Risk Factors for Recurrence of Follicular Thyroid Cancer: A Systematic Review

Abstract

Background:

In risk assessment of recurrence, papillary thyroid cancer (PTC) and follicular thyroid cancer (FTC) are often grouped together as differentiated thyroid cancer (DTC). However, while risk factors affecting recurrence of PTC are well established, risk factors for recurrence of FTC are not. This systematic review examines risk factors for recurrence of FTC and evaluates their significance.

Methods:

A systematic search on PubMed and Embase was performed in September 2020, including studies evaluating risk factors for recurrence of FTC. A quality assessment of the enrolled studies was performed.

Results:

Nine studies (n = 1544 patients) from eight countries were included. The average recurrence rate was 13.6%, and distant metastasis (DM) constituted 64.8% of the recurrent cases. The risk factors examined were sex, age at diagnosis, primary tumor size, degree of invasiveness, focality, positive resection margin, lymph node (LN) metastasis, and DM at diagnosis. Risk factors correlated with recurrence of FTC were age older than 45 years, primary tumor size above 40 mm, widespread invasion, multifocality, positive resection margin, LN metastasis, and DM at diagnosis. Sex was not a statistically significant risk factor.

Conclusions:

We identified seven risk factors associated with recurrence of FTC. Age and multifocality were found to be of greater impact regarding recurrence risk of FTC compared with PTC. Future research needs to address the impact of different risk factors for recurrence of FTC particularly including age, primary tumor size, angioinvasion, and mutational status.

Introduction

In Western countries, thyroid cancer is the most common endocrine neoplasm and is nearly three times more frequent in females (1 –4). The incidence of thyroid cancer has increased substantially since 1980, in parallel with survival rates (1 –3). Some studies explain the increase in incidence due to better access to diagnostic imaging including ultrasonography, leading to diagnosis of many asymptomatic thyroid nodules (1,3,5). Papillary microcarcinomas (<1 cm) particularly account for this increase. Another explanation is that the average amount of external radiation in the environment has increased (2,5).

Recurrence of thyroid cancer is estimated to occur in 6–30% of treated cases (6 –8). Papillary thyroid cancer (PTC) is the most common and least aggressive subtype of thyroid cancer and its prognostic factors are well investigated (8,9). Recently, Guo and Wang (8) published a meta-analysis on risk factors of recurrence of PTC. Conversely, risk factors specific for recurrence of follicular thyroid cancer (FTC) are not well examined.

FTC is the second most frequent subtype of thyroid cancer and constitutes 11–18% of the total number of thyroid cancer cases (2,3,10). The incidence ranges between 0.71 and 1.0 cases per 100,000 (3,4,11,12).

In the literature as well as in staging systems, FTC is often grouped together with PTC as differentiated thyroid cancer (DTC). Yet, these subtypes of thyroid cancer differ in several ways (6,9). Because of the relatively large number of PTC cases compared with FTC cases, most of the current staging systems of DTC are mainly based on PTC data (9,13).

Compared with PTC, FTC is characterized as being a more aggressive malignancy and as having a worse outcome with respect to recurrence and survival (3,7,9,14). In terms of recurrence and in contrast to PTC, FTC recurs more often as distant metastases (DM) than as locoregional (LR) lymph node (LN) metastases, suggesting a hematogenous route of metastasis. Thus, recurrences of PTC are often more amenable for local treatment compared with FTC recurrences (6,7,9,13,15).

The aim of this article was to evaluate potential risk factors for tumor recurrence of patients with FTC to improve prognostication of the course of FTC and thereby improve the treatment.

Materials and Methods

The authors confirm that the present study meets the ethics guidelines, including adherence to the legal requirements of Denmark. Ethical approval was not relevant for this study, since it is solely based on literature.

This systematic review followed the procedures of the PRISMA guidelines (16).

A systematic search was performed using the databases PubMed and Embase by one author (M.P.G.), last updated on September 1, 2020. The main keywords in the search strategy were “follicular thyroid cancer,” “recurrence,” and “risk factors.” (See Supplementary Data for the complete search strategy.)

Studies examining pretreatment risk factors for recurrence of FTC were included. Treatment modalities and postsurgical treatment were not included as risk factors due to the risk of confounding. Studies including 10 or less patients were excluded alongside systematic reviews and meta-analyses. Studies that did not differentiate between subtypes of thyroid cancer in general or in analysis were excluded as well.

The studies were screened in two rounds. The first screening evaluated title and abstract for eligibility, while the second screening reviewed full text and analysis. If there was doubt whether a study met the inclusion criteria, it was discussed with the author group.

