Systematic review of PET imaging features and clinical applications in pulmonary NUT carcinoma

1 Introduction

Nuclear protein in testis (NUT) carcinoma is a rare and highly aggressive malignancy driven by NUTM1 gene rearrangements, most commonly with BRD4 (BRD4–NUTM1 fusion, −70%), followed by BRD3–NUTM1 (15%–20%) and NSD3–NUTM1 (−6%) (1, 2). Although initially described in midline structures, NUT carcinoma can arise at non-midline sites including the lung (3). The 2021 WHO Classification of Thoracic Tumors formally recognised pulmonary NUT carcinoma as a distinct entity (4).

Primary pulmonary NUT carcinoma is the most common thoracic manifestation, accounting for 51%–64% of cases (5). The median age at onset is 35–45 years; most patients are never-smokers or light smokers and present with nonspecific respiratory symptoms (6, 7). The clinical course is exceedingly aggressive, with median overall survival of only 4.4–6.7 months (7–9). No standardised therapeutic regimen currently exists, and early diagnosis remains the cornerstone for improving outcomes.

Diagnostic difficulty substantially contributes to the poor prognosis of pulmonary NUT carcinoma. Routine pathological examination is often insufficient because of morphological overlap with poorly differentiated squamous cell carcinoma, small cell carcinoma, and other round blue cell tumours (10). NUT IHC (clone C52B1) demonstrates near-perfect sensitivity and specificity (11), whereas standard DNA-based next-generation sequencing misses >75% of cases (12). Nearly 80% of patients are misdiagnosed initially, and −20% receive ineffective therapy before definitive diagnosis (13).

Clinically and pathologically, pulmonary NUT carcinoma presents a diagnostic challenge due to its morphological overlap with poorly differentiated squamous cell carcinoma, small cell lung cancer, and other round blue cell tumors (10). Distinctively, it is characterized by a lack of association with smoking in the majority of patients (over 70% are never-smokers), a younger median age at onset (35–45 years) compared to other lung cancers, and the pathognomonic NUTM1 rearrangement, most commonly involving BRD4. Nearly 80% of patients are misdiagnosed initially, underscoring the critical need for a high index of suspicion and specific immunohistochemical testing (NUT IHC) for accurate diagnosis (13).

¹⁸F-fluorodeoxyglucose positron emission tomography/computed tomography (¹⁸F-FDG PET/CT) is a molecular imaging modality that integrates anatomical and functional metabolic information. It has been widely applied in the initial staging, treatment response assessment, and recurrence surveillance of lung cancer. For pulmonary NUT carcinoma, which is characterised by poor differentiation, high aggressiveness, and early widespread metastasis, ¹⁸F-FDG PET/CT enables comprehensive, single-session assessment of the primary tumour burden, regional lymph node involvement, and distant metastatic dissemination, thereby overcoming the limitations of conventional imaging in detecting occult metastases (14). However, because of the rarity of primary pulmonary NUT carcinoma, no systematic review specifically addressing its ¹⁸F-FDG PET/CT imaging characteristics currently exists. The available literature consists predominantly of isolated case reports or small case series. A unified consensus on the typical imaging phenotype and differential diagnostic criteria has yet to be established, and systematic summaries of metabolic parametres, staging modification value, and efficacy for treatment response monitoring remain conspicuously lacking (6, 14, 15). This evidence gap leaves radiologists and clinicians without reliable imaging references when encountering suspected cases, potentially further delaying diagnosis.

This systematic review aimed to synthesize the ¹⁸F-FDG PET/CT imaging manifestations and clinical value of primary pulmonary NUT carcinoma, characterize its metabolic features and imaging phenotype, evaluate its staging value and metastasis detection efficacy, assess its utility for treatment response monitoring, and explore differential diagnostic points relative to common pulmonary malignancies.

2 Materials and methods

This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020) statement and was prospectively registered with PROSPERO (registration number: CRD420261292692). Because this study synthesized published literature, ethical approval and informed consent were not required.

