Visual Abstract
Abstract
Head and neck squamous cell carcinoma (HNSCC) remains a malignancy with high rates of locoregional recurrence and poor prognosis for recurrent cases. Early detection of subclinical lesions is challenging but critical for effective patient management. Imaging surveillance after treatment, particularly 18F-FDG PET/CT, has shown promise in the diagnosis of HNSCC recurrence. The aim was to evaluate the diagnostic performance of 18F-FDG PET/CT according to delay after treatment in detecting subclinical recurrence (SCR) in HNSCC patients. Methods: In this retrospective study, all 18F-FDG PET/CT scans were performed at a single center. All adults with histologically proven HNSCC who were treated with curative intent between January 1, 2006, and December 31, 2021, were included. They had a normal clinical examination before each scan. Patients who underwent an intensive follow-up strategy after treatment had 18F-FDG PET/CT with an intravenous contrast agent at 3–6 mo and annually thereafter for 5 y. The primary endpoint was diagnostic performance (positive and negative predictive values, sensitivity, specificity, and accuracy). Results: In total, 2,566 18F-FDG PET/CT scans were performed among 852 patients, with an average of 3 scans per patient. The overall diagnostic performance measures were as follows: positive predictive value (88%), negative predictive value (98%), sensitivity (98%), specificity (89%), and accuracy (93%). There were no significant differences in diagnostic performance over time. The scans detected 126 cases of SCR (14.8%) and 118 cases of metachronous cancer (13.8%). The incidence of SCR decreased over time, with the highest detection rate in the first 2 y after treatment. Positive predictive value improved over time, reaching 90% for the digital Vision 600 system (third period) compared with 76% for the analog Gemini GXLi system (first period, P < 0.001). Multivariate analysis identified advanced stage, high body mass index, and initial PET/CT upstaging as predictive factors for detection of SCR. Conclusion: Our study demonstrates that 18F-FDG PET/CT has high diagnostic performance in detecting SCR during follow-up after treatment of HNSCC, especially in the first 2 y. Advanced tumor stage, initial PET/CT upstaging, and high body mass index were associated with a higher likelihood of SCR detection. The routine use of 18F-FDG PET/CT during follow-up seems justified for patients with HNSCC.
Head and neck squamous cell carcinoma (HNSCC) is the sixth most common malignancy worldwide, with approximately 800,000 new cases annually (1). Despite recent advances in curative-intent treatments, locoregional recurrence occurs in approximately 40%–50% of patients with advanced-stage disease, mostly within the first 2 y after treatment (2). Furthermore, the prognosis for patients with recurrent HNSCC remains poor, with a median survival of less than 1 y (3). However, a metanalysis of 1,080 patients with recurrent HNSCC demonstrated the benefit of salvage surgery by improving their median 5-y survival to 39% (4). These findings highlight that early detection of small subclinical lesions remains a major challenge to guide clinicians in patient management.
According to the guidelines, standard follow-up is based mainly on iterative clinical examinations and less on imaging (5,6). In this context, imaging surveillance after treatment may have a role to play in the diagnosis of HNSCC subclinical recurrence (SCR).
18F-FDG PET/CT is recommended for HNSCC after treatment to resolve doubt if recurrence is clinically suspected and to look for distant metastasis if recurrence is confirmed (6,7). Although several studies have already demonstrated its high performance in diagnosing disease SCR during routine patient follow-up (8–10), 18F-FDG PET/CT is still optional and concerns only patients with locally advanced disease at diagnosis. Its usefulness remains controversial (11), mainly because the cost-effectiveness ratio is debated (12,13), the prognostic impact has rarely been studied, and there is no consensus on the frequency and duration of the imaging schedule.
A metanalysis of 7 studies and 907 patients found a global detection rate of 14.2% for HNSCC SCR using 18F-FDG PET/CT, with pooled sensitivity and specificity of 89% and 92% (14), respectively; however, results were not subanalyzed by delay after treatment. Only data from 2 prospective series provide estimates of detection rates at 6 mo (18.9%) and 1 y (32.9%) after treatment (8,9), and a multicenter randomized prospective trial is ongoing to assess the value of performing imaging every 6 mo for 3 y (15).