The following data were extracted from the included studies: publication year, country, number of patients with FTC, number of recurrences, female to male ratio, mean age, follow-up time, and time to recurrence. Furthermore, quantitative measures and outcomes of selected risk factors for recurrence of FTC were extracted if reported. Risk estimates extracted from multivariate analyzes were prioritized.

A quality assessment of the included studies was carried out using a rating tool for observational cohort and cross-sectional studies made available by the National Heart, Lung, and Blood Institute (NHLBI) of the United States Department of Health and Human Services. The rating process consisted of 14 questions for each study about study design, follow-up, and transparency. The percentage of positive answers determined the rating of quality of each study. The possible outcomes were poor, fair, and good.

Statistical analysis

If a study did not report a quantitative measure of a risk factor, we included our own calculations in the results when possible. Our calculations were carried out by extracting data of occurrence and distribution of patients with or without recurrence, when it was possible to extract these data from the studies.

Calculations of risk estimates were performed with the software R-statistics (version 3.6.0) using the Epi-package and twoby2 function (17). Relative risk and associated 95% confidence interval [CI] and p-value were reported. Fisher's exact t-test was used if one or more groups contained zero cases.

Results

The search of PubMed and Embase databases identified 326 studies after removal of duplicate studies. A total of nine studies (14,18 –25) met the inclusion criteria, comprising 1544 patients in total. Three studies were found in the references from the full text review of studies (Fig. 1). Four studies were from Asia (18,20,22,23), three from Europe (14,19,21), and two from North America (24,25). The mean follow-up times ranged from 5 to 23.7 years. The reported recurrence rates ranged from 2% to 29%, with an overall recurrence rate of 13.6%. The mean times to recurrence were 2–14.7 years. By geographic region, the recurrence rate in the studies from Asian countries were 10.6% (n = 877 patients) versus 17.2% (n = 667 patients) in the studies from Western countries. Where type of recurrence was reported, 64.8% of recurrent cases were DM after exclusion of patients with DM at diagnosis. The most common sites of DM were bones and lungs, while liver and brain metastasis were also reported with less frequency. The characteristics of the included studies are shown in Table 1.

FIG. 1.

PRISMA flowchart.

Table 1.

Characteristics of the Included Studies Comprising 1544 Patients with Follicular Thyroid Cancer

Reference	Country	Publication year	Female/male (ratio)	Age (mean), years	Number of patients with recurrence/total patients (%)	DM/LR recurrences	Follow-up time (mean), years	Time to recurrence (mean)
Asari et al. (14)	Austria	2009	148/59 (2.5:1)	57.8	23/207 (11%)	20/3	9.7	—
Chow et al. (20)	Hong Kong	2002	159/56 (2.8:1)	49	53/215 (25%)	30/23	10.8	—
D'Avanzo et al. (25)	United States	2004	92/40 (2.3:1)	49	21/132 (16%)	—	7.5	—
Enomoto et al. (22)	Japan	2013	18/2 (9:1)	17.3	3/20 (15%)	2/1	23.7	14.7 years
Kim et al. (23)	South Korea	2014	470/93 (5.0:1)	43	14/563 (2%)	—	6	—
Mueller-Gaertner et al. (21)	Germany	1991	113/27 (4.2:1)	52.4	24/140 (17%)	16/8	7.1	80% within 6 years
Ngo Lo et al. (18)	Philippines	2015	72/7 (10.3:1)	48.9	23/79 (29%)	—	6.9	2.2 years
Tzavara et al. (19)	Greece	1999	109/46 (2.4:1)	49	44/155 (28%)	—	7	—
Welch Dinauer et al. (24)	United States	1998	28/5 (5.6:1)	18	5/33 (15%)	2/3	5	2 years^a 90% within 7 years

DM, distant metastasis; LR, locoregional.

Median.

Eight risk factors were reported by more than one study: sex, age, primary tumor size, degree of invasiveness, focality, positive resection margin, LN involvement at diagnosis, and DM at diagnosis. Some studies reported data on radioactive iodine treatment as a risk factor, but the data were not extracted. Thyroglobulin (Tg) levels were not evaluated as a risk factor since only one study reported data on this. Multivariate analysis was performed in three studies (18,20,23) and was adjusted for demographic, clinical, and treatment factors. Table 2 summarizes the data extracted and our calculations.

Table 2.