2.1 Search strategy

A comprehensive literature search was performed in PubMed, Embase, Cochrane Library, and Web of Science Core Collection from database inception to March 20, 2026, with no language restrictions. Grey literature (ProQuest Dissertations and conference abstracts from ASCO, ESMO, and IASLC) and the reference lists of included studies and relevant reviews were also searched.

The core search strategy for PubMed (adapted for other databases) was: Studies that mentioned pulmonary primary cases in the full text but were not captured by #2 were identified through full-text review.

• #1 (“NUT carcinoma” OR “NUT midline carcinoma” OR “NUTM1-rearranged carcinoma” OR “NUTM1 fusion carcinoma”)
• #2 (“lung” OR “pulmonary” OR “bronchial” OR “respiratory”)
• #3 #1 AND #2

2.2 Eligibility criteria

2.2.1 Inclusion criteria

Study design: Cohort studies, case series, case reports, and diagnostic accuracy studies (prospective or retrospective). Clinical letters containing original case data with extractable outcomes were considered equivalent to case reports.

Participants: Pathologically confirmed primary pulmonary NUT carcinoma (tumour within lung parenchyma or bronchus) diagnosed by positive NUT immunohistochemistry (≥50% nuclear staining) and/or FISH/NGS detection of NUTM1 rearrangement. Studies with an ambiguous primary site were eligible if: (i) the dominant mass was located in the lung; (ii) there was no evidence of prior head, neck, or mediastinal involvement; and (iii) questionable cases were adjudicated by two investigators and labelled as “uncertain primary” for sensitivity analysis.

Intervention: ¹⁸F-FDG PET/CT examination.

Outcomes: At least one PET/CT-related outcome, including metabolic parametres, staging modification, metastasis detection, treatment response monitoring, or differential diagnostic features.

Full text accessible with complete extractable data.

2.2.2 Exclusion criteria

No definitive pathological confirmation of NUT carcinoma.

Mixed-cohort studies without extractable pulmonary-specific data.

PET/CT using non-¹⁸F-FDG tracers or studies lacking relevant PET/CT outcomes.

Reviews, commentaries, opinion letters without original data, editorials, animal/in vitro studies, and conference abstracts without adequate data.

Duplicate publications (only the most recent and complete version was included).

Extrapulmonary primary NUT carcinoma with lung metastases, or indeterminate primary site not meeting the operational definition above.

2.3 Study selection and data extraction

Two trained investigators independently screened studies in duplicate. After duplicate removal, titles and abstracts were screened, followed by full-text assessment of potentially eligible studies. Disagreements were resolved by a third reviewer. The selection process was documented using a PRISMA flow diagram.

Data were independently extracted using a standardised form and cross-checked by two investigators. Extracted information included: (1) study characteristics (first author, year, country, design, sample size); (2) patient demographics (age, sex, smoking history, fusion partner, stage, treatment); (3) PET/CT parametres (scanner, acquisition protocol, tracer dose); (4) outcome data (SUVmax values, staging modification rates, metastasis detection details, treatment response assessments, differential diagnostic features); and (5) quality assessment items.

Given the retrospective nature of the included studies, contacting authors for missing data was deemed infeasible. Outcomes with a reporting rate below 50% were described case by case rather than summarised.

2.4 Quality assessment

Study quality was independently assessed by two investigators using standardised tools: the Newcastle-Ottawa Scale (NOS) for cohort studies, QUADAS-2 for diagnostic accuracy studies, and JBI critical appraisal checklists for case series and case reports. Disagreements were resolved by a third reviewer. Quality assessment results informed sensitivity analyses and GRADE evidence rating.

2.5 Data synthesis

Given the extreme rarity of this disease and substantial inter-study heterogeneity, a structured qualitative synthesis was performed without meta-analysis. Continuous variables with reporting rates >80% were summarised using ranges and medians; otherwise, case-by-case description was used.

2.6 Publication bias assessment

Funnel plots and Egger's test were not planned because of insufficient statistical power given the limited number of studies and their design heterogeneity. Publication bias was assessed qualitatively by evaluating: (1) completeness of outcome reporting within individual studies; (2) inherent publication bias favoring unusual or positive findings in case reports; and (3) incorporation of these assessments into GRADE evidence rating.

