Current Controversies in Low-Risk Differentiated Thyroid Cancer: Reducing Overtreatment in an Era of Overdiagnosis

Abstract

Context

Low-risk differentiated thyroid cancer (DTC) is overdiagnosed, but true incidence has increased as well. Owing to its excellent prognosis with low morbidity and mortality, balancing treatment risks with risks of disease progression can be challenging, leading to several areas of controversy.

Evidence Acquisition

This mini-review is an overview of controversies and difficult decisions around the management of all stages of low-risk DTC, from diagnosis through treatment and follow-up. In particular, overdiagnosis, active surveillance vs surgery, extent of surgery, radioactive iodine (RAI) treatment, thyrotropin suppression, and postoperative surveillance are discussed.

Evidence Synthesis

Recommendations regarding the diagnosis of DTC, the extent of treatment for low-risk DTC patients, and the intensity of posttreatment follow-up have all changed substantially in the past decade. While overdiagnosis remains a problem, there has been a true increase in incidence as well. Treatment options range from active surveillance of small tumors to total thyroidectomy followed by RAI in select cases. Recommendations for long-term surveillance frequency and duration are similarly broad.

Conclusion

Clinicians and patients must approach each case in a personalized and nuanced fashion to select the appropriate extent of treatment on an individual basis. In areas of evidential equipoise, data regarding patient-centered outcomes may help guide decision-making.

The management of low-risk differentiated thyroid cancers (DTCs) remains controversial at all stages, from diagnosis through treatment and follow-up, mainly because of the indolent nature of the disease and the challenge of balancing morbidity of treatments with the risk of disease progression (1). The definition of “low-risk” thyroid cancer has evolved over time and has varied considerably between studies. The American Thyroid Association (ATA) defines low-risk DTC as intrathyroidal papillary or follicular cancers, without aggressive histology, evidence of local invasion, vascular invasion, or distant metastases, and fewer than 5 cervical nodal metastases (provided all are < 0.2 cm) (2). This mini-review uses the ATA definition of low-risk DTC for tumors 1 to 4 cm in size, unless otherwise specified (Table 1). These cancers have a risk of structural recurrence of approximately 1% to 10%, depending on patient and tumor characteristics, extent of treatment, and length of follow-up (2–4). The most recent edition of the ATA guidelines for the management of DTC made several critical changes to the recommendations for extent of surgery and the use of adjuvant radioactive iodine (RAI) for these patients (5). Following publication of these revised guidelines, more patients are now being treated with lobectomy alone (6–8), and fewer are receiving adjuvant RAI (8, 9), complicating long-term surveillance. Undoubtedly, this trend will affect future recommendations for posttreatment management, including thyrotropin (TSH) suppression, the utility of surveillance imaging, and the role of thyroglobulin (Tg) measurement.

Table 1.

Adapted American Thyroid Association criteria for low-risk differentiated thyroid cancer used in this review (2)

Low-risk differentiated thyroid cancer
Papillary or follicular histology
Tumors ≥ 1.0 cm and < 4.0 cm in greatest dimension
Entirely contained within the thyroid gland
Fewer than four foci of vascular invasion (for follicular carcinomas)
No aggressive histology (tall-cell, columnar, or hobnail variants)
Fewer than 5 lymph node micrometastases (all < 0.2 cm in greatest dimension)
No distant metastases

This mini-review will focus on controversies regarding extent of treatment and challenges in the posttreatment surveillance for patients with low-risk DTC. In particular, we highlight studies that investigate patient-centered outcomes, such as quality of life and cost of care, which may help to inform decision-making when the scientific consensus regarding more traditional outcomes (eg, recurrence rates) is ambiguous.

Search Strategy

The MEDLINE database was queried using the PubMed search engine for English-language articles using the term “low-risk thyroid cancer” in combination with the following additional terms: “incidence,” “overdiagnosis,” “active surveillance,” “extent of surgery,” “lobectomy,” “quality of life,” “radioactive iodine,” “TSH suppression,” “post-operative surveillance,” “ultrasound surveillance,” “thyroglobulin,” and “decision making.” Case reports and case series were excluded. Articles were selected for inclusion by the authors based on their relevance and effect (in terms of citation by guideline committees), with preference given to more recent publications. Additional references were selected directly from articles cited in this manuscript and the authors’ knowledge of the literature.

Incidence of Low-Risk Differentiated Thyroid Cancers: Overdiagnosis, True Increase in Incidence, or Both?