Results of Selected Risk Factors for Recurrence of Follicular Thyroid Carcinoma Among Included Studies

Reference	Sex			Age			Primary tumor size			Invasiveness^f
Reference	Definition	Result	Significance	Definition	Result	Significance	Definition	Result	Significance	Definition	Result	Significance
Asari et al. ^a (14)	Male vs. female	MI: RR 1.97^b WI: RR 1.81^b	CI: 0.46–8.37 CI: 0.76–4.31	>45 vs. ≤45 years	MI: t-test^c WI: t-test^c	p = 0.11 p = 0.34	>40 vs. ≤40 mm	MI: RR 15.17^b WI: RR 1.05^b	CI: 1.89–121.59 CI: 0.43–2.55	WI vs. MI	RR 5.26^b	CI: 2.31–11.99
Chow et al. (20)	—	—	—	>45 vs. ≤45 years	RR 3.1^d	CI: 1.6–7.8ⁱ	—	—	—	—	—	—
D'Avanzo et al. (25)	—	—	—	—	—	—	—	—	—	AI vs. MI WI vs. MI	RR 1.08^b RR 3.38^b	CI: 0.35–3.30 CI: 1.27–8.94
Enomoto et al. (22)	Male vs. female	t-test^c	p = 0.28	Average	t-test	p = 0.116	Average	t-test	p = 0.922	AI: yes vs. no WI: yes vs. no	t-test^c t-test^c	p = 0.074 p = 1
Kim et al. (23)	Male vs. female	HR 0.8	CI: 0.2–3.6	Per 1 year increase	HR 1.0	CI: 1.0–1.1	Per cm increase	HR 1.3ⁱ	CI: 1.1–1.6	AI: yes vs. no WI: yes vs. no	HR 3.3 HR 2.1	CI: 1.2–9.4 CI: 0.7–6.1
Ngo Lo et al. (18)	Male vs. female	OR 3.15	p = 0.077	>45 vs. ≤45 years	OR 2.50	p = 0.043	>40 vs. ≤40 mm	OR 0.86	p = 0.775	—	—	—
Mueller-Gaertner et al. (21)	Male vs. female	RR 1.40^b	CI: 0.61–3.18	>50 vs. ≤50 years >60 vs. ≤60 years	RR 2.06^b RR 2.19^b	CI: 0.94–4.50 CI: 1.07–4.47	—	—	—	Tumor capsule present vs. absent	RR 2.19^b	CI: 1.05–4.57
Tzavara et al. ^h (19)	—	—	—	>60 vs. 41–60 years	RR 7.3	p = 0.01	>40 vs. ≤40 mm	RR NA	p = 0.4	—	—	—
Welch Dinauer et al. (24)	Male vs. female	NA	—	16 vs. 20 years	χ ²-test	p = 0.065	—	—	—	—	—	—

Reference	Focality			Resection margin			Cervical LN metastasis at diagnosis (N1a/N1b)			DM at diagnosis (M1)
Reference	Definition	Result	Significance	Definition	Result	Significance	Definition	Result	Significance	Definition	Result	Significance
Asari et al. ^a (14)	Multi- vs. unifocal	MI: t-test^c WI: RR 1.81^b	p = 1 CI: 0.74–4.49	—	—	—	N1x vs. N0	MI: NA WI: RR 1.33^b	CI: 0.50–3.59	—	—	—
Chow et al. (20)	—	—	—	Postsurgical macroscopic residual disease (no vs. yes)	RR 0.037^e	p < 0.01	N1x vs. N0	RR 2.6^d	CI: 1.4–4.8ⁱ	M1 vs. M0	RR 4.4^e	CI: 1.76–11.2ⁱ
D'Avanzo et al. (25)	—	—	—	—	—	—	—	—	—	Excluded^g	—	—
Enomoto et al. (22)	—	—	—	—	—	—	N1x vs. N0	t-test^c	p = 1	—	—	—
Kim et al. (23)	Multifocal (yes vs. no)	HR 0.7	CI: 0.2–3.0	—	—	—	N1x vs. N0	HR 30.4ⁱ	CI: 9.2–100.3	Excluded^g	—	—
Ngo Lo et al. (18)	Multifocal (yes vs. no)	OR 4.25	p = 0.027	Incomplete thyroidectomy (yes vs. no)	OR 5.32	p < 0.01	N1x vs. N0	OR 1.63	p = 0.435	M1 vs. M0	OR 21.56ⁱ	p < 0.001
Mueller-Gaertner et al. (21)	—	—	—	—	—	—	N1x vs. N0	χ ²-test RR 0.97^b	p = 0.491 CI: 0.36–2.57	Excluded^g	—	—
Tzavara et al. ^h (19)	—	—	—	—	—	—	N1x vs. N0	RR 2.10^b	CI: 0.87–5.10	M1 vs. M0	RR 3.78^b	CI: 2.51–5.69
Welch Dinauer et al. (24)	Multi- vs. unifocal	OR 22.0	CI: 1.3–362.9	—	—	—	—	—	—	—	—	—

95% confidence interval presented as level of significance (95% CI). If not reported, then p-value is used.