2.7 GRADE evidence rating

Evidence quality was rated using the GRADE system (high, moderate, low, very low). Observational studies started as low quality; upgrade considerations included large effect size or dose–response relationships.

2.8 Sensitivity analysis

Sensitivity analyses were planned to verify the robustness by restricting to NGS-confirmed cases, excluding low-quality studies (NOS < 5 or high JBI risk of bias), and excluding uncertain-primary-site cases.

3 Results

3.1 Study selection

A total of 305 records were initially identified through database searches and reference list screening. After a stepwise selection process—duplicate removal (222 records remaining), title/abstract screening, and full-text review of 42 records—23 studies (6, 8, 14–34) ultimately met the eligibility criteria and were included in this systematic review, comprising one retrospective cohort study, five case series, and 17 case reports. The study selection process is illustrated in Figure 1 (PRISMA flow diagram).

3.2 Characteristics of included studies

The 23 included studies encompassed 86 patients. Ages ranged from 17 to 82 years (median, 35–45 years); −54% were male, and >70% were never-smokers or light smokers. All cases were confirmed pathologically, with BRD4–NUTM1 being the most common fusion partner (−65%). Detailed characteristics are presented in Table 1.

Table 1. Summary of included studies: basic characteristics and core PET/CT findings.

y, year; mo, month; IHC, immunohistochemistry; FISH, fluorescence in situ hybridization; NGS, next-generation sequencing; RNA-seq, RNA sequencing; PET/CT, positron emission tomography/computed tomography; FDG, fluorodeoxyglucose; LN, lymph node; PR, partial response; PD, progressive disease; EP, etoposide plus platinum; AFP, alpha-fetoprotein; NR, not reported; NUT, nuclear protein of the testis.

This cohort included patients with both pulmonary and mediastinal primary NUT carcinoma; the original article did not separately report the number of cases with pulmonary primary. Only PET/CT data relevant to the pulmonary subgroup were extracted for this review.

Of the 8 patients in this series, 4 underwent ¹⁸F-FDG PET/CT, whereas the remaining 4 were evaluated by CT only. Only PET/CT-related data were extracted for this review.

This study comprised two parts: the authors’ own case series (n = 6, of whom 1 had original PET/CT data) and a review of the literature. The table lists only PET/CT data from the authors’ series; pooled data from the literature review are described in Section 3.3 and Table 2 of the main text.

The primary site of this case (right hilum) is uncertain. In accordance with the study protocol, it was categorized as “uncertain primary” and excluded in sensitivity analyses to verify the robustness of the conclusions.

3.3 Metabolic characteristics on ¹⁸F-FDG PET/CT

3.3.1 Metabolic features of primary lesions

Twelve studies (52.2%) reported primary tumour SUVmax values. All showed markedly hypermetabolic features, with SUVmax ranging from 5 to 40 and mean concentrated between 12 and 14 (Table 2). Because the reporting rate was below the 80% threshold, data are presented descriptively.

Table 2. Summary of primary tumor SUVmax reported in the included literature.

SUVmax showed some heterogeneity, ranging from 6.8 (22) to 25.4 (19), with frequently heterogeneous metabolic distribution showing central necrotic zones and ring-like peripheral FDG avidity (8, 14, 23).

In summary, primary pulmonary NUT carcinoma shows: (1) markedly high FDG uptake (SUVmax 5–40; mean, 12–14) with heterogeneous uptake due to central necrosis; (2) synchronous high uptake in regional nodal metastases (SUVmax 60%–90% of primary); and (3) distant metastases with comparable metabolic levels.

It is important to note that while this triad is characteristic of many cases, it should not be used as an exclusion criterion for NUT IHC or genetic testing, as individual cases may present with atypical features.

3.3.2 Metabolic relationship between primary tumors and metastases

Eight studies concurrently reported metabolic activity of both primary tumours and metastatic sites (8, 14, 15, 22, 23, 26, 28, 29).