The incidence of thyroid cancer, particularly low-risk disease, has risen dramatically in the past several decades (10–13). Between 1974 and 2013, thyroid cancer incidence in the United States tripled, increasing from 4.5 to more than 14 cases per 100 000 people (10, 13). The dramatic increase in incidence is largely attributed to increased detection of indolent disease, that is, overdiagnosis (12, 14). Vaccarella and colleagues (15) estimated that between 1988 and 2007 approximately 470 000 women and 90 000 men had been overdiagnosed with thyroid cancer in 12 industrialized countries alone. In South Korea, by 2011 screening had contributed to an increased thyroid cancer incidence 15 times the rate observed in 1993 (16).

The expanded use of medical imaging, particularly ultrasound, seems to be the primary culprit in overdiagnosis. The connection between expanded utilization of medical imaging and the prevalence of thyroid cancer has been studied since the early 1990s (17). Current estimates are that approximately 16% to 18% of patients who have a computed tomography scan and 20% to 70% of patients who have any cervical ultrasound have incidental thyroid nodules identified (18). A recent analysis of the Surveillance, Epidemiology, and End Results (SEER) Medicare database demonstrated that patients with ultrasound-detected thyroid cancers had improved disease-specific survival, a result of lead-time bias and further evidence of imaging-based overdiagnosis (19). It is therefore not surprising that identification of these thyroid incidentalomas has led to dramatic increases in fine-needle aspiration (FNA) biopsies of thyroid nodules, diagnoses of thyroid cancer, and subsequent surgeries (20).

Autopsy studies provide additional evidence of overdiagnosis, with historical studies estimating occult thyroid cancer prevalence as high as 35% (21, 22). LeClair and colleagues (22) compared papillary thyroid cancer (PTC) prevalence in autopsy studies with disease incidence in men and women in the SEER database. They found that the incidence of small (< 2 cm) PTCs in women in SEER was greater than 4 times the incidence in men; however, in their meta-analysis of autopsy studies, that ratio was closer to 1 to 1. The authors argue that this is additional evidence that small thyroid cancers are overdiagnosed, particularly in women.

Others have suggested that environmental factors and changing demographics (in particular, obesity) have contributed to the increased incidence of thyroid cancer (12). Epidemiologic studies have identified an association between body weight or body mass index and the incidence of thyroid cancer (23–26). Kitahara et al (27) found that adults with overweight (25.0-29.0) and obese body mass index (≥ 30.0) experienced a 1.26- and 1.30-fold increased risk of thyroid cancer, respectively. They estimated that the prevalence of overweight and obesity accounted for 13.6% of new thyroid cancer diagnoses in the United States from 1995 to 2015. Obesity is also associated with more aggressive or advanced thyroid cancers, including larger tumor size, extrathyroidal extension, and persistent or recurrent disease (28–30). Thus, increasing obesity rates have likely contributed to the increased incidence of thyroid cancer over the past several decades.

Epidemiologic evidence also suggests there are true increases in thyroid cancer incidence that cannot be entirely explained by overdiagnosis (12, 13, 31–33). Lim et al (13) found that incidence of thyroid cancer increased for all tumor sizes and American Joint Committee on Cancer stages between 1974 and 2013, including large or advanced tumors. Furthermore, this study demonstrated an overall increase in incidence-based thyroid cancer mortality of approximately 1.1% per year. These results corroborated earlier studies that found a similar pattern of increased incidence for tumors of all sizes and stages in California alone (31), nationally in the United States (33), and abroad (32). Recently, likely due to efforts to curb overdiagnosis, thyroid cancer incidence trends have shifted and are no longer rising (34, 35). From 2014 to 2018, the incidence of thyroid cancers smaller than or equal to 1 cm has declined in the United States, while it has plateaued for tumors larger than 1 cm. Meanwhile, incidence-based mortality has continued to increase (35).

Extent of Treatment

Active Surveillance vs Surgery

In the 1990s, a group of Japanese surgeons at Kuma Hospital began to offer observation alone to their patients with papillary thyroid microcarcinomas (PTmCs), defined as tumors smaller than 1 cm (36). Overall, 70% of patients experienced no change or a decrease in size over a mean of 47 months of follow-up. A subsequent study found that the detection rate of new lymph node metastases during active surveillance was comparable to the recurrence rate following upfront surgery: 2.1% vs 3.0%, respectively (37). Similar results were also documented by a separate Japanese group (38).

