Clinical prognostic risk assessment of different pathological subtypes of papillary thyroid cancer: a systematic review and network meta-analysis

Abstract

There are multiple pathologic subtypes of papillary thyroid carcinoma (PTC), each with distinct clinical prognoses. However, the available data on the clinicopathologic risks associated with several common PTC subtypes are controversial and require more comprehensive evaluation. To address this, we conducted a systematic search of English-language databases, including PubMed, EMbase, Cochrane Library, and Web of Science, for studies on six PTC subtypes, including classic papillary thyroid carcinoma (CPTC), papillary thyroid microcarcinoma (PTMC), follicular variant of papillary thyroid carcinoma (FVPTC), tall cell variant of papillary thyroid carcinoma (TCVPTC), diffuse sclerosing variant of papillary thyroid carcinoma (DSVPTC), and columnar cell papillary thyroid carcinoma (CCVPTC). Our case-control study of clinicopathological prognostic analyses of six subtypes, with a search date of January 2000 to May 2024. Two researchers independently screened the literature, extracted data, and assessed quality and risk of bias according to set criteria. R software gemtc package, Stata 15.1 software were applied to perform reticulated Meta-analysis methods were applied to compare the clinicopathological features and prognostic assessment of classic papillary thyroid carcinoma and the other five subtypes in all the studies.The risk of distant metastasis was higher in patients with CCVPTC, TCVPTC, and DSVPTC than in CPTC.FVPTC and PTMC exhibit a lower risk of in situ tumor relapse compared to CPTC. The tumour size of TCVPTC was significantly larger than that of CPTC, while there was no significant difference in the tumour size of CCV, DSV, FVPTC, CPTC and PTMC. DSVPTC was significantly more multifocal than the other subtypes. This network meta-analysis confirms the aggressive biological behavior and poor prognosis associated with TCVPTC, DSVPTC, and CCVPTC. Therefore, these subtypes should be managed aggressively with total thyroidectomy and lymph node dissection if diagnosed preoperatively. In contrast, FVPTC and PTMC are less aggressive and have a better prognosis, suggesting that treatment and follow-up strategies for PTC should be tailored according to the histopathological subtype.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00423-025-03841-2.

Introduction

Papillary thyroid carcinoma (PTC) is the most common endocrine malignancy, accounting for 85–90% of all thyroid cancers. PTC generally has a favorable prognosis, with a 10-year survival rate exceeding 95% [1]. However, approximately 19% of patients experience recurrence after initial surgery, suggesting that PTC is a heterogeneous group of tumors with widely varying macroscopic and microscopic features, leading to different clinical outcomes [2]– [3]. According to the World Health Organization (WHO) tumor classification, there are more than ten distinct pathological subtypes of PTC, each characterized by unique histomorphometric features. The most common subtype is the conventional type of PTC, while other notable subtypes include papillary thyroid microcarcinoma (PTMC), follicular variant of PTC (FVPTC), diffuse sclerosing variant of PTC (DSVPTC), tall cell variant of PTC (TCVPTC), and columnar cell variant of PTC (CCVPTC) [4].Papillary thyroid microcarcinoma (PTMC), defined as papillary carcinoma with a tumor diameter of less than 1 cm, is the most common subtype of PTC. Although most PTMC cases have a good prognosis, a small number of patients with PTMC combined with other poor prognostic subtypes exhibit worse outcomes. The follicular variant of PTC (FVPTC), first described by Lindsay, is now widely recognized as a distinct subtype, accounting for 23–41% of hyperdifferentiated thyroid cancers. Studies on FVPTC have shown conflicting results: Lang et al. [5] reported that patients with FVPTC have more favorable clinicopathological and biobehavioral characteristics compared to those with conventional PTC (CPTC), while Hagag et al. [6] found that FVPTC could be more aggressive than CPTC in long-term studies. Despite these differences, current treatment and prognosis for FVPTC and CPTC are similar [7].The diffuse sclerosing variant of PTC (DSVPTC) is a rare subtype, accounting for less than 5% of all PTC cases, and it predominantly affects younger patients. Originally described by Vickery et al. in 1985 [8], DSVPTC is characterized by diffuse invasion of one or both thyroid lobes, often without the formation of a distinct mass. Fujimoto et al. [9] treated 14 female patients with a mean age of 20 years and found no tumor-specific deaths, suggesting that DSVPTC can be a curable disease, similar to CPTC. However, Carcangiu and Bianchi [10] reported a higher incidence of neck and lung metastases and a decreased disease-free survival rate in DSVPTC patients.Tall cell variant PTC (TCVPTC), first reported by Hawk and Hazard, represents approximately 6% of PTC cases [11]. TCVPTC is characterized by a later age of onset and larger tumor size. Studies comparing TCVPTC and CPTC patients have consistently shown that TCVPTC is associated with higher disease stages and grades, and exhibits greater aggressiveness [12]. Consequently, some researchers advocate for a more aggressive treatment approach for TCVPTC patients [13]– [14].On the contrary, other studies have found no significant difference in prognosis between TCVPTC and CPTC patients with the same tumor stage, thus suggesting that the treatment of TCVPTC patients should be individualized according to the treatment criteria for CPTC patients [15]. CCVPTC is classified as one of the special histologic subtypes by WHO due to its specific cell type, which is of low incidence and clinically rare, accounting for 0.15%−0.2% of all papillary thyroid carcinomas [16]. Some studies have shown that CCVPTC is more aggressive than CPTC; however, others have suggested a more inert biological behavior. The biological behavior and prognostic implications of these histologic subtypes remain controversial [16–18].

