The Feasibility of ¹⁸F-FDG PET Scans 1 Month After Completing Radiotherapy of Squamous Cell Carcinoma of the Head and Neck

Patients

Between August 2001 and September 2004, 97 SCCHN patients were diagnosed and treated with definitive radiotherapy at the Asan Medical Center, Ulsan University College of Medicine, Seoul, Korea. Eligible patients included those with poorly differentiated or undifferentiated carcinomas and no distant metastases. This clinical study was approved by the institutional review board of our hospital, and all patients provided oral or written informed consent.

The staging work-up consisted of taking a medical history and performing a physical examination, panendoscopy with tumor measurements, a biopsy, a dental evaluation, a bone scan, chest radiography, a complete blood count, liver function testing, and contrast-enhanced CT or MRI of the head and neck. At the time of the second ¹⁸F-FDG PET scan, CT or MRI was not routinely performed, but all patients underwent a routine examination including fiberoptic endoscopy of the nasopharynx or larynx. CT or MRI of the nasopharynx, including the neck, was performed if the primary tumor invaded the parapharyngeal or intracranial space.

PET

All patients fasted for at least 8 h before undergoing ¹⁸F-FDG PET and, to stimulate ¹⁸F-FDG excretion from the renal calyces and subsequent voiding, drank 1 L of water just before the scan began. In each patient, including those with diabetes mellitus, the serum glucose concentration had to be under control and less than 120 mg/dL; the average blood glucose concentration before intravenous injection of ¹⁸F-FDG was 99 ± 12 mg/dL. All patients rested for at least 1 h before undergoing PET. About 555 MBq (15 mCi) of ¹⁸F-FDG, of radiochemical purity greater than 99%, were injected into a peripheral vein of the anterior upper arm. Sixty minutes later, PET studies from the skull base to the pelvis were obtained with septa (2-dimensional mode) on a full-ring scanner—88 on an ECAT Exact HR+ PET scanner (CPS/Siemens) and 9 on a Biograph Sensation PET/CT scanner (CPS/Siemens). These scanners have an axial field of view of 15.5 cm and a full width at half maximum of 4.6 mm at the center of the field of view. The PET scans were obtained at 4 and 6 min per bed position for transmission and emission scanning, respectively. PET images were reconstructed on a 128 × 128 matrix using an ordered-subsets expectation maximization algorithm for 8 subsets and 2 iterations, with a 6-mm gaussian filter. A segmented attenuation-correction algorithm was used with ordered-subsets expectation maximization. Attenuation correction was performed using 2 different methods. The ECAT Exact HR+ was used for attenuation correction of the PET images obtained on that scanner, and the CT data from the Biograph Sensation were used for attenuation correction of the PET images obtained on that scanner. The maximal standardized uptake value (SUV_max) (18) was measured for 1 pixel of each region of interest, and the SUV was determined using the whole-body attenuation-corrected image according to the following equation: SUV = maximum regional activity (Bq/mL)/(injected activity [Bq]/lean body weight [g]). The region of interest was defined by visual interpretation on the workstation computer. No correction was made for glucose or partial volume. All PET images were interpreted by a professor of nuclear medicine. An SUV_max reading of 3.0 was used to distinguish between malignancy and normal tissue. If the SUV was 2.5–3.5, a biopsy was performed whenever possible. If a biopsy could not be performed, the clinical outcome was determined by follow-up status.

Response Evaluation and Follow-up

Immediately after and 1 mo after the completion of radiotherapy, patient response was evaluated by physical examination and fiberoptic endoscopy. If residual disease was suspected or abnormal uptake of ¹⁸F-FDG was seen, a biopsy was performed for confirmation. All patients were followed up for at least 6 mo or until death. In this study, the period of response evaluation was 6 mo after the completion of radiotherapy. Recurrence was defined as evidence of disease after 6 mo. However, if the neck node was palpable, or if there was a residual mass in the nasopharynx, larynx, or hypopharynx as shown by fiberoptic endoscopy, we performed further studies such as CT, MRI, or neck sonography 2 mo after radiotherapy.

Study Design and Statistical Analysis

The baseline ¹⁸F-FDG PET scan of each patient was obtained no more than 4 wk before the initiation of radiotherapy. All patients also underwent whole-body ¹⁸F-FDG PET within 4–6 wk of the completion of radiotherapy, as described previously (17). CT or MRI was not routinely performed at the same time as the postradiotherapy ¹⁸F-FDG PET scan. If a residual mass was observed in the primary tumor site and the residual node was palpable, CT or MRI was performed 1 mo later with strict clinical follow-up. The cutoff SUV was 3.0, with SUVs of less than 3.0 defined as negative and SUVs of 3.0 or more defined as positive, the same value as that used in a previous study (19). However, because SUVs are semiquantitative, it is not possible to determine the specific value for reference. The results for primary sites and lymph nodes were separately evaluated to determine the sensitivity, specificity, and accuracy of ¹⁸F-FDG PET. If a patient had 2 or more metastatic lymph nodes, only the one with the highest SUV was measured. P values of less than 0.05 were considered statistically significant. Categorizations of continuous variables used medians as cutoff values. The distributions of continuous outcome variables were compared among subgroups, with the Mann–Whitney U test used for binary outcomes.

