A Comprehensive Study on the Diagnosis and Management of Noninvasive Follicular Thyroid Neoplasm with Papillary-Like Nuclear Features

Introduction

In 2016, a subset of tumors previously classified as encapsulated follicular variant of papillary thyroid carcinoma (EFVPTC) was designated as noninvasive follicular thyroid neoplasm with papillary-like nuclear features (NIFTPs). This revised nomenclature from a carcinoma to low-risk neoplasm followed convincing data using strict, but evolving diagnostic criteria and >10 years of clinical follow-up in the initial retrospective cohort. ¹ Despite some subjectivity regarding a diagnostic threshold of features to qualify as NIFTP, published data have confirmed limited biological potential when strict criteria are met. Given the novel NIFTP designation and concern for long-term biological potential, its validity continues to be closely scrutinized. ² To date, most published studies have retrospectively assessed EFVPTC as a surrogate for NIFTP, despite more stringent required inclusion criteria for its diagnosis. ³

In assessing the notion that NIFTP have exceptionally low long-term biological potential, most published studies, except for two recent studies, ^4,5 have included cases diagnosed as EFVPTC from before 2016, despite discordant diagnostic criteria, ^{6 –12} and those patients were treated according to their diagnosis. Prospectively developing care plans for patients initially diagnosed with NIFTP, after 2016, requires an approach more in line with the revised conservative treatment recommendations from the American Thyroid Association (ATA) for well-differentiated thyroid cancers promoting de-escalation, including active surveillance. ¹³

Two recent studies evaluated NIFTP cases since 2016 and exclusively diagnosed as NIFTP. ^4,5 Each offers unique insight into the evolution of NIFTP, although one study included cases classified as NIFTP with BRAF p.V600E ⁴ and the other included cases with concomitant lesions. ⁵ A pure NIFTP cohort, with attempted elimination of confounding neoplasia, is desired to assess isolated tumor behavior over time. These cases may be compared in parallel with those having concomitant low-risk and higher-risk thyroid lesions to avoid cohort bias. Stratifying NIFTP-associated carcinomas into low-risk and higher-risk biological scenarios may help clinicians to better counsel patients for optimal therapeutics and surveillance protocol development. In a large institutional cohort, we studied the clinical, pathological, and molecular characteristics of 319 patients diagnosed with NIFTP and focus on a subgroup of 183 NIFTP with no associated carcinoma.

Materials and Methods

Results

The clinicopathological and molecular characteristics of all NIFTP groups are demonstrated in Table 1. In our full cohort of 319 patients, 23% presented with multiple nodules (n = 73) and 12% of NIFTP were multifocal (n = 39). FNA was performed on 80% of NIFTP specimens and was either nondiagnostic or benign in 18% of cases (n = 46; 12 were nondiagnostic and 34 were benign). Around half of NIFTP were Bethesda category III (atypia of undetermined significance/follicular lesion of undetermined significance; AUS/FLUS; n = 125).

Seventeen percent of cases were designated as Bethesda IV (follicular neoplasm/suspicious for follicular neoplasm; FN/SFN; n = 43). In 16% of cases (n = 42), NIFTP were designated as either suspicious for malignancy (n = 32) or malignant (n = 10). The histological, cytological, and immunohistochemical features of NIFTP, compared with circumscribed classic PTC with predominantly follicular growth, are demonstrated in (Fig. 1). The distribution of NIFTP cytological diagnoses since 2016 is demonstrated in (Fig. 2A). Thirty percent of biopsied NIFTP had FNA of a second nodule (n = 95). Bethesda categories for the second FNA are described in Table 1.

OPEN IN VIEWER

FIG. 1.

Comparison of NIFTP versus circumscribed classic PTC with predominantly follicular architecture. NIFTP is well circumscribed [200 × , H&E but unencapsulated (arrows, A), and in this example, 1.5 cm in size with NRAS p.Q61R]. Cytology of NIFTP with microfollicles and nuclear atypia (Bethesda IV; inset, 1000 × , A). Higher magnification of NIFTP (1000 × , H&E, B) highlights crowded nuclei with clearing and some grooves. Immunohistochemistry to assess for HBME1 (upper inset, 1000 × , B) showing a strong, membranous staining pattern and BRAF p.V600E (lower inset, 1000 × , B) is negative. Well-circumscribed (arrow heads) PTMC (0.4 cm) (200 × , H&E, C) shows minimal infiltration into adjacent thyroid parenchyma, has more pronounced PTC-like nuclear features (1000 × , H&E, inset, C) and is positive for BRAF p.V600E (200 × , BRAFVE immunostain, D), excluding an NIFTP diagnosis. BRAFVE, monoclonal BRAF p.V600E-specific antibody; H&E, hematoxylin and eosin; NIFTP, noninvasive follicular thyroid neoplasm with papillary-like nuclear features; PTC, papillary thyroid carcinoma; PTMC, papillary thyroid microcarcinoma.

