Clinical value of 18FDG PET/CT in screening for distant metastases in head and neck squamous cell carcinoma

The detection of distant metastases is of major importance in management of head and neck squamous cell carcinoma patients.


F-fluorodeoxyglucose positron emission tomography ( FDG
PET) is able to detect distant metastases not detected by chest CT.
The combination of whole-body 18 FDG PET and contrast-enhanced (CE)-CT of the chest revealed a higher sensitivity than chest CT or whole-body 18 FDG PET as separate modalities. 13 As a result, integrated PET/CT has taken a major role in screening for distant metastases in head and neck cancer, as it combines functional (metabolic) imaging (PET) with anatomical details (CT).
Several studies have reported on the value of 18 FDG PET/CT in screening for distant metastases in head and neck cancer, but with variable results. 14 Sources of heterogeneity comprise the patient spectrum (tumour type, stage and risk factors), as well as the reference standard (particularly the duration of follow-up) and locoregional control during follow-up. Cost-effectiveness analyses for the use of the 18

FDG PET/CT screening for distant metastases have
shown that its use results in overall cost savings by reducing the number of futile radical treatments. 15,16 The aims of our study were to assess in a large cohort of HNSCC patients at increased risk of developing distant metastases: (i) the accuracy of 18 FDG PET/CT to identify patients with distant metastases (ii) compare overall survival between patients diagnosed with distant metastases during initial screening and patients diagnosed with distant metastases during follow-up.

| Patients and study design
We retrospectively reviewed the medical records of consecutive patients diagnosed with HNSCC and a high risk of distant metastases, 10,17,18 who underwent routinely 18  year follow-up because of other reasons were excluded from accuracy analysis and censored in survival analysis. The minimum of 12 months follow-up was determined based on the time of detection of distant metastases during follow-up found in previous studies. 4,8,9,19 Patients with distant metastases were treated with palliative intent, whereas patients were treated with curative intent after negative screening. If equivocal results remain equivocal after discussion in the multidisciplinary team, patients were treated with curative intent and imaging was repeated during follow-up.
As part of the pre-therapeutic work-up, all patients underwent panendoscopy, and contrast-enhanced CT and/or magnetic resonance imaging (MRI) of the head and neck. If considered indicated, (ultrasound guided) fine needle aspiration of cervical lymph nodes was performed. Post-treatment follow-up was performed by regular visits to the outpatient clinic (every 6-8 weeks in the first year, with increasing intervals in following years). No routine imaging was planned to screen for distant metastases, but additional examination was performed when suspicion arose based on patient history or physical examination (eg weight loss, lesions/complaints suspicious of recurrence).

| Combined 18 FDG-PET/CT imaging
During our study period, both the Gemini TF-64 and Ingenuity TFintegrated PET/CT systems (Philips Medical Systems, Best, The Netherlands) were used to perform whole-body (from mid-thighs to skull vertex) 18 FDG PET/CT scans. 20,21 The CT scans were low-dose non-contrast made for attenuation correction and anatomical correlation of PET, as well as CE-CT of the chest during the same session.
Patients fasted for at least 6 hours prior to scanning, which started T A B L E 1 High-risk patients are defined as having one or more of the risk factors as assessed by palpation and radiological examinations. Second primary and recurrent tumours are proven by pathology. Number and percentages of risk factors present in total study population (n = 190). Some patients had more than one risk factor We scored what suspicion arose from the 18 FDG PET/CT using the written report of each scan. Most lesions suspicious of being malignant on 18 FDG PET/CT were confirmed using additional (follow-up) imaging, endoscopic work-up and/or biopsy, using a rational approach. In some cases, findings of 18 FDG PET/CT were considered equivocal proof of distant metastases and consensus was reached not to perform additional work-up by the multidisciplinary team (giving the patient the benefit of the doubt). If tissue of both the primary process and the suspicious lesion outside of the head and neck region was obtained, and consisted of identical cell types, preferably a loss of heterozygosity analysis was performed.

| Statistical analysis
The reference (gold) standard used was follow-up of 12 months.
Sensitivity, specificity, positive and negative predictive values of the 18 FDG PET/CT for detection of distant metastases were calculated. 18 FDG PET/CT findings suspicious of being metastases were considered positive. Equivocal findings were scored as negative because patients with equivocal findings cannot be withheld from curative treatment. As in clinical practice, these patients were treated with curative intent. If no suspicious lesion or lesions suspicious of being either benign or second primary tumours were found, the scan was considered negative. The 18 FDG PET/CT findings were compared to the findings of further initial work-up and findings during follow-up.
We considered negative findings on 18  Patients with second primary tumours outside the head and neck region, which were found during screening, were described separately. Significance was attributed to a P-value less than .05 (P < .05).
All calculations were performed using SPSS 22.0 for Windows.

| RESULTS
A total of 190 patients who underwent an 18

| Survival
Median OS in patients with distant metastases was significantly worse compared to patients without distant metastases (P < .001)

| Second primary tumours
In twenty-four patients (12.6%), the initial 18 FDG PET/CT was suspicious for second primary tumours outside of the head and neck region.
In ten patients, the detected lesions were suspicious of being either distant metastases or second primary tumours. Five of these lesions proved to be distant metastases through further work-up and were classified as true positive in Table 3. Of the remaining 19 patients (all classified as true negative for distant metastases), work-up revealed 11 lesions (57.9%) to be second primary tumours, whereas in the other cases further work-up of the lesion showed no signs of malignancy. Table 4 depicts the number of lesions found in each organ (system).

| Prevalence
In our cohort, 51 of 190 patients (26.8%) developed distant metastases, which is at the high end of the percentages found in previous studies. [3][4][5][6][7][8]12 This can be explained by our definition of the high-risk patient 10,18 and our follow-up duration. We included also patients with a previous malignancy of the head and neck region and/or recurrent disease, whereas other studies included only patients with previously untreated head and neck cancer. 7,9 The prevalence of distant metastases found by Haerle et al 8

| 18 FDG PET/CT
At first glance, the sensitivity of PET/CT to detect distant metastases in our study seems to be considerably lower than in other studies (Table 5). 8,9 This difference can partly be explained by the duration of follow-up: we considered all distant metastases identified during follow-up and negative on initial 18  Regarding the limited sensitivity of 18

| Limitations
In the present study, we did not analyse the tumour burden of dis- resected, and the distant sites are ablated (surgically or with radiation), a prolonged disease-free interval, and possible cure, may be achieved. 27 More research is needed to develop a new protocol for screening for distant metastases after implementation of the concept of treating oligometastases in HNSCC.
Another limitation of this study is potential bias in comparing survival of patients with distant metastases diagnosed at initial screening and follow-up due to different treatment regimens in these groups of patients. However, overall survival of patients with distant metastases detected during follow-up suggests that initial locoregional treatment with curative intent may be worthwhile in these patients with negative screening by PT/CT at initial work-up.

| CONCLUSION
Head and neck squamous cell carcinoma patients with low jugular lymph node metastases or regional recurrence have the highest risk of distant metastases. Patients diagnosed with distant metastases at initial screening have a significantly worse expected survival (from time of initial screening) compared to the group diagnosed during follow-up. In HNSCC patients with high-risk factors, 18 FDG PET/CT (including chest CE-CT) has a high negative predictive value for the detection of distant metastasis and should be part of the initial diagnostic work-up, although the sensitivity is limited when long-term follow-up is used as reference standard.

CONF LICT OF I NTEREST
The authors have nothing to disclose.