The aim of this study was to evaluate the diagnostic performance of 18F-FDG PET/CT according to delay after treatment for the detection of SCR in a large retrospective single-center cohort of HNSCC patients.
MATERIALS AND METHODS
Population
Our retrospective study included 852 adult patients with newly diagnosed HNSCC who received curative treatment between 2006 and 2021 (Fig. 1). All patients underwent 18F-FDG PET/CT scans 3–6 mo after completion of treatment and annually for 5 y thereafter, replacing annual chest CT. They were required to have a normal clinical examination before each scan. Patient demographics, treatment information, and follow-up data were collected from electronic medical records.
Flowchart.
The primary objective of our study was to evaluate the diagnostic performance of 18F-FDG PET/CT in detecting SCR of HNSCC according to delay after treatment. The secondary objectives were to analyze its efficacy according to disease stage, treatment modality, and PET system used and to search for predictive factors for the detection of SCR using 18F-FDG PET/CT.
The study was approved by the Institutional Ethics Committee (29BRC22.0040), and all patients gave written informed consent. The study complied with the French General Data Protection Regulation.
Follow-up
All patients follow-ups consisted of a conventional work-up according to guidelines (6), including clinical examinations every 2 mo during the first year, every 3 mo during the second year, every 4 mo during the third year, every 6 mo during the fourth and fifth years, and then annually. As recommended, 18F-FDG PET/CT was performed 3 mo after radiotherapy for patients with node-positive disease to assess the necessity of neck dissection (16). All imaging studies were reviewed and interpreted by a panel of experienced nuclear medicine physicians. SCR was defined as lesions detected on 18F-FDG PET/CT that were not clinically or radiographically apparent at the time of imaging. A qualitative criterion was used to identify recurrence, that is, any nonphysiologic uptake above the surrounding background noise (Fig. 2). Because of the rapidly progressive nature of this type of cancer, if a lesion was detected by 18F-FDG PET/CT within 3 mo of the examination, the result was considered false-negative. However, if the lesion was detected more than 3 mo later, it was considered undetectable and the result was classified as true-negative. Suspected recurrence or synchronous cancer was confirmed by histologic sampling or, if not possible, by morphologic or metabolic progression on subsequent imaging within 3 mo.
18F-FDG PET/CT performed 18 mo after surgical treatment of T2N0M0 left tonsillar squamous cell carcinoma showing, in coronal (top left), sagittal (top right), and axial (lower left) views, recurrence of squamous cell carcinoma (circle). This was confirmed by CT-guided biopsy (lower right). B = background; SUV-bw = body-weight SUV; T = tumor.
18F-FDG PET/CT Imaging
The examinations were performed on several consecutive PET/CT hybrid scans as follows: from 2006 to 2012 on an analog Gemini GXLi (Philips Healthcare) system, from 2012 to 2018 on an analog Biograph-mCT (Siemens) system, and from 2019 to 2021 on a digital Vision 600 (Siemens) system. PET/CT imaging was performed 1 h after an intravenous injection of 3–5 MBq/kg of 18F-FDG (IBA Molecular Imaging). The detailed parameters of the PET/CT imaging are available in the supplemental materials (supplemental materials are available at http://jnm.snmjournals.org).
Statistical Analysis
Statistical analysis was performed using SPSS software (version 25.0; IBM). Descriptive statistics were used to summarize the demographic and clinical characteristics of the patient cohort.