Material and methods

Search strategy and study identification

A comprehensive search was conducted using the PubMed, EMBASE, Cochrane Library, and Web of Science databases. The search terms included “papillary thyroid carcinoma,” “papillary thyroid tumor,” “subtype,” “microscopic carcinoma,” “follicular,” “diffuse sclerotic,” “hypercellular,” “columnar cellular,” “clinicopathology,” and “histopathology.” In addition to database searches, we performed manual searches and examined the citations of included publications and reviews. Search terms were refined and adjusted based on initial search results to ensure comprehensive coverage. The search period spanned from January 2000 to June 2024.

Included and excluded criteria

Two reviewers (A and B) independently screened titles and abstracts using predetermined selection criteria. We used the following inclusion criteria: (i) the type of study was a controlled study of the biological behaviour and prognosis of two or more papillary thyroid carcinomas of different subtypes, with no restriction in terms of randomisation method, blinding, or allocation concealment; (ii) the language of the literature was restricted to English; (iii) there were no restrictions on age, gender, ethnicity, or duration of disease, but comparable baselines were required; and vi) the efficacy criterion was that the outcome metrics included an overall effective rate or. Exclusion criteria were (i) datasets that were considered overlapping, missing or erroneous data, (ii) control groups that did not have more than two comparable PTC subtype groups, and (iii) literature in the form of systematic evaluations, meta-analyses, guidelines, reviews, experimental studies, conferences, books, posters, and so on. Disagreements between the two reviewers were resolved by discussion led by another expert.

Data extraction

Two reviewers (A and B) read through the full text of the articles and extracted the data in tabular form according to the information design table. If there were differences in the results, the reviewers and experts resolved them through discussion.

Statistical analysis

The study followed the Priority Reporting Entry for Systematic Evaluation and Net Meta Analysis (PRlSMA Extension for Network Meta analysis) for the study. Network Meta-analysis was performed by applying gemtc package of R software, Stata 15.1 software to plot the reticulogram, prediction intervals, and comparative-corrected funnel plots of publication bias, and performing Network Meta-analysis method to compare the clinicopathological features and prognostic assessment of papillary thyroid carcinoma classic and the other 5 subtypes in all the studies. The quality of all included studies was assessed using the Newcastle-Ottawa Scale (NOS), independently assessed by two authors (Supplementary Material S2).

Results of GRADE quality of evidence evaluation

The quality of evidence for the outcome indicators of the included studies was graded using the GRADE system, in which the following five factors may affect the quality of evidence of a study: (1) risk of bias; (2) inconsistency; (3) imprecision; (4) indirectness; and (5) publication bias. The level of evidence was evaluated as high quality (no downgrade), moderate quality (downgrade 1), low quality (downgrade 2), and very low quality (downgrade 3). The results showed a total of five outcome indicators for the reticulation meta, and the overall quality of evidence for the outcome indicators was medium- to high-quality.The results of the GRADE quality evaluation are presented in Supplementary Material S12.