Regional nodal metastases exhibited synchronous high FDG uptake with the primary tumour, with SUVmax characteristically falling within 60%–90% of the primary lesion (Table 3). Distant metastases showed metabolic levels comparable to the primary tumour (23, 28). The occult metastases detected by PET/CT—leading to upstaging in 55% of patients—were significantly associated with poor prognosis (29).

Table 3. Comparison of SUVmax between primary lesions and matched metastases.

A total of 8 studies included in the systematic review concurrently reported the metabolic activity of primary lesions and matched metastases. Among them, 3 studies (Study 10, Study 20, and 1 additional study) provided only qualitative descriptions or pooled mean values without paired quantitative SUVmax data, and were therefore not included in this quantitative comparison table.

LN, lymph node; SUVmax, maximum standardized uptake value.

3.4 Characteristic imaging phenotype: A proposed “triad”

Primary pulmonary NUT carcinoma presents a characteristic phenotype on PET/CT, summarised as a suggestive triad: Additional features: The contralateral lung is mostly spared, and brain metastases at initial diagnosis are extremely rare (8, 15), possibly because the tumour progresses so rapidly that patients succumb before developing detectable brain metastases.

• Central, bulky mass in the hilar/mediastinal region with high metabolic activity: Most lesions are centrally located soft-tissue masses (5.0–8.6 cm) with frequent internal necrosis and heterogeneous enhancement (8, 23, 28). Even small lesions retain aggressive features (e.g., T1-stage nodule with SUVmax >10.8 (30).
• Ipsilateral mediastinal/hilar lymph node involvement: At initial diagnosis, 80%–100% of patients present with ipsilateral hilar and mediastinal lymphadenopathy coalescing with the primary mass (8, 23, 28).
• Extensive ipsilateral pleural involvement: Most cases (50%–100%) present with ipsilateral pleural invasion; PET/CT more clearly delineates minute pleural metastases than conventional CT (8, 23, 28).

3.5 Differential diagnostic features relative to common pulmonary malignancies

Five studies (8, 15, 23, 28, 34) compared imaging features of pulmonary NUT carcinoma with common pulmonary malignancies (Table 4).

Table 4. Comparison of PET/CT and clinical features between pulmonary NUT carcinoma and common pulmonary malignancies.

FDG, fluorodeoxyglucose; LN, lymph node; PET, positron emission tomography; SUVmax, maximum standardized uptake value; y, year.

Differentiation relies on a combined assessment of clinical, radiographic, and metabolic characteristics. Features suggestive of NUT carcinoma include: young, non-smoking patient; bulky central mass with high FDG avidity (SUVmax >10); extensive coalescent regional nodal metastases; widespread ipsilateral pleural involvement; relative sparing of the contralateral lung; PET-positive osteolytic bone metastases; and negative brain imaging at initial diagnosis. This combination may prompt NUT IHC or genetic testing.

3.6 Clinical utility of ¹⁸F-FDG PET/CT

¹⁸F-FDG PET/CT demonstrated substantial clinical value across multiple dimensions. One cohort study reported that 55% (6/11) of patients with thoracic NUT carcinoma were upstaged from M0 to M1 because PET/CT detected distant metastases missed by conventional imaging (29). The most frequent distant metastasis sites were bone (osteolytic, −60%–75%), distant lymph nodes (−50%), liver (−20%–33%), adrenal glands (−17%), and pleura/soft tissue (8, 27, 34). PET/CT was superior to bone scintigraphy in detecting osteolytic bone metastases (8).

Four studies described serial PET/CT for treatment response monitoring (17, 22, 24, 26). The overall pattern was “initial response followed by rapid progression”: transient metabolic improvement after chemotherapy or radiotherapy was typically followed by new distant lesions within months. Although PET/CT sensitively captures treatment response, metabolic remission is generally not durable because of the highly aggressive nature of NUT carcinoma.

3.7 Quality assessment, heterogeneity, and evidence profile

The single cohort study scored 5/9 on the Newcastle-Ottawa Scale (moderate risk of bias). The JBI checklists were applied to the five case series and 17 case reports. No studies were excluded on the basis of quality assessment (Table 5; detailed scores in Supplementary Tables S1–S3).