Exposed groups at risk calculation: male sex, higher age, larger tumor size, tumor capsule absent, extrathyroidal invasion, widely invasive or angioinvasive compared with minimally invasive, multifocal tumor, N1a/N1b status, and M1 status.

WHO definition of invasiveness: minimally invasive characterized by histologically confirmed focal capsular invasion. Angioinvasive defined as tumor growth in blood or lymph vessels with or without capsular invasion. Widely invasive characterized by widespread tumor cell invasion of the thyroid parenchyma and extrathyroidal tissue (31).

Data are grouped by MI and WI.

Relative risk was calculated by extracting table content from the respective article.

Relative risk was not possible to calculate because 0 was present in one or more groups. Fisher's exact t-test was performed instead.

Significant for recurrence as DM.

Data are extracted from a multivariate analysis and are adjusted for treatment type (type of surgery and RAI ablation) as well as various demographic and clinical data similar to several of the risk factors included in the table.

Patients were grouped by stages: stage 1 represents intrathyroidal disease only, stage 2 +cervical LN involvement (N1a or N1b), stage 3 +extrathyroidal invasion, and stage 4 +DM (M1).

Significant for LN recurrence.

Patients initially diagnosed with DM at presentation were excluded from the statistical analysis.

AI, angioinvasive; HR, hazard ratio; LN, lymph node; MI, minimally invasive; NA, not available; OR, odds ratio; RAI, radioactive iodine; WHO, World Health Organization; WI, widely invasive.

Six studies assessed sex as a risk factor for recurrence of FTC (14,18,21 –24), and none found that male sex was associated with a significantly increased risk of recurrence.

Age was reported as a risk factor in seven studies. Three studies (14,18,20) divided their patients into age groups of ≤45 and >45 years at diagnosis, two of which (18,20) found older age was associated with a significantly increased risk of recurrence. Two studies (19,21) divided their population by decades of age and found a significant association with increased risk of recurrence in patients older than 60 years.

Two other studies evaluated the recurrence pattern of FTC in children and adolescents (22,24) and found the recurrence rate to be similar to FTC in adults (7,14,21,25). Moreover, Welch Dinauer et al. (24) found that younger age at diagnosis was associated with an increased risk of recurrence (median age 16 vs. 20 years, p = 0.065).

Size of primary tumor was examined as a risk factor in five studies (14,18,22 –24), but only two studies found it to be significantly associated with risk of recurrence. Kim et al. (23) measured the hazard ratio (HR) per centimeter increase of primary tumor size and found a HR of 1.3 ([CI 1.1–1.6], p < 0.01). Additionally, they compared the recurrence risk of patients with a primary tumor size of ≤2 cm with a primary tumor size of >4 cm and found a HR of 23.1 ([CI 6.7–80.3], p < 0.01) for recurrence for the larger tumor size.

Five studies assessed the correlation between the degree of invasiveness and FTC recurrence. One study (14) compared minimally and widely invasive tumors, whereas three others (22,23,25) additionally included angioinvasion in the risk calculation. Both angioinvasion and/or widespread invasion were associated with a higher recurrence rate in all studies. Mueller-Gaertner et al. (21) examined the effect of tumor growth outside the tumor capsule and found that extrathyroidal extension was associated with an increased risk of recurrence.

Patients with unifocal versus multifocal tumors were compared in four studies (14,18,23,24), two of which (18,24) found that multifocal tumor was significantly associated with an increased risk of recurrence. Ngo Lo et al. (18) reported that multifocal tumors had an increased risk of recurrence with an odds ratio (OR) of 4.25 (p = 0.027).

Positive surgical margin as a risk factor was reported by two studies (18,20) and was associated with a significantly increased risk of recurrence of FTC. Ngo Lo et al. estimated the recurrence risk of FTC after an incomplete thyroidectomy with an OR of 5.32 (p < 0.001).

The presence of malignancy in LR LNs at diagnosis (N1a and N1b) was assessed as a risk factor in seven studies (14,18 –23). Five studies (14,18 –20,23) found a higher risk of recurrence in cases with LN involvement, two of which (20,23) found significance. Chow et al. and Kim et al. found a RR of 2.6 [CI 1.4–4.8] and HR of 30.4 [CI 9.2–100.3], respectively.

Three studies (18 –20) reported data of DM at presentation (M1) and found it to be significantly associated with an increased risk of recurrence of FTC. Ngo Lo et al. (18) reported an OR of 21.56 (p < 0.001) for recurrence in patients with M1 tumors.