Result

Description of included studies

As of May 30, 2024, a total of 1734 articles were retrieved using the above search strategy. Among these studies, 22 studies met the eligibility criteria, and the detailed procedure of study retrieval screening is shown in Fig. 1. Overall, 43,0008 patients with CPTC, 46,561 with FVPTC, 534 with PTMC, 2,470 with TCVPTC, 718 with DSVPTC, and 1,753 with CCVPTC were included. Table 1 summarises the clinicopathological characteristics of the patients in our meta-analysis.All the included articles were assessed to be of moderate quality.(Supplementary Material S3).

Fig. 1.

PRISMA flowchart. PRISMA = Preferred Reporting Items for Systematic Reviews and Meta-Analyse

Table 1.

Characteristics of the 22 included studies

FVPTC indicates follicularvariant of papillary thyroid cancer; PTMC indicates papillary thyroid microcarcinoma; TCVPTC indicates tall cell variant; DSVPTC indicates diffuse cell variant; and PTMC indicates papillary thyroid microcarcinoma. DSVPTC indicates diffuse sclerosing variant of papillary thyroid cancer; CCVPTC indicates columnar cell variant of papillary thyroid cancer; “NR” stands for “Not Reported”.

Results of reticular Meta-analysis of biological behaviour and clinicopathological prognosis of PTC subtypes

• 3.2.1 In terms of focal multifocality: the reticulated Meta produced a total of 15 two-by-two comparisons (Supplementary Material S4), of which DSVPTC had the highest probability of developing focal multifocality (Table 2), DSVPTC (OR = 1.62,95% CI (1.14,2.31)) was significantly more likely than CPTC, and PTMC had the lowest probability of focal multifocality occurring (Supplementary Material S5).

• 3.2.2 In terms of tumour size, TCVPTC tumour diameter was significantly larger than the individual subtypes (Table 2), followed by CCVPTC, DSVPTC, FVPTC, CPTC, and PTMC tumour diameter was the smallest.The tumour diameter of TCVPTC was significantly larger in comparison to CPTC (WMD = 0.90 cm, 95%CI (0.02,1.78)) (Supplementary Material S5).

• 3.2.3 In terms of tumour-specific mortality, with the exception of PTMC for which no relevant data were available, the reticulation meta for the other subtypes yielded a total of 10 two-by-two comparisons (Supplementary Material S4), with TCVPTC having the highest tumour-specific mortality rate (Table 2), followed by DSVPTC, CCVPTC, CPTC, FVPTC. compared to CPTC, TCVPTC had a significantly higher tumour-specific mortality rate (OR = 5.94,95% CI (1.75,20.10)) and FVPTC had the lowest tumour-specific mortality rate (Supplementary Material S5).

• 3.2.4 In terms of in situ recurrence outcomes, this was reflected in a star-shaped structure centred on CPTC (Supplementary Material S4). The results of reticulated Meta-analysis showed that DSVPTC had the highest rate of in situ recurrence (Table 2), followed by TCVPTC, CPTC, FVPTC, and PTMC tumours had the lowest probability of in situ recurrence (Supplementary Material S5).

• 3.2.5 In terms of tumour distant metastasis: the analysis showed that CCVPTC had the highest rate of distant metastasis (Table 2), followed by TCVPTC, DSVPTC, CPTC, and FVPTC had the lowest rate of metastasis. CCVPTC, TCVPTC, and DSVPTC had higher rates of distant metastasis (OR = 4.03, 95% CI (2.62,6.20); OR = 2.52, 95% CI (2.08,3.04); and OR = 2.23, 95% CI (1.43,3.49)) relative to CPTC tumours, respectively (Supplementary Material S5).

Table 2.

Ranking tables with respect to (A) Incidence of multiple lesions, (B) Tumor size, (C) Tumor-specific mortality, (D) Tumor recurrence and (E) Tumor mortality. Tumor recurrence and (E) Distant tumor metastasis

Conclusions

This network meta-analysis confirmed the aggressive biological behavior and poor clinical prognosis of three subtypes of papillary thyroid carcinoma: tall cell variant (TCVPTC), diffuse sclerosing variant (DSVPTC), and columnar cell variant (CCVPTC). Patients diagnosed preoperatively with these subtypes should be considered for aggressive treatment strategies, including total thyroidectomy with lymph node dissection. In contrast, the follicular variant (FVPTC) and papillary thyroid microcarcinoma (PTMC) exhibit low invasiveness and are associated with a good prognosis. Therefore, the treatment strategy and follow-up plan for papillary thyroid carcinoma (PTC) should be tailored according to the histopathological subtype.