Since the publication of these studies, several groups have sought to expand criteria and eligibility for active surveillance to include patients with small low-risk cancers (39–41). Tuttle et al (39) published their experience with active surveillance of 291 low-risk PTC patients, including 59 patients (20.3%) with tumors between 1.0 and 1.5 cm in size. They found that only 11 patients (3.8%) experienced tumor growth greater than or equal to 3 mm; however, they estimated a cumulative incidence of growth (≥ 3 mm) of 24.8% by 5 years of observation. Importantly, initial tumor size (< 1 cm vs 1.0-1.5 cm) was not predictive of future growth (hazard ratio [HR] 0.93; 95% CI, 0.36-2.41). Ho et al (41) demonstrated similar results for patients with tumors as large as 2 cm. Another study from Japan included a series of 61 patients with 1 to 2 cm low-risk PTCs who requested surveillance over upfront surgery (40). Over a mean follow-up of 7.9 years, 4 patients (7%) experienced tumor growth greater than or equal to 3 mm, 7 patients (11%) had growth greater than or equal to 50% by volume, and 2 patients (3%) developed lateral neck nodal metastases. Overall, the 5-year disease progression rate was estimated to be 12%, which was the same for a comparison observation cohort with tumors smaller than 1 cm. In both studies, all patients who crossed over from observation to surgery had no evidence of persistent or recurrent disease during the postoperative surveillance period. In a separate subgroup analysis of the Kuma cohort, only 8 of 496 PTmC patients (1.8%) older than 60 years experienced disease progression while undergoing active surveillance, as compared to 15 of 169 patients (8.9%) younger than 40 years (42). Older patients may be at higher risk of surgical complications, due to comorbid conditions and frailty (43). Therefore, older adults may be the best candidates for active surveillance of low-risk DTC.

To date, the few studies on active surveillance of known DTCs have been nonrandomized, heterogeneous, and generally low quality (39–41, 44). Chou et al (44) performed a systematic review of the available data regarding the relative benefits and harms of active surveillance vs upfront surgery for low-risk DTC. They included 14 studies, only half of which had a comparison surgical cohort, and found that all were high-risk for bias given unclear selection criteria for active surveillance vs surgery, nonblinded analyses, and in many cases clear differences in patient and tumor characteristics between observation and comparison surgical cohorts (44). They concluded that active surveillance and upfront surgery may have similar outcomes, but low-quality data precluded any strong recommendations. Additionally, active surveillance requires substantial patient compliance and clinician diligence in follow-up that may make it untenable in certain situations. Ultimately, while active surveillance may be safe for highly selected patients with low-risk DTC, particularly older adults, more data are needed to make definitive conclusions about the comparative safety and efficacy of active surveillance vs upfront surgery.

Extent of Surgery: Lobectomy vs Total Thyroidectomy

The 2015 ATA guidelines changed recommendations for extent of surgery for patients with low-risk DTC, stating that lobectomy and total thyroidectomy were equally acceptable treatment options (2). A large retrospective analysis of the National Cancer Database by Bilimoria and colleagues (45) included more than 52 000 patients who underwent surgery for PTC from 1985 to 1998 and found that patients treated with thyroid lobectomy appeared to have a 57% increased risk of recurrence and 21% increased risk of death compared to patients who underwent total thyroidectomy. Their subgroup analysis of patients whose tumors were between 1 and 2 cm in size found similar results; lobectomy increased the risk of recurrence and death by 24% and 49%, respectively, compared to total thyroidectomy. However, this study did not adjust for comorbidities, multifocality, extrathyroidal extension, and completeness of resection. Furthermore, recurrence data in National Cancer Database are heterogeneous, often incomplete, and of questionable reliability (46). Adam and colleagues (47) challenged the Bilimoria et al study several years later. Their analyses, including nearly 62 000 patients, adjusted for comorbidities and tumor variables including multifocality, extrathyroidal extension, and completeness of resection. After adjustment, they found no difference in overall survival for patients undergoing lobectomy vs total thyroidectomy (HR 0.96; 95% CI, 0.84-1.09). As DTC has an excellent prognosis, power calculations have demonstrated that a prospective, randomized controlled trial of lobectomy vs total thyroidectomy is impractical and prohibitively expensive (48).

