Does ¹⁸F-FDG PET/CT Improve the Detection of Posttreatment Recurrence of Head and Neck Squamous Cell Carcinoma in Patients Negative for Disease on Clinical Follow-up?

MATERIALS AND METHODS

RESULTS

Ninety-one patients, 13 women and 78 men, with a mean age of 57.4 ± 9.4 y were included in the study. Patient characteristics and pathologic status are summarized in Table 1.

The whole-body ¹⁸F-FDG PET/CT results were negative in 52 patients and positive in 39. For 30 of 39 patients with positive findings, recurrence was confirmed. Fourteen patients had only local recurrence (Fig. 1) (6 in the oral cavity, 1 in the hypopharynx, 6 in the oropharynx, and 1 in the laryngeal area); for 2, an associated cervical lymph node extension was confirmed by biopsy. Twelve had distant metastasis confirmed on the basis of progression seen on CT or histologically. Among them, 2 had both local recurrence confirmed by biopsy and distant metastases (1 in the oral cavity and 2 in the larynx, both with a lung metastasis) confirmed by CT, and 1 had local recurrence, a neck lymph node extension, and distant metastasis. For 17 patients, salvage surgery was performed; 10 patients underwent chemotherapy or palliative treatment; 2 patients refused therapy; and for 1 patient nothing was done. Nine of the 39 patients who exhibited abnormal ¹⁸F-FDG uptake in the laryngeal area did not have recurrent HNSCC despite many biopsies (false-positive; Figs. 2 and 3). The detailed results of these patients are shown in Table 2.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1. 

Transaxial ¹⁸F-FDG PET/CT images demonstrate true-positive focus (arrow) representing recurrent oral cavity carcinoma 12 mo after surgery.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2. 

Transaxial ¹⁸F-FDG PET/CT images demonstrate false-positive focus (arrow) not confirmed by ultrasonography and follow-up of oral cavity carcinoma 12 mo after surgery.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3. 

Transaxial ¹⁸F-FDG PET/CT images demonstrate false-positive focus (arrow) corresponding after biopsy to Bartonella henselae infection 12 mo after surgery and radiochemotherapy of oral cavity carcinoma.

All 52 patients with negative findings on¹⁸F-FDG PET/CT remained free of disease at 3 mo after each PET/CT examination, and for no one was evidence of malignancy on physical examination found during this period.

As shown in Table 3, the sensitivity and specificity of ¹⁸F-FDG PET/CT in this study for the diagnosis of HNSCC recurrence were 100% (30/30) and 85% (52/61), respectively. The positive predictive value was 77% (30/39). The negative predictive value was 100% (50/50). The overall accuracy was 90% (82/91).

DISCUSSION

A delay in the detection of recurrent HNSCC has been shown to be deleterious to clinical outcome after treatment. Patients with recurrent, early-stage HNSCC who undergo salvage surgery have a 70% 2-y relapse-free survival, whereas those with recurrent, advanced-stage HNSCC undergoing surgical salvage have just a 22% 2-y relapse-free survival (13). Early diagnosis and accurate identification of recurrent HNSCC are therefore critically important for successful treatment. Sequential physical examinations are generally accepted for posttherapy follow-up of patients with head and neck cancer. The National Comprehensive Cancer Network panel suggests a general consensus surveillance schedule: physical examination every 1–3 mo during the first year, every 2–4 mo during the second year, every 4–6 mo during the next 3 y, and every 6–12 mo thereafter (14,15). However, physical examination can be compromised by granulation, fibrosis, tissue edema, and necrosis. ¹⁸F-FDG PET can identify viable tumor on the basis of higher glycolytic rates in neoplasm than in necrotic or reactive tissues (16).

Our study showed that ¹⁸F-FDG PET/CT is a useful tool in detecting recurrent HNSCC during posttreatment surveillance despite negative findings on follow-up physical examinations. In addition to its ability to detect the presence of locoregional disease in these patients, ¹⁸F-FDG PET/CT provided significant additional information. Thirty of 91 patients had a proven recurrence that was not discovered by the usual sequential physical examinations. This high level of recurrence agrees with the usual recurrence rates (>30% during the first 3 y (4,17), even in asymptomatic patients (18)). The study suggests an earlier detection of the recurrence that probably will be found later during the usual follow-up. For 9 patients, ¹⁸F-FDG PET/CT had positive results despite the absence of recurrent disease. These patients probably had local inflammation, and one had mandibular osteoradionecrosis, which is known to cause false-positive results. This is in agreement with many studies that have proven the need to delay imaging after the end of radiotherapy to minimize persistent inflammatory changes (19,20). Moreover, false-positive results may occur with ¹⁸F-FDG PET because of infections; physiologically increased uptake in structures such as the palatine tonsils and salivary glands (salivary gland cancers are typically not ¹⁸F-FDG–avid) and the masticator, oral cavity, neck, and laryngeal muscles; uptake in reactive nonneoplastic lymph nodes; uptake after surgery; and uptake in noninfectious inflammation and granulation at the surgical site (flare phenomenon). Uptake in muscles may be asymmetric and is usually due to anxiety, with laryngeal and masticator activity after ¹⁸F-FDG injection. Coregistered images with ¹⁸F-FDG PET/CT allow a direct correlation between ¹⁸F-FDG uptake and anatomic structures, thus reducing false-positive results. Attenuation-correction artifacts result from erroneous overcorrection of PET emission data by software that uses CT transmission data for attenuation correction. These artifacts occur in areas that have a high attenuation on corresponding CT images (e.g., metallic implants) but can usually be detected by evaluating the uncorrected emission ¹⁸F-FDG PET data (21).