Table 5. Summary of quality assessment of included studies.

The maximum NOS score is 9★. The single cohort study scored 3★ for selection (representativeness of the exposed cohort, ascertainment of exposure, demonstration that outcome of interest was not present at start of study) and 3★ for outcome (assessment of outcome, sufficient follow-up duration, adequacy of follow-up). Comparability was not applicable due to the absence of a control group. The detailed itemized scores for the JBI checklists are provided in Supplementary Tables S1–S3.

JBI, Joanna Briggs Institute; NOS, Newcastle-Ottawa Scale.

The included studies exhibited notable heterogeneity in study design, patient selection, PET/CT parameters, and outcome reporting. Only one study was a retrospective cohort; the remainder were case series or case reports. The predominance of case reports carries inherent publication bias favoring positive or unusual cases.

In accordance with the GRADE framework, evidence quality was rated as follows: primary tumour SUVmax range (moderate; 12 studies, −60 cases); staging modification value (low; one small cohort); metastasis detection efficacy (low; case series/reports); treatment response monitoring (very low; four case reports); and differential diagnostic features (low; consistent observations but no diagnostic accuracy studies). The highest achievable evidence level was low to moderate.

Sensitivity analyses showed that excluding the uncertain-primary-site study (33) had no substantive impact on conclusions. All studies achieved 100% on JBI scoring, precluding exclusion on the basis of quality. A subgroup analysis restricted to NGS-confirmed cases could not be conducted because only −30% of studies employed NGS.

4 Discussion

4.1 Principal findings and academic positioning

This systematic review synthesised the PET/CT imaging characteristics and clinical utility of primary pulmonary NUT carcinoma from 23 studies (86 patients). The core findings are: (1) consistent hypermetabolic phenotype with primary tumour SUVmax 12–14; (2) proposed imaging triad (bulky central mass with necrosis, ipsilateral coalescent lymphadenopathy, extensive ipsilateral pleural involvement) with relative sparing of the contralateral lung and rare brain metastases; and (3) substantial clinical value in staging modification (>50% upstaging), bone metastasis detection, and treatment response monitoring.

The 2025 international guidelines recommend NUT IHC for screening poorly differentiated malignant tumours but acknowledge that no standardised therapeutic strategy exists (35). This review fills the evidence gap by providing the first systematic synthesis of PET/CT imaging features in pulmonary NUT carcinoma.

4.2 Interpretation of metabolic characteristics and the imaging “triad”

The SUVmax of primary pulmonary NUT carcinoma (5–40; mean, 12–14) exceeds typical levels of many lung adenocarcinomas, reflecting active glycolytic metabolism consistent with its aggressive nature (15). Mechanistically, the BRD4–NUTM1 fusion oncoprotein drives a potent oncogenic transcriptional program (36). Recent genetically engineered mouse model studies have confirmed that BRD4::NUTM1 alone can initiate NUT carcinoma-like tumors in vivo (36, 37); however, the precise downstream effectors responsible for the hypermetabolic phenotype require further investigation.

Metabolic synchronicity between primary tumours and metastases suggests that NUT carcinoma maintains consistent molecular driver characteristics during evolution, corroborating its single fusion oncogene-driven pathogenesis. Although preliminary studies on ⁶⁸Ga-FAPI PET/CT suggest potential complementary value for pleural metastasis assessment (37, 38), no FAPI studies were included in this review; our conclusions pertain exclusively to ¹⁸F-FDG PET/CT.

The proposed imaging triad builds on previous descriptions by Sholl et al. (8) and Chang et al. (23), adding two supplementary features: relative sparing of the contralateral lung and rarity of brain metastases. NUT carcinoma exhibits pronounced lymphangitic invasiveness and pleural tropism (33); for radiologists, this triad in a young, non-smoking patient should prompt NUT carcinoma in the differential diagnosis.