Several retrospective studies have raised concerns of increased recurrence rates following lobectomy for low-risk DTC (49, 50). In a single-center series of 331 patients with low-risk DTC, those treated with total thyroidectomy (291 patients, 87.9%) vs lobectomy (40 patients, 12.1%) had lower recurrence rate: 5% vs 12.5% (49). However, the total thyroidectomy group had smaller tumors (0.9 cm vs 1.4 cm; P = .032), and only one patient in the lobectomy group had a structural recurrence. A single-institution study of almost 9000 PTmC patients from South Korea found a 60% reduction in the risk of locoregional recurrence after total thyroidectomy, although 58% of recurrences in the lobectomy group occurred in the contralateral thyroid lobe (50). When contralateral lobe recurrences were excluded, there was no difference in recurrence rates between total thyroidectomy and lobectomy alone (HR 0.88; P = .640). These studies stand in contrast to other retrospective series (51, 52) and multiple contemporary large database studies that found no difference in recurrence rates or survival following lobectomy vs total thyroidectomy (53–55).

Others have concluded that treating low-risk DTC patients with lobectomy alone will increase the need for completion thyroidectomies to allow for adjuvant RAI (56–58). Overall, it had been estimated that 20% to 50% of patients treated with lobectomy may need subsequent completion thyroidectomy (56–58). The majority of these patients would be upgraded from low to intermediate risk based on the presence of microscopic extrathyroidal extension or lymphatic invasion (58), which are controversial indications for RAI (59, 60). Furthermore, high-quality ultrasound can detect extrathyroidal extension with a negative predictive value ranging from 70% to 100% (61–63). This may explain why, despite the increasing number of patients treated with lobectomy, there does not yet seem to be an increase in frequency of completion thyroidectomy (6).

Ultimately, the decision to proceed with total thyroidectomy vs lobectomy has to take into account the risks vs benefits of each procedure and the individual patient's preferences advised by a multidisciplinary team (Table 2). Total thyroidectomy is associated with increased operative risk, including higher rates of laryngeal nerve injury, hypocalcemia, and postoperative hemorrhage (6, 7, 54). Patients who are treated with lobectomy are more likely to be managed as outpatients and have shorter length of stay, which substantially reduces health care costs (6, 7). All patients will require lifelong thyroid hormone supplementation after total thyroidectomy, while approximately one-third of patients undergoing lobectomy for cancer may avoid thyroid hormone supplementation (64, 65). Finally, other factors—such as comorbid conditions, the presence of contralateral nodules, preexisting hypothyroidism—can tip the balance to favor one approach over another in any given case.

Table 2.

Advantages of lobectomy vs total thyroidectomy for low-risk differentiated thyroid cancer

Favors lobectomy	Favors total thyroidectomy
Fewer postoperative complications (eg, hypocalcemia, nerve injury)	Allows for adjuvant RAI if necessary
Improved short-term (< 6 mo) quality of life	No risk of reoperation for completion thyroidectomy
Preserves thyroid tissue; one-third will not need hormone replacement	Facilitates postoperative surveillance (eg, ultrasound and Tg levels)
Increased frequency of outpatient perioperative management	May reduce recurrence ratesa
Decreased surgical costs

Abbreviations: RAI, radioactive iodine; Tg, thyroglobulin.

While some studies have suggested improved disease-free survival following total thyroidectomy, there are mixed results in the literature and this remains controversial. See text discussion.

The complexity and nuance involved in decision-making regarding extent of surgery has prompted investigation into which factors may push clinicians and/or patients toward lobectomy or total thyroidectomy (66, 67). McDow and colleagues (66) surveyed members of the ATA, the American Association of Endocrine Surgeons, and the American Head and Neck Society with a variety of hypothetical patient scenarios to identify factors associated with recommendations the clinicians provided for extent of surgery. They found that clinicians who were more likely to recommend total thyroidectomy (over lobectomy or active surveillance) in any scenario also reported being more risk averse and “uneasy” with uncertainty in patient care. To assess the patient perspective, Ahmadi et al (67) performed a discrete choice experiment that enrolled 150 patients who planned to undergo thyroid surgery. Using normalized preference weights, they found that the majority of the patients’ preferences were explained by risk of cancer recurrence (35%), risk of needing a second surgery (28%), and voice change (19%). Assuming an identical 5-year risk of recurrence but a higher postoperative complication rate for total thyroidectomy, they calculated that the average patient would favor lobectomy if the risk for completion thyroidectomy was less than 30% (67).