OPEN IN VIEWER

FIG. 2.

NIFTP Diagnostic and surgical parameters. Retrospective distribution of NIFTP classification using The Bethesda System for Reporting Thyroid Cytopathology, 2016-6/2022 (A, n = 256) with distribution of molecular findings for the same cases by Bethesda category (B). Distribution of molecular findings, including specific findings from ThyroSeq and categorical findings from other gene classifiers (C). Surgical choice is shown in (D).

Ninety-three percent of molecular tests were performed for Bethesda III/IV (n = 106). The proportion of NIFTP with molecular testing and associated molecular alterations are shown in Figure 2B.

The distribution of NIFTP molecular alterations is shown in Figure 2C. Most NIFTP harbor RAS mutations (82%; n = 49), with NRAS p.Q61R being most common (n = 32). One patient had two bilateral nodules (Bethesda III) with subsequent ThyroSeq testing identifying NRAS p.Q61R mutation in both nodules. The second most common alteration was THADA fusion (n = 8) with two reported partners (IGF2BP3; n = 7 and TRA2A; n = 1). In 44 cases (42%), Afirma testing was suspicious with no subsequent XA. In groups 1 and 2, there has been a trend toward more HT since NIFTP emerged in 2016 (Fig. 2D).

In the thyroid, BRAF p.V600E is exclusive to a diagnosis of carcinoma. BRAFVE immunostain (Ventana, IN) was used to rule out BRAF p.V600E-mediated PTC and was negative in 99% of NIFTP (n = 193/195). In two cases, there was high background, including on repeat stain, but no BRAFVE positive case was detected. BRAFVE did not stain any of RAS-aberrant or the BRAF p.K601E NIFTP (n = 33 and n = 2, respectively).

In group 3 NIFTP, 19 patients presented with lymph node metastases, with 6 tested metastatic lymph nodes showing positive BRAFVE immunostain. Corresponding NIFTP in three of these patients were BRAFVE negative, and the associated PTC in four patients were BRAFVE positive (Table 2). In two NIFTP cases (5.5 and 3.7 cm) resected with HT, metastatic PTC was identified in the central compartment lymph nodes (both cases BRAFVE positive nodes), but no primary PTC within the thyroid lobe was found (the entire lobe was submitted in each case).

HBME1 (Ventana, IN) immunostain highlighted cells with nuclear membrane irregularities in both NIFTP (Fig. 1B, top inset) and PTC. HBME1 was positive in 91% of NIFTP (n = 148).

Discussion

NIFTP was first described in 2016 ¹ and recognized by the WHO in 2017. ²⁰ Although universally recognized by those involved in diagnosis and treatment of thyroid neoplasia, its novelty, ¹ along with modifying inclusion criteria, ^2,21,22 require increased data reporting. Lacking long-term outcomes data, NIFTP is described to be of low biological potential. ^{23 –25} Recently, the ATA has modified treatment recommendations regarding well-differentiated thyroid carcinomas to allow for treatment de-escalation and active surveillance for lower-risk carcinomas. ^13,26 As NIFTP is not classified as malignant, our data reinforce the need for new management protocols. ¹⁹ We subdivided NIFTP into three groups: group 1, NIFTP-only; group 2, NIFTP with a single PTMC; and group 3, NIFTP with a more significant associated carcinoma burden.

In one study, 29% of NIFTP were incidental, ²⁷ similar to the 20% in our cohort. Pathologist interobserver variability may also be a significant factor in NIFTP incidence but was beyond the scope of our study. Carcinoma has been associated with NIFTP in 15–46% of patients, ^{27 –29} and our cohort is similar (41%). Incidental PTMC did not change the prognosis in patients with NIFTP, similar to Taneja et al. that included PTMC in their NIFTP cohort. ³⁰ We considered NIFTP with concomitant lesions separately (Groups 2 and 3).

NIFTP was Bethesda category III in 49% of cases, consistent with prior reports, ^{31 –34} and cellular morphology alone is limited in differentiating NIFTP from follicular variant papillary thyroid carcinoma (FVPTC). ³⁵ BRAF p.V600E should exclude the diagnosis of NIFTP, a finding favoring PTC. ²²

Molecular studies are increasingly employed for surgical planning. ³⁶ In our cohort, NRAS p.Q61R is the most common genetic variant detected, over HRAS and KRAS, consistent with prior studies. ^28,33,37,38 BRAF p.K601E and THADA fusion are common in follicular-patterned tumors and are considered RAS-like mutations. ^{37,39 –41} Mutations in miRNA, copy number alterations, and DNA methylation can be present in NIFTP but with unclear significance.

