyroidism and thyrotoxicosis. Older patients, who are more likely to develop cardiovascular disease (atrial fibrillation and arrhythmia) and bone loss as a result of subclinical thyrotoxicosis, should be managed with less tight goals (thyrotropin 1-2.5 mU/L) than younger patients (thyrotropin 0.5-1 mU/L), who are less sensitive to the adverse effects of treatment.81 82 The European and ATA guidelines for patients with low risk thyroid cancer recommend a thyrotropin concentration of 0.5-1 mU/L.8 10
After thyroid surgery, postoperative treatments (RRA and thyrotropin suppressive therapy) have not shown significant clinical benefit for patients with low risk thyroid cancer. However, for some patients, depending on the context and their preferences, even conventional surgery might not be desirable. Over the past decade, other management options have been studied for patients with low risk thyroid cancer.
Alternative surgical techniques
Although open surgery with a residual neck scar is the traditional technique for thyroidectomy, recent studies have reported that endoscopic thyroidectomy is also feasible for patients who wish to avoid a cervical incision. Endoscopic surgery, using extracervical access (for example, axillary, anterior chest, or breast), is an approach to the thyroid gland that uses video-endoscopic equipment through small and hidden incisions. One form of this endoscopic technique is robotic thyroidectomy, where the surgeon is assisted by an advanced robotic system that may allow better visualization and dexterity.
A meta-analysis of randomized clinical trials that compared video assisted interventions with conventional thyroidectomy in thyroid nodules and low risk papillary thyroid cancer found no significant difference in the risk of transient laryngeal nerve palsy or hypoparathyroidism.83 In addition, video assisted techniques reduced postoperative pain at six hours and improved cosmetic results, although operative times were longer. Of note, the robotic technique has not received US Food and Drug Administration approval for thyroidectomy because of concerns regarding unreported adverse effects.84
Minimally invasive therapy for patients with low risk thyroid cancer
When surgery is neither indicated nor desired, non-surgical minimally invasive therapies are available. These treatments are based on the principle that focused and accurate destruction of tumor tissue will induce small vessel thrombosis and coagulative necrosis within the tumor, and they have been proposed by some as excellent options for primary treatment of low risk papillary thyroid cancer.
Ultrasound guided percutaneous ethanol ablation involves the direct intratumoral injection of 95% ethanol under careful ultrasound guidance and local anesthesia. Although this technique is effective and safe (small risk of temporary hoarseness) for benign cystic thyroid nodules and nodal metastasis in papillary thyroid cancer,85 86 limited evidence is available on its efficacy in low risk thyroid cancer. A case series described three patients, with five intrathyroid foci of papillary microcarcinoma, whose biopsy confirmed papillary thyroid cancer lesions were treated by ethanol ablation within the intact thyroid.87 In all cases, the lesions became avascular and smaller and one disappeared. These promising results are limited by a lack of a comparator group and selection bias.
Others, however, have raised concerns that this treatment can cause random distribution of ethanol within the thyroid gland, with possible seepage into surrounding cervical tissues and local side effects.88 89 They propose that thermal ablation with lasers may be a better alternative treatment for thyroid cancer owing to its predictable and well defined area of necrosis.90 A recent study evaluated the efficacy of laser ablation of low risk papillary thyroid cancer as primary treatment in three patients.91 All patients underwent percutaneous laser treatment of thyroid cancer lesions in the operating room immediately before surgical removal of the thyroid. Subsequent pathological and immunohistochemical analyses of the extirpated thyroid glands showed destruction of the malignant cells.
Radiofrequency ablation is another possible non-invasive treatment for primary low risk papillary thyroid cancer. This technique has been used successfully to ablate tumors in the liver, lung, and kidney.92 A recent systematic review of prospective studies found that it is safe and effective in the treatment for symptomatic benign thyroid nodules and is effective in treating locoregional recurrence of thyroid cancers.92 93 However, no studies have yet been reported for primary treatment of low risk thyroid cancer.
Two large Japanese observational studies of 1465 patients with thyroid cancer were conducted on the basis of the hypothesis that most low risk papillary thyroid cancers do not need immediate or eventual thyroid surgery.94 95 Patients were offered the option of active surveillance or thyroidectomy. Those who chose active surveillance were followed closely with neck ultrasound at six months and then annually for an average of five years (range 1-19). By the end of the follow-up only a minority of patients had lymph node metastasis (<2%) or experienced asymptomatic lesion growth (5%). No cases of disease specific mortality were seen in the observed patients. None of the other traditional risk factors for lymph node metastasis (multicentricity or size at diagnosis) were linked to any adverse outcomes.
Following the example of these Japanese studies, the Memorial Sloan-Kettering Cancer Center in New York City has begun an active surveillance program for patients with low risk papillary thyroid cancer who do not wish to proceed with surgery.96 The table⇓ presents the most important characteristics and finding of these cohorts.