Other groups have focused on patient-centered outcomes and quality of life. A recent survey of nearly 2000 thyroid cancer patients investigated patient attitudes toward extent of surgery and RAI (68). Less than 19% reported that lobectomy was discussed as a surgical option; however, 39% reported that if oncologic outcomes were equivalent, they would choose lobectomy. Long-term quality of life outcomes following lobectomy and total thyroidectomy did not differ (69, 70). In a survey of 270 Canadian patients following either lobectomy (22%) or total thyroidectomy (78%), global quality of life was similar between groups (69). Patients who were treated with lobectomy expressed more anxiety regarding the possibility of cancer recurrence (2.4 vs 2.1 on a 5-point Likert scale; P = .021). A longitudinal, prospective quality of life survey of 1149 Chinese patients with DTC demonstrated better short-term postoperative quality of life following lobectomy as compared to total thyroidectomy (70). At 1 and 3 months postoperatively, the total thyroidectomy cohort reported lower cognitive and social function, including anxiety, depression, and voice symptoms. However, by 6 months, these differences disappeared, and cognitive and social function remained similar between groups at 12 months.

In conclusion, current evidence shows that lobectomy and total thyroidectomy have similar oncologic outcomes in terms both of recurrence-free and overall survival for patients with low-risk DTC. Lobectomy has superior short-term outcomes, including decreased complication rates and better short-term quality of life. However, lobectomy may complicate follow-up, as will be discussed further in this review, and patients may experience more anxiety regarding cancer recurrence following lobectomy. Importantly, current studies have found that patients have similar long-term postoperative quality of life after lobectomy or total thyroidectomy (69, 70). As such, care needs to be individualized based on each patient's risks and preferences (see Table 2).

Radioactive Iodine

In contrast to extent of surgery for low-risk DTC, where the 2015 ATA guidelines are ambivalent to lobectomy vs total thyroidectomy, adjuvant RAI for low-risk DTC “is not routinely recommended” (2). At the time these guidelines were written, the available data were predominantly from retrospective studies that found no difference in recurrence or survival for patients with low-risk DTC treated with or without RAI (2). Recently, the results of a randomized, controlled, noninferiority trial at 35 centers in France demonstrated no difference in recurrence-free survival for patients with low-risk DTC who underwent total thyroidectomy with adjuvant RAI vs total thyroidectomy alone (71). The study enrolled 776 adult patients with thyroid cancers either as a single tumor small than or equal to 2 cm or multifocal tumors (all < 1 cm with total sum of tumor diameters ≤ 2 cm) and no nodal disease, extrathyroidal extension, or aggressive variant histology. Patients were randomly assigned to treatment with 30 mCi of ¹³¹I or postoperative surveillance alone following total thyroidectomy. Enrolled patients were predominantly female (83%) with PTCs (96%) with a mean age of 52 years. Biochemical, pathological, or imaging evidence of recurrence was considered recurrent disease. Recurrence-free survival was 95.6% for the no-radioiodine group vs 95.9% for the radioiodine group at 3 years post randomization, a difference of −0.3% (CI, −2.7 to 2.2%) demonstrating noninferiority of no radioiodine.

Pasqual et al (8) analyzed SEER data and found that RAI use declined following publication of the 2015 ATA guidelines. However, they noted that in 2018, 23% of patients with intrathyroidal tumors 1 to smaller than 2 cm and 36% of patients with tumors 2 to smaller than 4 cm were treated with total thyroidectomy and RAI (8). RAI may be indicated in some low-risk DTC patients with hereditary cancer syndromes, radiation exposure, or other unique risk modifiers not captured in SEER; however, these findings suggest that at least one-fifth of low-risk DTC patients may still be overtreated with RAI (8). Overtreatment leaves patients vulnerable to adverse effects of RAI, such as taste and smell impairment, sialoadenitis, and increased long-term cancer risks (72, 73). More than half of the participants (8/15 patients, 53%) in a small qualitative study reported negative short- or long-term emotional and physical effects of RAI treatment (74).