Demographic Data

Of the 97 patients, 85 were men and 12 were women, with a median age of 57 y (range 17–83 y). The patient characteristics are summarized in Table 1. The follow-up period ranged from 6 to 39 mo (median, 20 mo). Fifty patients had nasopharyngeal cancer, 16 had hypopharyngeal cancer, 16 had carcinoma of the larynx, 8 had carcinoma of the oropharynx, 4 had oral cavity cancer, and 3 had carcinoma of the paranasal sinus. After radiotherapy, 86 patients (89%) showed locoregional complete responses. Partial responses to radiotherapy were confirmed by biopsy or salvage surgery. Of the 86 patients who had locoregional complete responses, 14 had locoregional recurrent tumors and 8 had distant metastases. Of the 14 cases of locoregional failure, 7 occurred at the primary site, 5 at the regional nodes, and 2 at both sites.

¹⁸F-FDG PET Findings

Using ¹⁸F-FDG PET, we evaluated a total of 388 SUV sites before and after radiotherapy in 97 patients. Before radiotherapy, the median SUV at the primary sites was 6.5 (range, 2.3–23.0), and the median SUV at the neck nodes was 5.6 (range, 1.2–16.8). One month after the completion of radiotherapy, the median SUV at the primary sites was 1.8 (range, basal status value to 9.7), and the medium SUV at the neck nodes was 1.8 (range, basal status value to 8.6) (Table 2). Basal status is classed as a level of ¹⁸F-FDG uptake that is similar to the background ¹⁸F-FDG uptake level of normal tissue and is usually an SUV of about 1.0. This term was used because physiologic ¹⁸F-FDG uptake of normal tissue varies. At the primary sites and neck nodes, the SUVs for true- and false-positive lesions ranged from 3.1 to 9.7 at primary sites and from 3.2 to 4.4 at neck nodes. A positive PET result was defined as an SUV of at least 3. The mean SUV of true-positive lesions was significantly higher than that of false-positive lesions (3.7 ± 0.41 vs. 6.0 ± 2.02; P = 0.003, t test).

Postradiotherapy PET image analysis showed that 23 sites were positive for uptake and 171 were negative (Table 3). Of the 23 positive sites, 15 (65.2%) were true-positive, as shown by biopsy results or evidence of recurrence after at least 6 mo. Of the 171 negative sites, 2 (1.2%) were considered to be false-negative. Thus, the sensitivity of ¹⁸F-FDG PET in evaluating the postradiotherapy status of these SCCHNs was 88.2% and the specificity was 95.5%. The overall diagnostic accuracy of this method was 94.9%. The lack of false-negative findings for neck nodal disease is particularly interesting, making the sensitivity of this method for neck nodes 100%. The diagnostic accuracy was 90.7% at primary sites and 99.0% at the neck nodes (P = 0.002, χ² test). Figure 1 shows a typical example of ¹⁸F-FDG PET scans of a patient with tumors that had a complete metabolic response as shown by ¹⁸F-FDG PET but that were classed as morphologically nonresponsive on CT.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1. 

Example of complete metabolic response. CT images after radiation therapy showed residual disease (bottom right panel), but no pathologic evidence of disease was observed.

The nonspecific inflammation resulting from radiotherapy of the primary tumor site did not reduce ¹⁸F-FDG uptake below the basal level, although there were no residual tumor cells. However, reductions of less then 50% in initial gross tumor volume were observed in 9 patients, 8 of whom were pathologically positive and 1 of whom was pathologically negative. The SUVs of the pathologically positive cases ranged from 4.1 to 9.7 (mean, 6.3), whereas the SUV of the pathologically negative patient was 1.9. These 9 patients, who were diagnosed with clinically residual disease, were found to have residual tumor tissue on biopsy or fine-needle aspiration. With the exception of 1 case, all PET-negative nodes were less than 1.5 cm as measured by neck sonography.

The ¹⁸F-FDG PET–negative and CT-positive case is presented in Figure 1. This patient was a 29-y-old man who presented with a swollen right neck node, dysphagia, and odynophagia. Panendoscopy showed a large, bulky tumor occupying the entire nasopharyngeal cavity and extending into the right lateral pharyngeal wall. The clinical stage according to the 5th edition of the staging manual (20) of the American Joint Committee on Cancer was T2b N2 M0. The residual primary tumor and neck node were observed on CT 1 mo after the completion of radiotherapy, although ¹⁸F-FDG PET showed a complete metabolic response (SUV, 1.9). Three months after radiotherapy, salvage surgery was performed to remove the residual tumor. The pathologic finding was total necrosis. The patient is in a locoregional control state at present.