When the results of ¹⁸F-FDG PET are negative, the likelihood of recurrent malignancy is low (at least at 6 mo of follow-up), but it should not underestimate a possible false-negative, as when malignancy is present in structures with a physiologically elevated metabolism (e.g., tonsillar carcinoma), when the tumor is smaller than the resolution of current PET/CT scanners (typically 10 mm for PET scanners), or when the tumor is not ¹⁸F-FDG–avid. Therefore, careful surveillance and continued follow-up are still probably needed in those patients with a high risk of recurrence (21). A negative predictive value of 100% gives clinical confidence that a false-negative result is unlikely. Moreover, with a sensitivity of 100%, if malignancy is present, ¹⁸F-FDG PET will produce positive findings. Nevertheless, with a specificity of 85%, unnecessary explorations are performed in less than 15%.

Few other studies have evaluated other imaging modalities such as CT and MRI in such treated and asymptomatic patients. However, in a recent review of trials evaluating PET for detecting residual or recurrent HNSCC in undifferentiated patients treated by radiotherapy or chemoradiotherapy, among 27 of 1,871 identified studies the pooled sensitivity and specificity of PET for detecting residual or recurrent HNSCC were 94% and 82%, respectively, whereas the CT mean pooled sensitivity and specificity were lower (67% and 78%, respectively), as were the sensitivity and specificity of MRI (81% and 46%, respectively) (22). In another recent review of 21 articles comparing ¹⁸F-FDG PET, CT, and MRI for the diagnosis of local residual or recurrent nasopharyngeal carcinoma, the pooled sensitivity estimates for PET (95%) were significantly higher than those for CT (76%) (P < 0.001) or MRI (78%) (P < 0.001) and the pooled specificity estimates for PET (90%) were significantly higher than those for CT (59%) (P < 0.001) or MRI (76%) (P < 0.001) (23). Moreover, in a prospective study comparing ¹⁸F-FDG PET, CT, and MRI for the initial evaluation of 134 oral SCC patients with a palpably negative neck, Ng et al. found ¹⁸F-FDG PET to have a sensitivity 2-fold higher than CT or MRI in the detection of nodal metastases (41.2% vs. 21.6%, respectively; P = 0.021) (24). These data confirm the possible superiority of PET in detecting asymptomatic recurrences.

Several posttreatment studies found ¹⁸F-FDG PET to have a similar accuracy for the detection of locoregional and distant recurrence of HNSSC. In these studies, sensitivities ranged from 92% to 100%, specificities from 64% to 100%, PPVs from 64% to 100%, NPVs from 92% to 100%, and overall accuracy from 88% to 91% (5–11). The lowest specificities are explained by false-positive findings in irradiated regions within 6 mo after therapy. Radiotherapy-induced inflammatory processes may lead to a transient increase in ¹⁸F-FDG uptake (25). In our study, the specificity was high because we performed, as proposed by Greven et al., the ¹⁸F-FDG PET at least 3 mo after the completion of treatment (26). However, in the recurrence studies, patients were commonly enrolled if recurrence was suspected. Lowe et al. and Ryan et al. have shown that ¹⁸F-FDG PET may be useful, too, in asymptomatic patients (18,27). Our results confirmed these data and the conclusion of our preliminary study (12).

CONCLUSION

The results of our study confirmed the high effectiveness of ¹⁸F-FDG PET/CT in assessing for recurrence of HNSCC in patients who have been considered cured of the disease. Our findings suggest that ¹⁸F-FDG PET/CT is more accurate than conventional follow-up physical examinations alone in such patients. The systematic use of PET/CT at 12 mo of the usual follow-up could be proposed, but cost-effectiveness and an evaluation of the survival impact remain to be evaluated.