Multiple studies have investigated the etiology of overtreatment with RAI in low-risk DTC (75, 76). Researchers from the University of Michigan surveyed 853 physicians involved in postsurgical management of thyroid cancer patients from 251 hospitals (response rate 63%), and found that extent of disease and completeness of surgical resection were the most important factors affecting recommendation for RAI (75). Physicians treating fewer thyroid cancer patients per year were more likely to identify both patient and physician “worry about death” as quite or very important in influencing the recommendation for RAI. In another study from this group, physician preference was shown to have a strong influence on whether low-risk DTC patients are treated with RAI; patients whose doctors reported using RAI for low-risk cancers were 84% more likely to be treated with RAI (76). Another study found that 56% of patients surveyed felt they had no choice in whether they received RAI, and that those patients were 2.5 times as likely to be treated with RAI compared to those who felt that they had a choice (77). Direct involvement of patients in the decision-making process decreases patient regret regarding RAI use, regardless of the final decision (78). Finally, socioeconomic and demographic factors play a role in the rates of RAI utilization. Studies have linked insurance status (79), race and ethnicity (80–82), and socioeconomic status (82–84) with likelihood of receiving adjuvant RAI.

Patient empowerment and physician education offer opportunities to reduce overtreatment (76–78). Sawka et al (85) designed a computerized decision aid on adjuvant RAI treatment for patients with low-risk DTC who underwent total thyroidectomy. They recruited 74 participants in a randomized trial and found that the decision aid halved the patients’ self-reported decisional conflict scores regarding whether to proceed with RAI. However, there was no difference in the rate at which patients chose to proceed with RAI. A follow-up survey of the study cohort found that at a mean of 17 months after randomization, the patients in the intervention group rated themselves as more informed, aware of options, and knowledgeable about risks and benefits of RAI than control group counterparts (86). Larger studies are needed to determine whether decision aids such as these can decrease RAI overtreatment.

Thyrotropin Suppression

Traditionally, TSH suppression following surgery for DTC was recommended for all patients (87). However, recognition of the harms of iatrogenic hyperthyroidism, including cardiovascular disease (88–91), osteoporosis (91–93), and even poor sleep function (94), have led to narrowing of the eligibility criteria for TSH suppression, particularly for low-risk patients. Serum TSH levels correlate strongly with postthyroidectomy quality of life for thyroid cancer patients (95–97), in part because iatrogenic hyperthyroidism can cause anxiety, depression, insomnia, and decline in cognitive function (96, 97). Thus, the ATA currently recommends that low-risk patients be maintained with a low-normal TSH between 0.5 and 2.0 mIU/L (2).

However, controversy persists over concerns that higher TSH levels may increase risk for recurrence. An older retrospective cohort study of 366 consecutive DTC patients found that patients with a median postablative TSH greater than or equal to 2.0 mIU/L had significantly higher risk of recurrence; approximately 30% vs 10% (98). Critically, no subanalyses were performed by initial risk stratification; high-risk patients may have predominantly driven the study's conclusions. More contemporary studies have shown that for low-risk patients, TSH suppression likely does not significantly reduce risk of recurrence (87, 92, 99, 100). A Japanese randomized controlled trial of 441 patients comparing TSH suppression (goal TSH > 0.01 mIU/L to < 0.5 mIU/L) to euthyroid TSH management (TSH goal ≥ 0.5 mIU/L and < 4.5 mIU/L) after total thyroidectomy found no difference in recurrence rates by treatment strategy (100).

TSH suppression following lobectomy has not been shown to provide substantial benefit for low-risk patients (97, 99). Park et al (99) performed a propensity-matched retrospective analysis of 446 patients with low-risk PTC treated with lobectomy and found no difference in recurrence risk between those treated with levothyroxine and those managed without, despite a higher median TSH level in the untreated group. A larger retrospective study of 2297 patients treated with lobectomy for DTC also found no association between mean postoperative TSH and risk of recurrence in low-risk DTC (101). More data are forthcoming from South Korea, as a randomized controlled trial is currently accruing to study the role of TSH suppression in low- to intermediate-risk DTC following lobectomy (102).

Despite the ATA's recommendation, the lack of consistent evidence of a benefit for TSH suppression in low-risk patients, and its inherent risks, many clinicians still recommend this practice (103). Papaleontiou and colleagues (103) conducted a survey of 654 surgeons and endocrinologists who cared for thyroid cancer patients about their practices regarding TSH suppression (69% response rate). They found that 48.8% would recommend TSH suppression below 0.5 mIU/L for patients with low-risk DTC. Physicians treating more than 40 thyroid cancer patients per year were less likely to recommend TSH suppression than those treating fewer than 20 patients per year (P = .035). Therefore, clinician education and referral to high-volume specialists may help reduce overtreatment.