4.3 Clinical value in staging modification, metastasis detection, and treatment response monitoring

PET/CT holds substantial clinical value in initial staging. Kloker et al. demonstrated that 55% of patients were upstaged from M0 to M1 because PET/CT detected distant metastases missed by conventional imaging (29). Occult metastases represent a core adverse prognostic marker and directly alter therapeutic decision-making, supporting PET/CT as a staging modality at initial diagnosis.

PET/CT is superior to conventional bone scintigraphy in detecting osteolytic bone metastases (8). ¹⁸F-FDG PET/CT directly probes tumour glucose metabolism and is unaffected by osteoblastic reaction; its ability to detect minute pleural metastases is important for accurately defining disease extent.

The four included studies reveal a shared pattern of “initial response followed by rapid progression” (17, 22, 24, 26). Transient metabolic improvement after chemotherapy or radiotherapy is typically followed by new distant lesions within months, consistent with external reports showing lack of sustained responsiveness (39, 40). Although PET/CT sensitively captures treatment responses, metabolic remission is rarely durable. Serial PET/CT facilitates early recognition of disease progression, enabling timely therapeutic adjustment.

4.4 Differential diagnostic value and misdiagnosis prevention

Pulmonary NUT carcinoma differs from common lung cancers in several key respects: younger age at onset (35–45 vs. −70 years), predominantly non-smoking history (>70%), and higher rate of regional nodal involvement at initial diagnosis (80%–100% vs. 30%–40%) (41). Although no individual feature is specific, their combination constitutes a profile suggestive of NUT carcinoma.

The persistently high misdiagnosis rate (78% on initial reporting; median diagnostic delay, 10 weeks) is a core factor contributing to poor prognosis (13). Causes include radiologists' unfamiliarity with PET/CT features, pathologists' failure to routinely incorporate NUT IHC, and the inability of standard DNA-based NGS to detect >75% of cases (12). The imaging feature profile summarised in this review offers a radiology-level solution: when radiologists proactively describe this combination and raise the possibility of NUT carcinoma, they may effectively trigger NUT IHC testing and shorten diagnostic delays.

4.5 Limitations and future directions

Several limitations warrant caution: (1) predominance of case reports/series (73.9%) without prospective cohorts; (2) substantial inter-study heterogeneity in age, fusion partners, and PET/CT parameters; (3) publication bias favouring typical or unusual cases; (4) lack of standardised acquisition parameters limiting SUVmax precision; and (5) treatment response data from only four studies with variable follow-up.

Future directions include: (1) multicentre prospective imaging registries with standardised PET/CT protocols, integrated with BET inhibitor trials (35, 42); (2) exploration of fusion subtype–metabolism associations; (3) evaluation of prognostic value of metabolic parameters (SUVmax, MTV, TLG); and (4) assessment of novel PET tracers such as FAPI (43).

4.6 Clinical implications and conclusions

This review makes three contributions: (1) the first systematic review specifically addressing PET/CT imaging characteristics of pulmonary NUT carcinoma; (2) synthesis of the proposed imaging triad into clinically actionable recognition criteria (8, 23); and (3) evaluation of PET/CT clinical value with GRADE evidence ratings.

For radiologists, this study provides the imaging triad and differential diagnostic profile, facilitating early identification. For clinicians, it clarifies PET/CT’s central staging value—more than half of patients may have M-stage revised because PET/CT detects occult metastases (44). For pathologists, imaging reports raising suspicion of NUT carcinoma should trigger NUT IHC in the routine diagnostic workflow (35, 45, 46).

5 Conclusion

Primary pulmonary NUT carcinoma exhibits markedly elevated metabolism on ¹⁸F-FDG PET/CT, with a proposed imaging triad. PET/CT plays a crucial role in staging, metastasis detection, and response monitoring. Recognition of this profile may facilitate earlier diagnosis and appropriate clinical management.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fradi.2026.1933175/full#supplementary-material

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Footnotes

Edited by: Ping Wang, Michigan State University. United States

Reviewed by: Bin Gu, Michigan State University, United States

Elias Gallardo Navarro, General surgery, Mexico