The quality assessment of the included studies showed four studies as Fair and five studies as Good. Thus, the overall quality measured was Fair to Good. (See Table 3 for the results of the quality assessment.)

Table 3.

Quality Assessment of the Included Studies

Reference	1. Was the research question or objective in this article clearly stated?	2. Was the study population clearly specified and defined?	3. Was the participation rate of eligible persons at least 50%?	4. Were all the subjects selected or recruited from the same or similar populations? Were inclusion and exclusion criteria for being in the study prespecified and applied uniformly to all participants?	5. Was a sample size justification power description or variance and effect estimates provided?	6. For the analyses in this article were the exposure(s) of interest measured before the outcome(s) being measured?	7. Was the time frame sufficient so that one could reasonably expect to see an association between exposure and outcome if it existed?	8. For exposures that can vary in amount or level, did the study examine different levels of the exposure as related to the outcome?	9. Were the exposure measures (independent variables) clearly defined, valid, reliable, and implemented consistently across all study participants?	10. Was the exposure(s) assessed more than once over time?	11. Were the outcome measures (dependent variables) clearly defined, valid, reliable, and implemented consistently across all study participants?	12. Were the outcome assessors blinded to the exposure status of participants?	13. Was loss to follow-up after baseline 20% or less?	14. Were key potential confounding variables measured and adjusted statistically for their impact on the relationship between exposure(s) and outcome(s)?	Number: yes/number: possible	Result^a
Asari et al. (14)	Yes	Yes	Yes	Yes	No	Yes	Yes	NA	Yes	NA	Yes	NA	Yes	No	9/11	Good
Chow et al. (20)	Yes	Yes	Yes	Yes	No	Yes	Yes	NA	Yes	NA	No	NA	Yes	No	8/11	Fair
D'Avanzo et al. (25)	Yes	No	Yes	CD	No	Yes	Yes	NA	Yes	NA	Yes	NA	Yes	No	7/10	Fair
Enomoto et al. (22)	Yes	Yes	Yes	Yes	No	Yes	Yes	NA	Yes	NA	Yes	NA	Yes	No	9/11	Good
Kim et al. (23)	Yes	Yes	Yes	Yes	No	Yes	Yes	NA	Yes	NA	Yes	NA	Yes	Yes	11/12	Good
Mueller-Gaertner et al. (21)	Yes	Yes	Yes	Yes	No	Yes	Yes	NA	Yes	NA	No	NA	Yes	No	8/11	Fair
Ngo Lo et al. (18)	Yes	Yes	Yes	Yes	No	Yes	Yes	NA	Yes	NA	Yes	NA	Yes	Yes	10/11	Good
Tzavara et al. (19)	Yes	Yes	Yes	Yes	No	Yes	Yes	Yes	Yes	NA	No	NA	Yes	Yes	10/12	Good
Welch Dinauer et al. (24)	Yes	Yes	Yes	Yes	No	Yes	No	NA	Yes	NA	Yes	NA	Yes	No	8/11	Fair

CD, cannot determine; NA, not applicable.

Potential outcomes are good, fair, or poor determined by the number of yes's out of the total number of applicable questions.

Discussion

The American Thyroid Association (ATA) recommends staging of thyroid cancer by the 2015 American Joint Committee on Cancer (AJCC)/Union for International Cancer Control (UICC) classification system based on age (≤45 vs. >45 years) and TNM stage (26,27). However, this system was developed for mortality risk, not recurrence. In addition, the ATA developed a three-level stratification system in 2009, and updated it in 2015, to divide patients with DTC into low-, intermediate-, or high-risk groups for recurrence (26,27). In short, the risk factors taken into account comprise: presence of LR and/or DM, incomplete tumor resection, extrathyroidal extension, presence of aggressive histology, Tg levels, and mutational status for PTC (26). The studies included in this review examined several of the risk factors of the 2015 ATA and the 2015 AJCC/UICC recommendations, except for aggressive histology and mutational status.

Age is not taken into account regarding recurrence risk estimation of DTC by the 2015 ATA modified guidelines (27). However, this review found a higher recurrence risk of FTC when diagnosed after 45 years of age, but statistical significance was inconsistent. The association of age and risk of recurrence needs further examination regarding FTC.

Regarding the included studies specifically investigating children and adolescents, younger age at diagnosis increased the risk of recurrence (24), indicating a U-shaped age-recurrence-association with the highest recurrence risk at the extremes of age, which is consistent with the findings of Mazzaferri and Jhiang (7).