The overall diagnostic performance of 18F-FDG PET/CT was evaluated using sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy. Subanalyses were performed in subgroups as follows: interval after treatment (3–6, 7–12, 13–24, 25–36, and ≥37 mo), treatment modality (surgery only and with radiotherapy), disease stage (early stage I/II and advanced stage III/IV), and PET system (first period [P1] Gemini, 2006–2011; second period [P2] Biograph, 2012–2017; and third period [P3] Vision, 2018–2022),
For univariate analysis, a Mann–Whitney test was used for continuous variables and a Fisher exact test for categoric variables. Predictive factors found to be associated with subclinical detection were then entered into a multivariable logistic regression model. A P value of less than 0.05 was considered statistically significant.
RESULTS
Patient Characteristics
Among the 852 patients included, 2,566 18F-FDG PET/CT scans were performed, with an average of 3 scans per patient. Clinical characteristics are shown in Table 1.
Patient Characteristics (n = 852)
Patients were predominantly men (695/852, 82%), with a median age of 62 y (range, 32–94 y). Of the included patients, 161 (19%) had a previous history of cancer and 58 (7%) had HNSCC from another primary site. The most common tumor site was the oropharynx (292/852, 34%). With regard to the distribution of the main curative treatments, slightly more patients received initial radiotherapy than surgery (458/852 vs. 394/852, respectively).
Diagnostic Performance of PET for SCR
Of the 852 patients, 26% (221/852) had a recurrence of head and neck cancer, of which 57% (126/221) were detected by 18F-FDG PET/CT. The overall detection rate of SCR by 18F-FDG PET/CT was 14.7% (126/852). It was 4.8% (40/830) at 3–6 mo, 10.3% (45/435) at 7–12 mo, 3.5% at 13–24 mo (21/596), 2.6% at 25–36 mo (9/343), and 3.1% at more than 37 mo (11/358).
The overall diagnostic performance of 18F-FDG PET/CT for SCR (n = 126) was 88%, 98%, 98%, 89%, and 93% for PPV, NPV, sensitivity, specificity, and accuracy, respectively. For the subgroup of 95 patients without pretherapeutic 18F-FDG PET/CT imaging, the diagnostic accuracy of the first follow-up 18F-FDG PET/CT scan (92%, 98%, 97%, 95%, and 93% for PPV, NPV, sensitivity, specificity, and accuracy, respectively) was quite similar to that of patients with pretherapeutic 18F-FDG PET/CT imaging.
Per Interval After Treatment
The diagnostic performance of 18F-FDG PET/CT at different time intervals after curative treatment is summarized in Table 2.
Overall Diagnostic Performance of 18F-FDG PET/CT and by Time Interval After Treatment
Excluding cancers detected at the initial work-up, 18F-FDG PET/CT identified 118 metachronous cancers in 13% (113/852) of patients. Of these, 5 patients developed 2 metachronous cancers during follow-up. The most common sites were the lung (54%, 61/113), colon or rectum (11%, 12/113), head and neck (10%, 11/113), and esophagus (8%, 9/113).
The incidence of SCR decreased over time, with 67% (85/126) detected in the first year and 84% (106/126) detected in the first 2 y. Curative treatment of recurrences detected during follow-up increased over time, from 42% before 13 mo to 83% after 13 mo. Overall, 56% (70/126) of patients who had SCR during follow-up received curative treatment.
The incidence of asymptomatic metachronous cancers was slightly higher in the first year but tended to remain stable over time, around 20 per year. Similarly, the rate of curative treatment increased only slightly over time, from 81% to 96%.
Per PET System
There was an improvement in PPV over time, from 76% to 84% and then 90% from P1 (Gemini) to P2 (Biograph) and then P3 (Vision, P < 0.001); all results are shown in Supplemental Table 1. In particular, there was improvement in the characterization of tumor recurrence from P1 to P2–P3, from 76% to 90% (P < 0.001). For nodal status, there was an increase from 90% to 93% and then 97% from P1 to P2 and then P3 (P = 0.004), and for synchronous cancer detection, there was an increase from 60% to 70% and then 90% from P1 to P2 and then P3 (P < 0.001). The results for NPV and sensitivity were similar over time, with excellent values ranging from 98% to 100%. Specificity decreased from 88% to 81% and then increased to 93% from P1 to P2 and then P3 (P = 0.008). Accuracy increased from P3 to reach the best overall value of 95%.