Long-term Surveillance

Surveillance Imaging

High-quality cervical ultrasound is the mainstay for posttreatment surveillance for patients with low-risk DTC (2). The ATA does not recommend routine use of other imaging modalities, such as computed tomography, magnetic resonance imaging, or whole-body imaging, including RAI or ¹⁸FDG-PET (¹⁸F-fluoro-2-deoxy-D-glucose positron emission tomography) scans for surveillance of low-risk DTC. However, despite a recommendation for “periodic” ultrasound, the interval for repeat imaging is not well defined, is based on low-quality evidence, and has remained controversial (2). Thus, there is substantial heterogeneity regarding interval and length of posttreatment imaging follow-up often resulting in variation in care (104). One study reported a 5.3-fold increase in use of surveillance ultrasound for low-risk DTC patients between 2003 and 2012, despite a recurrence rate of only 0.4% (105). Frequent surveillance can lead to harm; Yang et al (106) found that 67% of postoperative DTC patients in their series had false-positive ultrasounds, which in turn may lead to unnecessary biopsies or treatment.

Inappropriate imaging can also be costly (104, 107, 108). A 2014 study estimated that US thyroid cancer care cost $1.6 billion, 37% of which was attributed to survivorship, including posttreatment surveillance (108). Surveillance is particularly cost-ineffective for low-risk DTC (107). Wang et al (107) performed a retrospective cost-effectiveness analysis of DTC surveillance based on ATA risk stratification at their single institution. They found that the cost to detect a single recurrence among low-risk DTC patients was $147 819, as compared to $22 434 and $20 680 for intermediate- and high-risk patients, respectively (107).

Several recent retrospective series have characterized the time-course of recurrence and the role of surveillance ultrasound in DTC. Park et al (109) reported on a series of 569 patients who underwent total thyroidectomy for PTC and were followed with serial ultrasound for a mean of 54.7 months; 31 patients (5.9%) developed recurrent disease, and the majority (21 patients, 67.7%) recurred 10 to 23 months postoperatively. An even larger series of 3106 PTC patients followed for a mean of 54.3 months found 94% of recurrences occurred within the first 5 years postoperatively (110). These results were corroborated in a separate multicenter retrospective review (111). These data suggest that an ultrasound 1 to 2 years postoperatively and a second 4 to 6 years postoperatively would detect the majority of recurrences (109).

Thyroglobulin Measurements

Following total thyroidectomy and RAI remnant ablation for DTC, TSH-stimulated Tg levels are highly predictive of persistent or recurrent disease (112). In combination with cervical ultrasound, stimulated Tg levels have a sensitivity and specificity for recurrent disease of 96.3% and 99.5%, respectively, in this setting (113). Thus, Tg levels have become an important tool in posttreatment DTC surveillance. The ATA's dynamic risk stratification system based on response to initial therapy relies in part on Tg levels, both for patients who have undergone total thyroidectomy with or without RAI and those treated with lobectomy alone (2, 114). In the modern era of highly sensitive Tg assays, stimulated Tg is no longer necessary for risk stratification (2, 115, 116). However, TSH and Tg antibody levels can affect Tg testing results (117). Additionally, the existing recommendations for utilization of Tg as a postoperative tumor marker are based on data from patients who underwent total thyroidectomy and RAI. In the current era, as more patients are treated with total thyroidectomy and lobectomy without RAI, the role of Tg monitoring and its interpretation is controversial, and prior evidence is likely no longer applicable (118).

A recent systematic review of the utility of Tg measurements following total thyroidectomy or lobectomy without RAI found very low-quality evidence for the utility of the test in either scenario (118). Following total thyroidectomy for low-risk DTC, the majority of patients have undetectable Tg levels (< 1 ng/mL) by 5 years postoperatively (119). A small prospective study of 57 low-risk DTC patients followed for a minimum of 3 years after total thyroidectomy without RAI found no recurrences by ultrasound, whole-body RAI or positron emission tomography scans, despite some patients having stimulated Tg values greater than 20 ng/mL (120). This is similar to other published findings in retrospective series with variable Tg cutoffs and testing regimens; none demonstrated a high specificity for Tg testing following total thyroidectomy without RAI (118, 119, 121–123).