Primary tumor size is included as T score in the 2015 AJCC/UICC classification system of DTC but not in the 2015 ATA modified guidelines for assessment of recurrence risk of DTC (26,27). This systematic review did not find consistently positive results regarding tumor size and recurrence risk. More studies on tumor size and recurrence risk of FTC are required before a conclusion can be drawn.

Multifocality, extensive angioinvasion (≥4 blood vessels with tumor invasion), wide invasion, and extrathyroidal extension are all individual risk factors in the ATA recommendations for recurrence assessment, placing a patient in the intermediate- or high-risk group. The current systematic review found nearly all these risk factors to be significant for recurrence of FTC. The importance of these risk factors has been widely reported (6,8,9,14,23,26 –28).

According to the 2015 ATA modified guidelines on recurrence risk, an incomplete tumor resection places a patient with DTC in the high-risk group for recurrence. Only two studies reported data on this risk factor on FTC with varying definitions. Ngo Lo et al. (18) assessed incomplete thyroidectomy and found an OR of 5.32 (p < 0.001). Chow et al. (20) found a RR of 0.037 (p < 0.001) for no postsurgical macroscopic disease compared with the presence of macroscopic tumor after surgery for LN recurrence. These findings suggest a strong correlation between residual cancer postsurgically and recurrence for FTC.

The presence of LR and/or DM at diagnosis were found to be closely associated with recurrence risk of FTC. These risk factors are also included in the 2015 ATA modified guidelines of recurrence risk assessment. Depending on the size and number of pathological LN metastasis, a patient with an N1 tumor can be placed in all three risk groups. This was added to the 2015 version of the ATA risk stratification system. Meanwhile, a patient with an M1 tumor is staged in the high-risk group regardless of T and N status. Some studies in this systematic review excluded patients with DM at presentation because of a chance of inaccurate risk assessment for recurrence (21,23,25). The three studies reporting data on DM at presentation as a risk factor (18 –20) are unclear whether patients with DM at presentation had complete remission before they were included in risk calculation for recurrence. This ought to be true for these patients to acquire a recurrent cancer. Since this is uncertain, it should be taken into consideration when interpreting the results. However, the findings indicate a strong association between both LR and DM at the time of diagnosis and recurrence of FTC.

Sex is not included as a risk factor in any of the aforementioned guidelines. The current systematic review found no significant association between sex and risk of recurrence of FTC.

In comparison, a systematic review and meta-analysis was conducted by Guo and Wang (8), which evaluated several potential risk factors for recurrence of PTC. They found male sex (OR 1.53 [CI 1.28–1.84], p < 0.01), extrathyroidal extension (OR 2.83 [CI 2.32–3.44], p < 0.01), LN metastasis (OR 3.24 [CI 2.61–4.02], p < 0.01), primary tumor size >2 cm (OR 2.69 [CI 2.06–3.50], p < 0.01), incomplete tumor removal (OR 2.38 [CI 1.81–3.12], p < 0.01), DM at diagnosis (OR 11.96 [CI 8.43–16.97], p < 0.01), and no adjuvant radioiodine treatment (OR 1.34 [CI 1.01–1.78], p = 0.04) to be significant risk factors for recurrence of PTC. Neither older age nor multifocality was significantly associated with a higher recurrence risk of PTC (8). Thus, some of the risk factors for recurrence are comparable between PTC and FTC, but the impact of each variable on recurrence risk seems to be different.

In summary, our findings support the risk factors: degree of invasiveness, multifocality, incomplete tumor removal, LR metastasis, and DM in the 2015 ATA modified guidelines of recurrence risk assessment regarding FTC. The guidelines did not include age nor primary tumor size; however, the results of this review suggest a correlation between both variables and the recurrence risk of FTC. However, more well-powered and well-designed studies on the topic of this review are required to support our results. This is especially true for age and tumor size before considering future addition of these to the guidelines.

Various other staging systems have been proposed (e.g., MACIS, AGES, AMES), but none have shown superiority nor differentiate between FTC and PTC (10,26,27). This is possibly leading to inaccurate prognostication of FTC due to various reasons. First, PTC is the predominant subtype in terms of DTC, thus the focus might be shifted from FTC in the staging systems (9,13). Second, even though the risk factors for recurrence of PTC and FTC are much alike, we found that the magnitude of each risk factor is apparently different between the two (8). Also, FTC turns out to be more present in iodine-deplete areas, suggesting that iodine levels interfere with the incidence and recurrence rates of FTC (2,14,29). Therefore, particularly areas with high incidence of FTC might not benefit optimally from the current assessment tools for recurrence risk. Based on our findings, the predictive value of the current recurrence risk assessment methods could be improved by distinguishing between PTC and FTC. Eventually, a better understanding of tumor behavior of FTC will aid the identification of patients benefiting from more intense monitoring.

