Introduction

Hürthle cell carcinoma (HCC) of the thyroid gland is a rare neoplasm which accounts for about 0.4- 10% of all differentiated thyroid cancers^1-3. This tumor in more prevalent in women (female-to-male ratio 3:1) between the fifth to seventh decade of life. Although the World Health Organization considers HCC a variant of follicular carcinoma, the information available on its prognosis compared with follicular carcinoma is not uniform. Hürthle cell carcinoma has been considered by some authors to be a distinctive subset of differentiated thyroid cancer based on its oncogene expression and specific molecular pathogenesis⁴.

The diagnosis is often made by fine needle aspiration (FNA) biopsy with cytological features suggesting follicular or Hürthle cell neoplasm. Oncocytic changes are defined as cellular enlargement characterized by an abundant cytoplasm because of the accumulation of altered mitochondria. However, FNA cannot differentiate between adenomas and carcinomas. Of all patients with a FNA biopsy consistent with a follicular or a Hürtle cell neoplasm, only 15% to 30% will have a carcinoma⁵.

Malignant Hürthle cell neoplasms are composed of 75% or greater Hürthle cells, and vascular and capsular invasion are the characteristics that differentiate it from an adenoma⁶. These characteristics are easy to identify before the definite surgical treatment and even in the intraoperative frozen section. For this reason, attempts have been made to identify prognostic factors of malignancy to develop more appropriate treatment algorithms. The risk factors used to predict the outcome of other differentiated thyroid tumors include age, sex, tumor size, extrathyroid extension, lymph node involvement, and distant metastases; however, their influence on HCC has not been conclusively established.

The treatment of this disease is controversial since there are no prospective randomized controlled trials validating the extent of thyroidectomy and the efficacy of adjuvant treatment with radioactive iodine or radiation therapy^5,4. Total thyroidectomy is the mainstay of treatment for HCC as it is multifocal in 15% to 35% of cases³. Elective ipsilateral central neck lymph node dissection should be considered as lymph node metastases may be present in 20-35% of the cases at the moment of diagnosis⁵. Hürthle cell cancer has a lower avidity for 131I; therefore, treatment with radioactive iodine remains unclear. 131I total-body scanning is recommended six weeks after surgery and ablation of any thyroid remnant with radioactive iodine also should be considered to facilitate the use of serum thyroglobulin during follow-up. Some studies have suggested that radiation therapy to the neck could improve survival³.

The largest series of patients with HCC do not exceed 200 cases due to the low prevalence of the tumor. There are no prospective studies analyzing the different possible treatments. The guidelines published by the American Thyroid Association (ATA) and the National Comprehensive Cancer Network (NCCN) establish different treatment algorithms. The ATA suggests radioactive iodine after total thyroidectomy although it may not be necessary in tumors < 1 cm. The NCCN recommends radioactive iodine for tumors > 2 cm or high serum Tg level or presence of histopathological criteria².

The aim of this study is to perform a systematic review and meta-analysis of the risk factors and prognostic factors affecting survival of patients with HCC.

Material and methods

Selection of relevant studies and data extraction

Two reviewers (VSM and GC) independently reviewed the articles after the literature search according to the pre-established exclusion and inclusion criteria. Disagreements were resolved by a third reviewer (OGA). Data extracted from both reviewers included the following variables: age, sex, tumor size, extrathyroid extension, lymph node involvement, distant metastasis, surgical treatment, radioactive iodine treatment and radiation therapy. The outcome measure evaluated was disease-specific survival.

The PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) checklist was used to select studies to be included in the meta-analysis (Fig. 1).

Figure 1 Flow chart. Selection of articles 

The Newcastle-Ottawa Scale was used to assess the methodological quality of observational cohort studies.

Results

A total of 139 studies were retrieved from the primary search. Following review of titles and abstracts, 111 articles were excluded, leaving 28 full text articles for the review. Six studies fulfilled the inclusion criteria for the qualitative analysis and meta-analysis (Fig. 1). The quality of the studies included was between 7 and 6 according to the Newcasttle-Ottawa Scale (Table 1). The number of studies included in the meta-analysis varied between 3 and 5 depending on the variables analyzed in each one (Tables 2 and 3).

Table 1 Quality analysis of the studies according to the Newcastle-Ottawa Scale 

Table 2 Data of risk factors extracted from the studies included 

Table 3 Data of treatments extracted from the studies included 

A total of 4363 cases were analyzed in the 6 studies included; the data extracted and the number of patients per publication are shown in Tables 2 and 3. Survival rate at 10 years was reported in 3 studies and ranged between 64 and 71% (Table 4).

Table 4 Survival reported in the studies included 

Analysis of risk factors

There was a significant association between older age and worse outcome, but the analysis could not be made due to differences in how the variable was categorized in the different studies analyzed (Table 5). Male sex was significantly associated with lower survival in three studies. In our meta-analysis, male sex decreased survival by 2.11 times (95% CI, 1.66-2.69). A tumor size > 4 cm had a significant negative impact on survival in 5 studies with HR of 2.38 (95% CI, 1.16-4.88). Extrathyroid extension increased mortality by 3.56 times (95% CI, 1.74-7.26). Lymph node involvement and distant metastasis were associated with adverse outcome (HR 2.56, 95% CI, 1.77, 3.72, and HR 2.64, 95% CI, 2.24-7.03, respectively) (Fig. 2 to 6).