Active surveillance protocols have also been developed for recurrent localized disease after thyroidectomy. In a retrospective study of 166 patients with low risk thyroid cancer with at least one abnormal lymph node after thyroidectomy, the median size of lymph nodes at the start of the observation period was 1.3 cm (range 0.5-2.7).97 After a median follow-up of 3.5 years, growth of at least 3 mm and 5 mm was seen in 33 (20%) and 15 (9%), respectively, whereas the nodules disappeared in 23 (14%). No local complications or disease related mortality were seen. A similar study from the same research group concluded that, when observed over a period of more than five years, only 10% of recurrent thyroid bed nodules from patients with low risk thyroid cancer increased substantially in size.98 These two studies support the hypothesis that most low risk thyroid cancer lesions follow an indolent course and that many can be monitored safely without active intervention.
Patient centered and evidence based approaches
Despite their excellent prognosis patients with low risk papillary thyroid cancer still receive similar treatment to patients with aggressive disease and a worse prognosis. Perhaps the continued use of unnecessary treatments reflects uncertainty in the definition of “low risk thyroid cancer,” which focuses on the lesions and not the patients.
Ideally, low risk papillary thyroid cancer should be managed in a way that achieves the lowest risk of mortality and morbidity with the lowest burden of treatment. For instance many argue that RRA facilitates disease related surveillance, but it clearly does not improve mortality and probably increases treatment related morbidity. Thus, this treatment might not be appropriate for a patient with low risk thyroid cancer.
Often, however, the balance between benefits (mortality and morbidity) and burden of treatment is not as clear as it is for surgical interventions (thyroidectomy v lobectomy). In such scenarios it is important to understand what the patient wants and needs, and what is most appropriate in the individual context.
Some patients might even opt for active surveillance because for them a surgical intervention brings more burden than benefit; fig 4⇓ shows a decision tool to help clinicians review these management options with their patients. Many others may, in the future, decide for minimally invasive treatments (such as ultrasound guided percutaneous ethanol ablation or laser ablation) as these therapies become better understood and are more widely practiced.
To help patients make better decisions we should combine efforts to bring them the best comparative evidence about the efficacy and potential harm caused by each intervention. In low risk cancers with a low rate of events, a thyroid cancer network that helps create large databases for observational studies and samples for randomized clinical trials would be useful. Perhaps such a network might follow the example of pediatric cancer research, where a network sponsored by the US National Cancer Institute supports the enrollment of children diagnosed as having cancer to research registries. These networks facilitate the understanding and translation of evidence that might yield improvements in practice.99
Finally, we should explore the possible causes of this increase in the incidence of low risk papillary thyroid cancer. We have suggested that it probably reflects the increased use of imaging technologies. If this is the case, more research is needed to recognize which imaging techniques are more likely to find incidental small papillary thyroid cancers and investigate the appropriateness and usefulness of these tests.
The diagnosis of thyroid cancer is rapidly increasing. This increase is occurring worldwide, but important geographic variations exist. Most of the new cases are small and localized papillary thyroid cancers that, because of their indolent course, are considered low risk. The main reason for this increase in the incidence of low risk papillary thyroid cancer is still unclear, but it probably reflects the discovery of a large reservoir of subclinical disease through the increased use of imaging.
Although clinicians are encountering patients with low risk thyroid cancer more often, guidelines and experts have not reached a consensus on the precise definition of this condition. This uncertainty in diagnosis is aggravated by the lack of high quality evidence from randomized clinical trials to elucidate the extent of the benefits and harms of currently available treatments. This lack of clarity has led many patients to receive treatments that are more appropriate for aggressive cancers. Many of these treatments can lead to adverse effects that might be perceived as unnecessarily harmful owing to the favorable prognosis of these lesions.
The challenges surrounding the diagnosis and management of low risk thyroid cancer should be seen as an opportunity to start collaborative networks around the globe that provide patients with faster and more reliable evidence about traditional and novel treatments. The outcomes of the trials should incorporate factors that help clinicians assess the burden of disease and also help them understand the burden of treatment and the outcomes that matter to patients. The era of a “one size fits all” management program for papillary thyroid cancer should end. Each patient is different and deserves a more individualized treatment approach.
Future research questions
What is the contribution of imaging tests to the increased incidence of low risk thyroid cancer?
Does any other risk factor explain the increased incidence of low risk thyroid cancer?
What is the role of promising molecular markers in predicting recurrence and mortality in patients with low risk thyroid cancer?
What are patients’ values and preferences regarding the treatment decision for low risk thyroid cancer?
What is the comparative effectiveness of active surveillance versus traditional management versus minimally invasive treatments for patients with low risk thyroid cancer?
What is the role of shared decision making for patients with low risk thyroid cancer?
Cite this as: BMJ 2014;348:g3045
Contributors: All authors contributed to the concepts and structure of this manuscript. JCM is guarantor.
Competing interests: We have read and understood BMJ policy on declaration of interests and declare the following interests: None.
Provenance and peer review: Commissioned; externally peer reviewed.