Per Disease Stage
There was no significant difference in diagnostic performance according to initial disease stage (Supplemental Table 2). PPV ranged from 84% to 87%, NPV and sensitivity ranged from 99% to 100%, specificity ranged from 88% to 89%, and accuracy remained constant at 93%.
Per Initial Treatment Modality
There was no significant difference in diagnostic performance, according to initial treatment modality (Supplemental Table 3). PPV ranged from 85% to 86%, NPV and sensitivity ranged from 99% to 100%, specificity ranged from 88% to 89%, and accuracy remained constant at 93%.
Predictive Factors for SCR
The analysis of predictive factors for SCR is detailed in Table 3.
Analysis of Predictive Factors for 18F-FDG PET/CT Recurrence Detection in Patients with Proven SCR
In univariate analysis, several factors predicted the detection of recurrence by systematic 18F-FDG PET/CT: body mass index of less than 22 (odds ratio [OR], 0.738; 95% CI, 0.579–0.941; P = 0.010), advanced stage (III or IV; OR, 1.904; 95% CI, 1.282–2.826; P = 0.002), upstaging by 18F-FDG PET/CT (PETup) at the initial assessment (OR, 1.481; 95% CI, 1.141–1.922; P = 0.005), and oral primary site (OR, 2.002; 95% CI, 1.233–3.253; P = 0.006) or oropharynx primary site (OR, 1.292; 95% CI, 1.06–1.575; P = 0.016). In multivariate analysis, 3 factors (advanced stage, body mass index < 22, and PETup) remained significantly associated with the detection of SCR (OR, 2.810; 95% CI, 1.061–7.440; P = 0.038; OR, 0.277; 95% CI, 0.116–0.658; P = 0.004; and OR, 25.582; 95% CI, 7.939–82.34; P < 0.001, respectively).
DISCUSSION
Our study showed that 18F-FDG PET/CT was highly effective in detecting 126 SCRs and 118 metachronous cancers during systematic surveillance of 852 HNSCC patients, corresponding to an occult event rate of 9.5% (244/2,566 scans). To our knowledge, this is the largest study evaluating the diagnostic performance of 18F-FDG PET/CT in this indication. We found good results, with PPV, NPV, sensitivity, specificity, and accuracy estimated at 88%, 98%, 98%, 89%, and 93%, respectively. These findings are consistent with several previous series investigating the usefulness of 18F-FDG PET/CT for the detection of SCR, with sensitivity of 88%–100%, specificity of 43%–95%, PPV of 51%–89%, NPV of 91%–100%, and accuracy of 81%–97%, regardless of anatomic site (tumor, node, or metastasis) or the delay after treatment (8,9,17–21).
In addition, the rate of curative treatment for SCR is higher than reported in the literature (23%–46%) (22–26). It has been known for several years that early management of recurrence can improve quality of care and overall survival (23,27). Recently, a cohort study of HNSCC patients demonstrated a probable protective role of imaging-based follow-up with 18F-FDG PET/CT (hazard ratio, 0.29) (28). A single-center retrospective case-control study of 782 patients with 18F-FDG PET/CT imaging showed similar results (OR, 0.71) (29). Considering these elements, and given the excellent performance of 18F-FDG PET/CT, its use in routine settings seems justified.
We chose to include only patients with a normal clinical examination at least 3 mo after completion of treatment for 2 reasons. First, it is difficult to distinguish true SCR from residual disease after treatment. Second, it is well known that there is an increased risk of false-positives because of inflammatory changes occurring less than 3 mo after treatment, especially with radiotherapy (30). In a retrospective study, Risør et al. (31) showed that the diagnostic accuracy of 18F-FDG PET/CT in the follow-up of 279 HNSCC patients improved mainly between the 0- to 3-mo and the 3- to 6-mo intervals after treatment, with a significant decrease in equivocal results (26.3% and 8.4%, respectively; P = 0.03). In our series, the diagnostic accuracy of 18F-FDG PET/CT in follow-up after treatment was stable (91%–95%), regardless of the time interval after 3 mo (32).