Table 5 Analysis of age as risk factor in the studies included 

Figure 2 Forest Plot. Sex 

Figure 3 Forest Plot. Tumor size 

Figure 4 Forest Plot. Extrathyroid extension 

Figure 5 Forest Plot. Lymph node stage 

Figure 6 Forest Plot. Distant metastasis 

Analysis of the different treatments

Four studies provided data comparing total thyroidectomy vs. partial thyroidectomy). Only Goffreddo and Petric found lower mortality associated with total thyroidectomy. Our meta-analysis shows a trend toward lower cause-specific mortality with a total effect of 0.56 (95% CI, 0.12-2.74; P = 0.12).

We did not observe a significant reduction in the risk of mortality with radioactive iodine and external radiation therapy (Figs. 7 to 9).

Figure 7 Forest Plot. Total thiroydectomy 

Figure 8 Forest Plot. Radioactive iodine treatment 

Figure 9 Forest Plot. Radiation therapy 

Discussion

Several articles have suggested that HCC had a worse outcome than follicular adenocarcinoma^4,7 due to its low radioactive iodine avidity and higher frequency of lymph node metastases. However, survival of HCC has increased over time because the histological diagnostic criteria for malignant thyroid tumors have changed over the past 35 years, decreasing the incidence of follicular carcinoma⁸. Other studies have reported opposite results with a worse outcome for HCCs⁹.

Some authors showed higher mortality rate in older patients^10,11 as opposed to other publications⁴. Age as a risk factor is considered in different ways by the studies analyzed^10,13-16; however, many of them, have found an increased risk of malignancy in patients > 45 years. Hürtle cell carcinoma is more common in men and usually presents worse outcome. In this study, male sex was significantly associated with lower disease-specific survival^17-19. While some studies suggested that HCC with a size > 4 cm have worse prognosis, other authors did not find a significant association^{4,15-17,20-22}. Tumor size has been analyzed in many ways, but the total effect shows a signification reduction in survival in those tumors > 4 cm. Most studies analyzed showed greater mortality associated with extrathyroid extension, lymph node involvement and distant metastasis^15,23-25. Our results are consistent with this hypothesis.

Other risk factors analyzed^4,12 as preoperative Tg levels¹, high BMI and a history of radiation therapy were associated with adverse outcome; yet, there are insufficient data to draw a conclusion.

Some authors have suggested that, in the presence of these risk factors, Hürthle cell tumors could be classified into two categories: low and high risk, and recommended more aggressive initial therapeutic approach for high risk malignancies before the pathological diagnosis of HCC. Other authors postulate that hemithiroidectomy should be the preferred initial surgical approach for the management of Hurthle cell neoplasms^15,26,27.

Surgery is the treatment recommended for HCC, but the extent of surgery is still controversial. Some authors recommend hemithyroidectomy in Hürthle cell tumors and, if malignancy is confirmed, thyroidectomy should be completed in high-risk patients^18,27. Instead, other authors recommend total thyroidectomy and ipsilateral central neck node dissection²⁵. In the present study, we observed a trend toward a therapeutic benefit of total thyroidectomy vs. lobectomy. However, we cannot conclude about the best surgical approach due to the high heterogeneity of the studies analyzed.

Despite HCC has low avidity for 131I, the efficacy of radioactive iodine is still a matter of debate^14,28. The meta-analysis did not find a clear benefit associated with this treatment. Tg level provides useful information on radioactive iodine ablation for postoperative follow-up.

Some authors have demonstrated that HCC is a radiosensitive tumor and recommend radiation therapy in patients presenting with symptomatic distant metastasis or unresectable disease in the neck and mediastinum, but its role in those tumors completely resected is not clear³. In this meta-analysis we did not observe a clear benefit in survival associated with postoperative radiation therapy.

Conclusions

Although the treatment of patients with HCC remains controversial, it is clear from the results obtained that male sex, tumor size, lymph node involvement, extrathyroid extension and distant metastases are undoubtedly the factors associated with adverse outcome.

In the case of treatment, we only found a trend toward better outcome with total thyroidectomy vs. lobectomy. We cannot draw any conclusions about the other factors due to differences in the approach of the study, and large-scale research on multimodal treatments is required.

The AACCyC (Asociación Argentina de Cirugía de Cabeza y Cuello) national registry is opened to all surgeons and its results, which will be available in 2023, will probably provide high quality evidence on the matter.

The authors are aware of the strengths and weaknesses of this presentation. This is the first study providing high level of evidence about prognostic factors and treatment of HCC. However, the degree of heterogeneity of the studies was very high for some variables and it was not possible to include in this analysis other risk factors that could affect the prognosis. We could not analyze a publication bias due to the few studies included, and we did not find prospective studies or clinical trials with a higher level of evidence to include in the meta-analysis.