Head and neck cancer is the sixth most common malignancy worldwide, with around 800,000 new cases and 320,000 deaths annually (1). These malignancies encompass cancers of the oral cavity, oropharynx, hypopharynx, and larynx and are squamous cell carcinomas (HNSCC) in 90% of the cases. The 5-year survival rate does not exceed 80% for patients with localized disease whereas it decreases to 50% in case of regional lymph node involvement, and to 20% when distant metastasis are present at diagnosis (2).

Surgery or radiation therapy with or without chemotherapy are the cornerstones of treatment, but multimodal management is often required in case of locally advanced disease (3). Hence, accurate cancer staging remains essential to select the appropriate treatment strategy. The gold standard based on conventional work-up (CWU) includes physical examination, endoscopy, computed tomography (CT), and/or magnetic resonance imaging (MRI) of the head and neck to assess the local and regional disease extension. Thoracic CT is also recommended as the lung is the most common site of distant metastasis and synchronous primary cancer (SPC) (3).

The use of 18F-fluorodeoxyglucose (18-FDG) positron emission tomography (PET)/computed tomography (CT) has significantly increased recently in the field of oncology. It allows a whole-body assessment in a single scan with a relatively low radiation exposure (4). Some studies have shown that 18-FDG PET/CT is superior to CWU to assess remote extension and to detect occult SPC (5–8). Nevertheless, pre-therapeutic 18F-FDG PET/CT is currently only recommended to assess distant extension and/or to detect SPC in locally advanced HNSCC and not in early stage cancers (3, 9–11). Modification of staging after integrating 18F-FDG PET/CT as part of the initial WU and the impact on prognosis and clinical management remains also poorly understood and not clearly reported in the literature.

We hypothesized that 18-FDG PET/CT may improve HNSCC staging after initial WU regardless of disease stage, resulting in potential change in patient management.

Therefore, the objective of this study was to assess the impact of the additional information provided by 18F-FDG PET/CT on HNSCC initial staging and whether integrating PETs into WU modify clinical management and prognosis, regardless of the CWU-based staging.

Our study demonstrated the real added value of 18-FDG PET/CT for (i) HNSCC initial staging in addition to the initial WU regardless of clinical stage and (ii) assessing remote disease extension and detection of SPCs with consequential alteration of the CWU-determined management plan. To the best of our knowledge, this study is the largest to assess the management impact of 18-FDG PET/CT in newly diagnosed HNSCC patients.

Foremost, our results suggested superior performance of sPET to assess nodal and metastatic status than sCWU [statistically significant prognostic stratification of sPET (p < 0.001)]. In fact, unlike sCWU, sPET strictly respected the prognostic stratification of the AJCC stages (13). sCWU IVB patients had inferior 3-year OS than sCWU IVC patients (11.7 vs. 30%). Conversely, sPET IVB patients had better survival than sPET IVC patients (26.7 vs. 18.5%), closer to the 3-year OS observed in the literature of ~30% (13, 20).

In our series, we found a high rate (46.3%) of modification of staging and/or detection of occult-sCWU SPCs by integrating sPET into initial WU. In comparison, Cacicedo et al. (17) reported a modification of staging rate of 38%, but without including occult-sCWU SPCs detected by sPET. These results are concordant with our modification of staging rate of 40.7%. Moreover, this rate was statistically higher for sCWU advanced stages patients compared to early stages (51 vs. 32% p = 0.02). This could be explained by the lower sensitivity of sPET in cN0 patients (without suspicious node), as already demonstrated (20, 21).

The SPC rate in our series was 13%, with 7.3% of occult-sCWU SPCs detected by sPET. These rates are in accordance with the literature: Jones et al. (21) found a SPC rate of 9.1% in a study of 3,436 patients. Moreover, Haerle et al. (22) and Strobel et al. (23) reported detection rates of 6.1 and 7.9%, respectively, in smaller cohorts of patients.

Of the 130 sCWU early stage I/II patients, 29 (22%) were sPET-upstaged to stage III/IV. Furthermore, in this early stage subgroup, 14 occult-sCWU SPCs (10.8%) were detected by sPET, altering management and prognosis (24). sPET led to a therapeutic escalation (medium or high treatment modification) in 24 patients (18.5%). This rate was statistically equivalent for advanced stage (stage III/IV) patients (p = 0.02), highlighting the sPET diagnostic accuracy even for early stages.

In the same way, amongst sCWU advanced stage III/IV patients, sPET-downstaged patients with therapeutic de-escalation (medium or high impact) with significantly better 3-year OS (78.6 vs. 44.4%; p = 0.02), whereas sPET-downstaged patients without therapeutic de-escalation had a slightly lower 3-year OS but also tended to be significantly higher than non-downstaged patients (p = 0.1). It can be assumed that the impact of therapeutic de-escalation is positive on survival. However, it could not be excluded that this lack of significance in downstaged patients without de-escalation could be attributed to decreased number of patients and/or selection bias. Our other hypothesis is that treatment-related morbidity and mortality is higher for downstaged patients without therapeutic de-escalation.