In our study, the incidence of disease recurrence decreased over time (67% in the first year and 84% in the first 2 y), corresponding to a detection rate of SCR by 18F-FDG PET/CT of 4.8% at 3–6 mo, 10.3% at 7–12 mo, and 3.5% at 13–24 mo (21/596). These findings are also consistent with the literature. Ho et al. (33) highlighted that PET/CT detection rates in clinically occult patients were 9% (15/175) at 12 mo and 4% (3/77) at 24 mo. Furthermore, in a retrospective study of 388 patients who underwent 18F-FDG PET/CT every 3 mo, Beswick et al. (34) concluded that 18F-FDG PET/CT is not useful after 2 y, because 95% of recurrences occur within this period. However, we found a slightly higher incidence of asymptomatic metachronous cancers in the first year, which then stabilized over time at a rate of 4.5% per year. This is in agreement with the retrospective study of 302 patients by Léon et al. (32), which showed a relatively constant incidence of metachronous cancers of 4% per year over a 10-y follow-up period.
With regard to our analysis per PET system, we found improvement in diagnostic performance, especially in terms of PPV from P1 (Gemini) to P2 (Biograph) and P3 (Vision, P < 0.001) and in terms of accuracy, with a best overall value calculated at 95% (P = 0.014). This makes sense because the advances in PET technology (i.e., point spread function + time-of-flight image reconstruction from P1 to P2 and digital detector implementation from P2 to P3) over the last decade have allowed improvement in lesion detectability (35–38), as well as increased physician expertise. Using P1 versus P2, in a study comparing the ordered-subset expectation maximization algorithm with or without the point spread function and time of flight, Akamatsu et al. (39) showed that the combination of point spread function and time of flight increased SUVmax by 43.3% in 41 lymph node metastases and improved detection of small lesions. Using P2 versus P3, Delcroix et al. (40) showed in a series of 98 patients significant improvement in image quality (noise and sharpness) and lesion detectability with the digital PET/CT dataset compared with the analoglike one.
Our study found no significant difference in performance based on initial treatment type or disease stage. The changes associated with the type of initial treatment suggest that rearrangements in tissues treated by surgery or radiotherapy do not affect this metabolic imaging. Although recurrences are less frequent and rearrangements are less significant in early stages, this does not alter the performance of PET/CT during follow-up.
In our series, multivariate analysis identified PETup at initial work-up (OR, 25.58; P < 0.001) and advanced stage disease III or IV (OR, 2.810; P = 0.038) as independent prognostic factors for subclinical events after treatment detected by 18F-FDG PET/CT, regardless of the initial stage. To date, 18F-FDG PET/CT is recommended only for HNSCC in advanced III or IV staging before treatment for distant extension (7). Nevertheless, several studies have shown its usefulness for restaging as an adjunct to conventional work-up at any stage (41–43). In a retrospective study of 477 HNSCC patients, Leclère et al. (42) found that 18F-FDG allowed PETup of disease after conventional work-up from stage I or II to advanced stage III or IV in 42 of 130 patients (32.3%). They showed that these changes were mainly related to lymph node status (38.2%) and significantly affected 3-y overall survival compared with patients whose conventional work-up and 18FDG PET/CT results were concordant (54.8 vs. 82.6%, P = 0.001). These findings highlight the importance of functional imaging in the initial assessment of head and neck cancer, because it allows greater accuracy in determining lymph node involvement. In a metanalysis of 32 studies including 1,236 patients, Kyzas et al. (44) showed sensitivity of 79% and specificity of 86% for 18F-FDG PET/CT in the diagnosis of lymph node extension in HNSCC. They also reported better sensitivity (82% vs. 74%) and specificity (82% vs. 76%) of PET than of CT in 16 studies and better specificity (85% vs. 80%) for equivalent sensitivity of PET than of MRI in 9 studies. Advanced stages are associated with a higher risk of recurrence, particularly in the lymph nodes (45). The improved diagnostic performance of 18F-FDG PET/CT over CT alone in the initial assessment of lymph node extension—the region most prone to recurrence—supports the use of PET/CT during follow-up to maximize the likelihood of detecting SCR. In a recent review, You et al. (46) highlighted that 18F-FDG PET/CT is a valuable tool for detecting nodal recurrences that may be missed by routine clinical examination or conventional imaging, with implications for patient management. In addition, a body mass index of less than 22 was found to be a protective factor for SCR of cancer in our results (OR, 0.277; P = 0.004). This tendency likely results from the increased feasibility of detecting recurrent squamous cell carcinoma in the neck by clinical palpation because of the reduced volume of adipose tissue. In addition, in patients with a body mass index of at least 22, the lower contrast because of fat may also limit the sensitivity or specificity of 18F-FDG PET/CT.