Findings are even more heterogeneous following lobectomy (118, 124). In contrast to total thyroidectomy, Tg values tend to slowly increase over the first 5 years following lobectomy (124). A retrospective review of 1451 patients treated with lobectomy alone for DTC found higher Tg levels in patients with recurrence than those without (59.5 ± 195.6 ng/mL vs 6.6 ± 29.1 ng/mL; P < .001) (125). However, the study included 43.5% of patients with ATA intermediate- or high-risk disease who were therefore inappropriate candidates for lobectomy alone. Thus, these results are likely not interpretable in the setting of ATA-concordant treatment. In summary, there is presently little evidence to guide frequency of testing or interpretation of Tg results in patients following lobectomy or total thyroidectomy without RAI.

Conclusions

Low-risk DTC is overdiagnosed, but true incidence also has increased (12, 13). Efforts to curb overdiagnosis and overtreatment have successfully slowed the rising incidence trend and led to more patients treated with lobectomy alone and without RAI (6–9, 34, 35). Active surveillance may be an option for select patients with small, low-risk DTCs. However, these shifts in disease management have complicated treatment algorithms and resulted in a growing burden of nuanced choices for physicians and their patients. For example, a less aggressive treatment strategy could result in a patient treated with lobectomy alone, euthyroid off thyroid hormone replacement, having undergone 1 Tg check, and 2 ultrasounds total within 5 years. By comparison, an equally acceptable but more aggressive strategy may involve total thyroidectomy, RAI, TSH suppression, and Tg and ultrasound checks as often as every 6 months (Table 3). Treatment by expert thyroidologists increases the likelihood that low-risk DTC patients will be offered thyroid lobectomy alone and avoid RAI (75, 126). As more patients are treated with lobectomy and without RAI, posttreatment surveillance has become more challenging. Unfortunately, access to expert care in the United States is limited, and this trend may exacerbate present treatment disparities (127, 128).

Table 3.

Spectrum of treatment options for patients with low-risk differentiated thyroid cancer

Less aggressive treatment	More aggressive treatment	Relevant ATA recommendations
Lobectomy alone	Total thyroidectomy	35B
No RAI	Ablative dose of RAI (30 mCi)	51A and 55A
Neck ultrasound 6-12 mo postoperatively and again at 5 y	Neck ultrasound every 6-12 mo postoperatively	65A and D
Single Tg measurement 3-4 wk postoperatively	Tg measurement every 6-12 mo	50B and C, 62 B and C, 64
Levothyroxine supplementation to maintain goal TSH 0.5-2.0 mU/L	Levothyroxine supplementation to maintain goal TSH 0.1-0.5 mU/L	59 C, D, and E

American Thyroid Association (ATA) recommendations are adapted from reference (2).

Abbreviations: RAI, radioactive iodine; Tg, thyroglobulin; TSH, thyrotropin.

Ambiguity in treatment recommendations can be difficult to navigate, particularly when data are equivocal between treatment options or even simply not available. Ultimately, the overwhelming majority of patients with low-risk DTC will have excellent outcomes in terms of disease-free and overall survival. Thus, clinicians should rely on indicators such as quality of life to better inform decisions when guidelines and other evidence are ambiguous. We expect that these patient factors will be important drivers of treatment decisions in areas of controversy in the future.

Financial Support

M.P. is supported by the National Institute on Aging (award No. K08 AG049684). J.A.S. is a member of the Data Monitoring Committee of the Medullary Thyroid Cancer Consortium Registry supported by Novo Nordisk, Astra Zeneca, and Eli Lilly. Institutional research funding was received from Exelixis and Eli Lilly.

Abbreviations

ATA

American Thyroid Association

DTC

differentiated thyroid cancer

FNA

fine-needle aspiration

HR

hazard ratio

PTC

papillary thyroid cancer

PTmC

papillary thyroid microcarcinoma

RAI

radioactive iodine

SEER

Surveillance, Epidemiology, and End Results

Tg

thyroglobulin

TSH

thyrotropin

Author Contributions

All authors contributed to the design, reference sourcing, writing, and editing of this work.

Disclosures

T.U. and M.P. have nothing to disclose. J.A.S. was a co-author of the American Thyroid Association guidelines for the management of adult patients with thyroid nodules and differentiated thyroid cancer published in 2015; she is co-chairing the next iteration of the ATA guidelines for adult patients with differentiated thyroid cancer.

Data Availability

Data sharing is not applicable to this article as no data sets were generated or analyzed during the present study.