A limitation to this review is that the definitions of recurrence are inconsistent among studies. For instance, some studies require a tumor-free time span of varying length between treatment and detection of a new tumor mass before accounting it as a recurrence, while others do not. Also, varying methods for detection of recurrence (e.g., cytological, histological, biochemical, diagnostic imaging, and clinical examination) are used. Likewise, the included studies originate from different countries with possibly dissimilar access to treatment, paraclinics, and follow-up affecting the recurrence rates. The fact that these factors vary among the studies creates a potential underreporting of recurrent FTC cases in some studies.

Moreover, the oxyphilic variant of FTC, Hürthle cell thyroid carcinoma (HCTC), was included in some studies (14,19,23), excluded in others (20) or not reported (18,21,22,24,25). Patients with HCTC tend to be older, have more lymphovascular invasions, and infrequently have DM (23). Furthermore, the ATA guidelines consider HCTC as an aggressive histological variant of FTC. However, the distinction between HCTC and FTC as two separate types of thyroid cancer was made by the World Health Organization (WHO) in 2017 (30,31). Hence, most studies published before 2017 might not be designed to distinguish between the two, thereby pooling the results of FTC and HCTC. Overall, the lack of consistency regarding inclusion of HCTC should be taken into consideration when interpreting the results.

No statistical method was used to examine the risk of publication bias. However, with the result of our quality assessment of the nine included studies being Fair to Good, our findings in this systematic review seem reliable with a minor chance of bias.

In conclusion, age older than 45 years, primary tumor size above 40 mm, angioinvasion and wide invasion, multifocality, incomplete tumor removal, LR LN metastasis, and DM at diagnosis were found to be risk factors for recurrence of FTC. Age and multifocality seem of greater magnitude of FTC recurrence risk, while male sex, as opposed to in PTC (8), was not found to be a risk factor for FTC recurrence. Thus, differentiation of PTC and FTC in risk assessment of recurrence might improve prognostication and aid in identifying patients benefiting from close follow-up. More prospective and retrospective studies focusing on FTC are recommended to clarify the full extent of risk factors for recurrence and thereby allowing a recurrence risk assessment system meant solely for FTC. In particular, age and primary tumor size were found as potential risk factors, and with more support in future studies, they could be included in future recurrence risk assessment systems. Also, future studies examining the molecular profile of the primary tumor and its correlation to recurrence of FTC are desired, which potentially would modernize the diagnostic process.

Footnotes

Authors' Contributions

Each author contributed substantially to this study and approved the final version for publication.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

No funding was received for this article.

Supplementary Material

Supplementary Data

References

Wiltshire

, Drake

, Uttley

, Balasubramanian

. 2016. Systematic review of trends in the incidence rates of thyroid cancer. Thyroid, 26:1541–1552.

Pellegriti

, Frasca

, Regalbuto

, Squatrito

, et al. 2013. Worldwide increasing incidence of thyroid cancer: update on epidemiology and risk factors. J Cancer Epidemiol, 2013:965212.

Mirian

, Grønhøj

, Jensen

, Jakobsen

, et al. 2018. Trends in thyroid cancer: retrospective analysis of incidence and survival in Denmark 1980–2014. Cancer Epidemiol, 55:81–87.

Lim

, Devesa

, Sosa

, Check

, et al. 2017. Trends in thyroid cancer incidence and mortality in the United States, 1974–2013. JAMA, 317:1338–1348.

Tufano

, Noureldine

, Angelos

. 2015. Incidental thyroid nodules and thyroid cancer: considerations before determining management. JAMA Otolaryngol Head Neck Surg, 141:566–572.

Schlumberger

. 1998. Papillary and follicular thyroid carcinoma. N Engl J Med, 338:297–306.

Mazzaferri

, Jhiang

. 1994. Long-term impact of initial surgical and medical therapy on papillary and follicular thyroid cancer. Am J Med, 97:418–428.

Guo

, Wang

. 2014. Risk factors influencing the recurrence of papillary thyroid carcinoma: a systematic review and meta-analysis. Int J Clin Exp Pathol, 7:5393–5403.

Daniels

. 2018. Follicular thyroid carcinoma: a perspective. Thyroid, 28:1229–1242.

10.

Hundahl

, Fleming

, Fremgen

, Menck

. 1998. A National Cancer Data Base report on 53,856 cases of thyroid carcinoma treated in the U.S., 1985–1995. Cancer, 83:2638–2648.