Our study had several limitations. First, its retrospective single-center design requires validation by multicenter randomized prospective studies to limit selection bias. Such a trial is under way to evaluate the value of PET-based follow-up up to 3 y after treatment (15). Second, our population included a high proportion of advanced stage disease (74%), which may have artificially increased the incidence of SCR or metachronous disease compared with an unselected population. Nevertheless, this was a real-life study, so this rate is consistent with the available data in our region over the last 20 y. Third, some patients (only 11%) did not undergo pretherapeutic 18F-FDG PET/CT, which may have led to an underdiagnosis of asymptomatic synchronous cancers and thus slightly increased our rate of subsequent metachronous cancers. Fourth, data on human papilloma virus status were missing for two thirds of the patients, because this was not recommended in the early years of the inclusion period. Fifth, 18F-FDG PET/CT imaging was performed with iodinated contrast agents unless contraindicated, a practice that may not be widespread in all centers. The use of contrast agents has been shown to improve the differentiation between anatomic structures and pathologic abnormalities, thus improving the diagnostic accuracy of PET/CT scans (47–49). This may explain the sensitivities and specificities observed in this study. Sixth, 56 patients were positive for human papilloma virus. The performance of 18F-FDG PET/CT is better in patients with cancers related to human papilloma virus (50), probably in part because the risk of recurrence is lower. Finally, during the long inclusion period (13 y), there has been a remarkable evolution in treatment methods and medical imaging techniques. This time interval may also have affected the PET analysis, depending on the learning curves of the nuclear medicine physicians.
CONCLUSION
In a large cohort of 852 asymptomatic patients, we have shown that 18F-FDG PET/CT has high diagnostic performance in detecting subclinical HNSCC recurrence, as well as metachronous cancers throughout follow-up after treatment. Several factors may predict the likelihood of a positive result, including advanced tumor stage (III or IV), PETup at initial work-up, and tumor location outside the oral cavity. Considering these performance results and recent studies showing the protective role of imaging-based follow-up, the routine use of 18F-FDG PET/CT during follow-up seems justified for patients with HNSCC.
DISCLOSURE
No potential conflict of interest relevant to this article was reported.
KEY POINTS
QUESTION: What is the diagnostic performance of 18F-FDG PET/CT during follow-up in clinically asymptomatic patients treated for HNSCC?
PERTINENT FINDINGS: In this retrospective study that included 852 adults, 2,566 18F-FDG PET/CT scans were conducted. The overall diagnostic performance measures were as follows: PPV of 88%, NPV of 98%, sensitivity of 98%, specificity of 89%, and accuracy of 93%.
IMPLICATIONS FOR PATIENT CARE: 18FDG PET/CT had excellent performance when used for the follow-up of asymptomatic patients treated for HNSCC.
Footnotes
Published online Jul. 11, 2024.
- © 2024 by the Society of Nuclear Medicine and Molecular Imaging.
REFERENCES
- Received for publication January 31, 2024.
- Accepted for publication May 13, 2024.
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.