Institutional management of encapsulated nodules, including NIFTP, is geographically variable (Fig. 3), a finding that may reflect cultural approaches to nodule management. Although surgical plans are multifactorial, for isolated NIFTP, a consensus approach should be considered to optimize cost, outcomes, and quality of life. ⁴²

OPEN IN VIEWER

FIG. 3.

Distribution of surgery for NIFTP in retrospective studies.

Although HT appears sufficient treatment for NIFTP-only and NIFTP with low-risk carcinomas, several patients in our cohort underwent TT (90/236). Indication for each TT is not further commented on in this study, but unlikely considered for NIFTP in isolation, as multivariable regression shows the decision to undergo TT was independently associated with the FNA result (favoring TT in Bethesda V/VI), FNA of an additional, non-NIFTP nodule, or preoperative aberrant thyroid function.

For 68 patients (37.8%), providers document that molecular testing of the dominant nodule informed their surgical decision-making, and the majority of NIFTP were characterized as Afirma suspicious. While genetic testing is nonstandard for Bethesda V/VI lesions, indeterminate categories (Bethesda III/IV) are often sent for molecular evaluation. As genetic testing has become standard on FNA material, and as costs for testing decline, more routine testing may be considered beyond the indeterminate FNA setting and extend to lesions with suspected malignancy to facilitate longitudinal management. ⁴³

Our NIFTP-only cases (Group 1) had a median follow-up of 35 months and our NIFTP plus PTMC cases (Group 2) had a median follow-up of 46 months. No recurrences or metastases were reported for NIFTP, and partnered with earlier studies with >10 years of follow-up of noninvasive encapsulated FVPTC, our study affirms lack of NIFTP recurrence or metastasis. ^{1,4,6 –8,11,12}

Postoperative follow-up for NIFTP patients has been variable. Surveillance US, 80 in total, was performed on 52 patients in group 1, those with NIFTP-only, including patients who had undergone TT HT with no contralateral nodule. The individual reasons for postoperative US in the 15 patients in Group 1 who underwent TT have not been scrutinized.

Our study has several caveats, and most glaring is the short clinical follow-up period given the typically long-term view of low-risk thyroid neoplasia. Given all cases reported in this study were diagnosed since NIFTP appeared in the literature, the maximum follow-up can be 6 years. ^26,44 Our cohort has no pediatric patients, and, in part, may be due to selection bias from our hospital's predominantly adult patient population, although consistent with Taneja et al., ³⁰ where the youngest reported patient was 21 years. Other studies have found NIFTP in children and adolescents. ^{45 –48}

Protocols for postoperative management of isolated NIFTP patients should be in line with those for other biologically low-risk thyroid neoplasias. Follicular adenomas and tumors categorized by the WHO as uncertain malignant potential are the most equivalent entities to NIFTP, architecturally and molecularly. With our current understanding of NIFTP, including the data presented in this study, we propose a postoperative management algorithm (Fig. 4). NIFTP alone should result in routine clinical follow-up. NIFTP plus any carcinoma should follow established management protocols for the coincidental lesion, from routine follow-up for incidental ipsilateral microcarcinomas to more lesion-specific protocols in all other cases.

OPEN IN VIEWER

FIG. 4.

Postoperative management algorithm after NIFTP diagnosis. Tumors are divided into Groups 1–3: Group 1 being NIFTP-only, Group 2 is NIFTP plus a single PTMC, and Group 3 is NIFTP plus additional malignancies (multiple microcarcinomas, PTC >1.0 cm, non-PTCs). Group 1 patients and patients in Group 2 with incidentally detected single foci of PTMC should be followed routinely, as any patient diagnosed with a benign thyroid tumor would be followed. Patients in Group 2 with nonincidental PTMC should be followed with active surveillance. Patients in Group 3 should be followed based on their concomitant disease process, indifferent to NIFTP, some with active surveillance and some with additional treatment.

Conclusion

NIFTP, so-named due to features of circumscription (no invasion), follicular-patterned architecture and atypical nuclear features, all seem to have variant RAS or RAS-like genetics and behavior. The historical binary concept of either benign or malignant is challenged with increased frequency, and, at least for clinical management purposes, our data favor the drift of isolated NIFTP toward a benign diagnosis, knowing that any true neoplastic process is inherent with at least some biological risk. As the ATA guidelines for management of low-risk malignancies now favors a de-escalated approach, guidance regarding that for borderline/biologically uncertain tumors, including NIFTP, is a reasonable next step.