In our study, therapeutic impact assessment was carried out using strict methodology, and only the certainties of treatment modification were retained. Probable changes in treatment but without certainty were not retained, especially when P-treatment was different from C-treatment and different from the treatment proposed by the multi-disciplinary team of experts (n = 6). Our medium or high treatment modification rate was 19.5% (n = 477). In the literature, this rate is variable: assessed at 13.7% (n = 233) for Lonneux et al., 15.7% (n = 248) for Ryu et al., 26.1% (n = 84) for Cacicedo et al., 33.8% (n = 71) for Scott et al., and up to 40% (n = 35) for Connell et al. (7, 14–17). This variability is probably due to different elements. First, the use of external radiotherapy varied widely from 25 to 74%, depending on the series. In our study, 59.8% of patients were curatively treated with external radiotherapy, and 77% of medium management changes were related to this treatment. Indeed, the large majority of sPET modification of staging in our series concerned regional lymph nodes involvement. This could have easily impacted radiotherapy treatment planning (25, 26). Second, patient management modification rates were lower in the more recent studies, probably related to improvement in imaging techniques. The more frequent use of MRI has probably helped to improve the sensitivity of the CWU. Third, the recent technical innovation of 18-FDG PET/CT may also improve lesion detection and the general use of contrast-enhanced combined CT may decrease the number of false positives results with resultant improved physician interpretation. Finally, the recent advent of dynamic whole-body PET/CT imaging supported by the latest clinical PET systems may also allow the use of more quantitative markers, beyond SUV and static uptake metrics, to further improve the clinical usefulness of sPET with respect to sCWU (27, 28).

Given our large patient cohort, we were able to perform subgroup analysis based on the localization of the primary tumor, unlike other series in the literature. Thus, our results demonstrated sPET outperformed sCWU for initial evaluation of oropharyngeal, hypopharyngeal, and oral cavity cancers with ~20% change in treatment plan (medium or high). This rate was significantly lower for laryngeal cancers (13.6%, p = 0.03), which was probably due to the low nodal involvement rate of this primary tumor location.

Our study had several limitations. First, the retrospective design requires validation by a randomized prospective trial. Second, Human Papilloma Virus (HPV)/p16 status was not routinely determined for HNSCC in our institution during the inclusion period. Therefore, this could lead to heterogeneity within our cohort with different prognoses of HPV+ and HPV- patients according to the 8th TNM classification, which was published after our last patient inclusion (29). Third, during the 10-year inclusion period, 18-FDG PET/CT examinations were performed on 3 different systems of different generations. However, this corresponds to a real-life study, the results of which could be applicable in different centers. Fourth, only two-thirds of PET scans were contrast-enhanced and, at worst, it decreases the diagnostic performance of 18-FDG PET/CT vs. CWU. Finally, this study was not designed to assess the cost-effectiveness of 18-FDG PET/CT in initial routine imaging workup.

Current guidelines optionally recommend the use of 18-FDG PET/CT for early CWU stages I/II to look for synchronous carcinoma that may modify the treatment plan (11). Our findings suggest that the systematic implementation of 18-FDG PET/CT in the initial WU of HNSCC, regardless of disease stage, could improve the staging and the treatment plan with significant consequences on survival outcome.

In conclusion, our study demonstrated that implementation of 18-FDG PET/CT in the initial WU of HNSCC provides valuable staging information with better prognostic stratification. Patient management was improved for any disease stage, even early stages I-II. Amongst patients with advanced disease, about half were restaged, and about a third of patients with early disease stage were restaged. In total, the management plan was altered for about 20% of patients. Furthermore, 18-FDG PET/CT modification of staging and treatment changes were significantly associated with OS. In addition, our study confirmed the interest of 18-FDG PET/CT for SPC detection, regardless of the stage.

The datasets generated for this study are available on request to the corresponding author.

The study was approved by the Ethics Committee Brest University Hospital (29BRC18.0012) (RENOVATE ClinicalTrials.gov NCT03841175) and all patients signed a written consent form. The study follows the French General Data Protection Regulation (GDPR).

RA, J-CL, and OD are the guarantors of the paper and realized statistics. RA, JR, GV, J-CL, and OD designed the study. RM, J-CL, and GV ensured inclusion and follow-up of patients. RA, OD, and JR managed imaging procedures. DG revised English language. RA, LO, and J-CL analyzed patient records. RA, MA, OD, and J-CL analyzed the data. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.