11.

Roche

, Niu

, Pawlish

, Henry

. 2011. Thyroid cancer incidence in New Jersey: time trend, birth cohort and socioeconomic status analysis (1979–2006). J Environ Public Health, 2011:850105.

12.

Colonna

, Grosclaude

, Remontet

, Schvartz

, et al. 2002. Incidence of thyroid cancer in adults recorded by French cancer registries (1978–1997). Eur J Cancer, 38:1762–1768.

13.

Lang

BHH

, Lo

, Chan

, Lam

, et al. 2007. Prognostic factors in papillary and follicular thyroid carcinoma: their implications for cancer staging. Ann Surg Oncol, 14:730–738.

14.

Asari

, Koperek

, Scheuba

, Riss

, et al. 2009. Follicular thyroid carcinoma in an iodine-replete endemic goiter region: a prospectively collected, retrospectively analyzed clinical trial. Ann Surg, 249:1023–1031.

15.

O'Neill

, Vaughan

, Learoyd

, Sidhu

, et al. 2011. Management of follicular thyroid carcinoma should be individualised based on degree of capsular and vascular invasion. Eur J Surg Oncol, 37:181–185.

16.

Moher

, Shamseer

, Clarke

, Ghersi

, et al. 2015. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev, 4:1.

17.

Carstensen

, Plummer

, Laara

, Hills M 2020 Epi: Statistical Analysis in

Epidemiology

. 2.42. Available at https://cran.r-project.org/package=Epi (accessed on May 11, 2021 ).

18.

Ngo Lo

, Canto

, Maningat

PDD

. 2015. Risk factors for recurrence in filipinos with well-differentiated thyroid cancer. Endocrinol Metab, 30:543–550.

19.

Tzavara

, Vlassopoulou

, Alevizaki

, Koukoulis

, et al. 1999. Differentiated thyroid cancer: a retrospective analysis of 832 cases from Greece. Clin Endocrinol, 50:643–654.

20.

Chow

, Law

, Mendenhall

, Au

, et al. 2002. Follicular thyroid carcinoma: prognostic factors and the role of radioiodine. Cancer, 95:488–498.

21.

Mueller-Gaertner

, Brzac

, Rehpenning

. 1991. Prognostic indices for tumor relapse and tumor mortality in follicular thyroid carcinoma. Cancer, 67:1903–1911.

22.

Enomoto

, Enomoto

, Uchino

, Yamashita

, et al. 2013. Follicular thyroid cancer in children and adolescents: clinicopathologic features, long-term survival, and risk factors for recurrence. Endocr J, 60:629–635.

23.

Kim

, Kim

, Jang

, et al. 2014. Follicular and Hurthle cell carcinoma of the thyroid in iodine-sufficient area: retrospective analysis of Korean multicenter data. Korean J Intern Med, 29:325–333.

24.

Welch Dinauer

, Tuttle

, Robie

, McClellan

, et al. 1998. Clinical features associated with metastasis and recurrence of differentiated thyroid cancer in children, adolescents and young adults. Clin Endocrinol (Oxf), 49:619–628.

25.

D'Avanzo

, Treseler

, Ituarte

PHG

, Wong

, et al. 2004. Follicular thyroid carcinoma: histology and prognosis. Cancer, 100:1123–1129.

26.

Cooper

, Doherty

, Haugen

, Kloos

, et al. 2009. Revised American Thyroid Association management guidelines for patients with thyroid nodules and differentiated thyroid cancer. Thyroid, 19:1167–1214.

27.

Haugen

, Alexander

, Bible

, Doherty

, et al. 2016. 2015 American Thyroid Association management guidelines for adult patients with thyroid nodules and differentiated thyroid cancer: the American Thyroid Association guidelines task force on thyroid nodules and differentiated thyroid cancer. Thyroid, 26:1–133.

28.

Huang

, Hsueh

, Liu

, Chao

, et al. 2011. Diagnostic and therapeutic strategies for minimally and widely invasive follicular thyroid carcinomas. Surg Oncol, 20:1–6.

29.

Feldt-Rasmussen

. 2001. Iodine and cancer. Thyroid, 11:483–486.

30.

Grani

, Lamartina

, Durante

, Filetti

, et al. 2018. Follicular thyroid cancer and Hürthle cell carcinoma: challenges in diagnosis, treatment, and clinical management. Lancet Diabetes Endocrinol, 6:500–514.

31.

Kakudo

, Bychkov

, Bai

, Li

, et al. 2018. The new 4th edition World Health Organization classification for thyroid tumors, Asian perspectives. Pathol Int, 68:641–664.