Discussion

Thyroid cancer is the fastest growing solid malignancy in terms of incidence, with papillary thyroid carcinoma being the most common, and the vast majority of papillary carcinoma subtypes can be treated with surgery and radioactive iodine with a good prognosis. Invasive PTC is relatively rare among PTCs and is usually associated with a poorer prognosis, which also contributes to the wide variation in the prognosis of PTCs. Clinicopathological risk stratification of papillary thyroid carcinoma subtypes is rare and controversial, resulting in a lack of effective treatment guidelines.

The aim of this study was to compare the clinical prognostic risk assessment of different pathological subtypes of papillary thyroid carcinoma, and we included a total of 22 retrospective controlled studies that assessed the prognosis based on the number of lesions, in situ recurrence of disease, distant metastasis, and cancer-specific mortality, and tumour size. The results of reticulated Meta-analysis showed that patients with TCVPTC, DSVPTC, and CCVPTC exhibited a higher risk of distant metastasis and tumour-specific mortality.FVPTC and PTMC had a lower risk of in situ tumour recurrence compared to CPTC.TCV tumour diameters were significantly larger than the various other subtypes, CPTC, FVPTC, DSVPTC, and CCVPTC, PTMC did not differ significantly in tumour diameter between them. In terms of tumour multifocality, DSVPTC patients exhibited a higher incidence of multifocality. In summary, we concluded that TCVPTC, DSVPTC, and CCVPTC exhibited more aggressive and poor prognosis, while FVPTC and PTMC had relatively better biological behaviours and better prognosis than CPTC.

The incidence of PTMC is rapidly increasing, accounting for 27.8% of PTC, and is the second most common papillary carcinoma after the conventional type [19]. With a 10-year survival rate of more than 90%, it is usually considered a pathological subtype with a benign biological behaviour. According to our findings, patients with PTMC have a lower incidence of multifocal disease than CPTC, and their tumour recurrence rate is not significantly different from that of CPTC. What’s more, some studies indicated that even in the presence of lymph node metastasis and local infiltration, the prognosis of PTMC patients is usually favourable [20]– [21]. Therefore, in general, a conservative approach should be taken in the management of patients with PTMC.

Some studies have found that lung metastasis is more common in FVPTC than in CPTC [22],and lymph node metastasis in FVPTC may be associated with activation of p-JNK, which accounts for 28.3% of FVPTC [23]. However, other studies have not found a significant difference in clinical prognosis between these two subtypes [5]. Our data suggest that FVPTC has a lower probability of in situ recurrence, and both distant metastasis and tumour-specific mortality are significantly better than the other subtypes. This is consistent with previous findings that the classical variant of PTC has a much higher incidence of both central interstitial lymph node involvement and lateral cervical lymphadenopathy [24, 25]. Whilst studies of the exact relative risk of lymph node metastasis between these two most common types of papillary cancer vary, the common conclusion is that the risk of such progression is much smaller in FVPTC. According to the latest edition of WHO staging in 2017, which classified some of the original FVPTC as NIFTP, it was previously noted in the literature that with the implementation of a stricter definition of FVPTC, the risk of lymph node metastasis rates between FVPTC and tumours with a classical PTC component were more similar than previously published. However, as the diagnosis of NIFTP currently relies solely on histological criteria, such a distinction cannot currently be made by imaging or fine needle aspiration. Although it is possible to identify papillary cell nuclei in NIFTP lesions by FNA techniques, it cannot reliably differentiate between NIFTP and invasive FVPTC [26]. Thus, the data included in this paper for inclusion included cases of NIFTP to better match the actual clinical situation. The current literature agrees that there is a status quo of overtreatment of FVPTC [27]– [28].

DSVPTC is a rare but distinct subtype of thyroid cancer characterised by the production of fibrotic sclerotic material by the cancer cells, which allows the tumour to show infiltrative growth and distinctive imaging features.DSVPTC is usually found in younger patients, and the prevalence is significantly higher in women than in men [29]. Clinically, DSVPTC often presents as a thyroid nodule or mass, and imaging studies such as ultrasound, CT and MRI can help assess the extent and aggressiveness of the tumour. Diagnosis relies on postoperative pathology, with typical papillary structures and fibrotic collagenous stroma seen on microscopic examination. Our data suggest that DSVPTC tumours have the highest risk of recurrence. Although DSVPTC is relatively rare, the risk of postoperative recurrence and metastasis is high due to its aggressive growth characteristics. Therefore, clinicians need to closely monitor patients’ recurrence and develop individualised treatment plans. In the future, targeted therapy and individualised treatment strategies for DSVPTC will be a key direction for research.

CCVPTC is more aggressive than CPTC, has a poorer overall survival rate, and has a higher risk of distant metastasis [30]. To address the risk of distant metastasis, it is recommended that patients with CCVPTC receive regular follow-up and monitoring by their physicians, and undergo timely imaging to assess disease progression in order to develop optimal treatment and management strategies.

TCVPTC, DSVPTC and CCVPTC are all considered aggressive subtypes, with TCVPTC being more common than the latter two.There is no significant change in the pathological diagnostic criteria between DSVPTC and CCVPTC, and TCVPTC has been adapted in the latest version of WHO staging. The TCVPTC cases included in this paper all conformed to the latest version of WHO staging, i.e., the percentage of tall cells was > 30%, the height of the cells was two to three times their width, and they showed cytological features of PTC. It has been shown that TCVPTC patients have the highest prevalence of lesions with late AJCC staging (T and N stages) and the worst 5-year prognosis for OS and DSS. The results of our Meta-analysis study showed that TCVPTC had the highest tumour-specific mortality rate among the five subtypes (PTMC has no relevant data), and the rate of distant tumour metastasis was comparable to that of CCVPTC, which demonstrated a higher metastasis rate among the five subtypes (PTMC has no relevant data), with the rate of tumour recurrence ranking second only to the rate of DSVPTC ranked second.The comprehensive analysis revealed that the biological behavior of TCVPTC, DSVPTC, and CCVPTC is more aggressive compared with CPTC, and their prognosis is significantly worse, which is consistent with the findings of most published studies. Interestingly, some studies found no significant difference in prognosis between TCVPTC patients who did not show aggressiveness and CPTC patients, and thus they concluded that the traditional PTC diagnosis and treatment could be used for such TCVPTC patients [15]. In the present study, we believe that, according to the biological behaviour of TCVPTC, surgical treatment should be performed aggressively in order to avoid the occurrence of poor long-term prognosis. This study concluded that aggressive treatment is recommended for these three subtypes.

The main limitation of this meta-analysis is that the included studies were retrospective, some pathological features were not specifically described, and classification criteria were not clearly indicated. It is recommended that the subtypes of papillary thyroid carcinoma should be specifically recorded, especially the more aggressive subtypes should be described with detailed pathological and histological results, so that the diagnostic classification criteria and prognostic analysis can be better studied.

In summary, based on the clinicopathological features and prognosis, this study concluded that different treatment strategies should be adopted for different PTC subtypes, and aggressive treatment should be adopted for TCVPTC, DSVPTC, and CCVPTC, while over-treatment is not recommended for FVPTC and PTMC subtypes, which have a better prognosis compared to CPTC.

Most of the available literature are retrospective studies, and due to the inert characteristics and good long-term prognosis of papillary thyroid carcinoma, the extremely low disease-specific mortality rate during the current follow-up period, and the lack of overall survival statistics, we did not perform OS analyses or plot the Kaplan-Meier survival curves, and we look forward to more RCTs with large sample sizes, longer follow-up times, multicentre, and high quality. RCTs to fit a more stable network model, as well as to establish a reliable OS assessment system to provide a favourable clinical reference.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Jiayi Zhao(Author 1) :(co-first author): Conceptualization, Methodology, Software, Investigation, Formal Analysis, Writing - Original Draft; Wanchen Zhang(Author 2) :(co-first author): Conceptualization, Methodology, Software, Investigation, Writing - Original Draft; Dongning Lu(Author 3): Data Curation, Writing - Original Draft; Chengying Shao(Author 4): Visualization, Investigation; Yiwei Chen(Author 5):Resources, Supervision; Xingyu Huang(Author 6):Software, ValidationYining Zhang(Author 7): Visualization, Writing - Review & EditingJiajie Xu(Author 8) (Corresponding Author): Conceptualization, Funding Acquisition, Resources, Supervision, Writing - Review & Editing.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Competing interests

The authors declare no competing interests.