FOCUS ON: Nuclear Medicine and Molecular Imaging
Review
FDG PET/CT in the Management of Nasopharyngeal Carcinoma
Audio Available | 
Share
OBJECTIVE. FDG PET/CT has a growing role in the diagnosis and management of nasopharyngeal carcinoma (NPC). FDG PET has greater efficacy for N and M staging than other modalities, which enables the treating oncologists to select the appropriate mode of treatment. FDG PET/CT helps in radiation therapy planning, provides valuable prognostic information, and is useful in the assessment of therapy response and in follow-up to detect recurrences.
CONCLUSION. FDG PET/CT is a valuable imaging test in the management of NPC.
Keywords: nasopharyngeal cancer, patient outcome, PET/CT, radiation therapy planning
Nasopharyngeal carcinoma (NPC) is a unique cancer differing from other head and neck cancers in its epidemiology, cause, clinical behavior, and treatment. Cancer of the nasopharynx resulted in 65,000 deaths in 2010, which was an increase from 45,000 deaths in 1990 [1]. NPC is a leading form of cancer in certain regions of the world such as the Cantonese population of Southern China and Hong Kong where the incidence is reported to be as high as 20 cases per 100,000 person-years. It is rare in most other regions of the world, including the United States, with incidence rates of less than 1 case per 100,000 person-years [2].
The World Health Organization classifies NPC into the following types: type 1, squamous cell carcinoma; type 2a, keratinizing undifferentiated carcinoma; and type 2b, nonkeratinizing undifferentiated carcinoma. Type 1 is less common and has the worst prognosis, whereas types 2a and 2b are more common and tend to have a better prognosis [3]. In endemic areas, undifferentiated carcinomas tend to dominate, accounting for as many as 93% of NPC cases [4]. In contrast, differentiated carcinomas account for as many as 50% of the cases in nonendemic areas [5]. Undifferentiated NPC is associated with Epstein-Barr virus (EBV) infection, whereas differentiated NPC is associated with smoking and alcohol intake. FDG PET/CT is useful in the management of many human solid tumors [6–15], and in this article we review the evolving role of FDG PET and FDG PET/CT in the management of patients with NPC.
| Role of Conventional Imaging |   |
|---|
Conventional imaging is useful in the management of patients with NPC. Because of its excellent spatial and soft-tissue contrast resolution, MRI has become an integral part in the management of NPC. Whereas the utility of CT is limited mainly to staging and radiotherapy planning, MRI plays a role in diagnosis, staging, treatment planning, and prognostication. MRI has been found to have high sensitivity, specificity, and accuracy of 100%, 93%, and 95%, respectively, in diagnosing NPC [16]. These rates are comparable with those of endoscopy with biopsy, which had corresponding values of 95%, 100%, and 98%, respectively [16]. MRI aids in the accurate staging of NPC by detecting primary tumor spread; parapharyngeal, orbital, and paranasal involvement; and spread to lymph nodes (LNs), especially the retropharyngeal LNs [17, 18]. The dosages of radiotherapy and the decision to add concurrent chemotherapy to the treatment regimen are dependent on the stage of the NPC. The gross tumor volume (GTV) can be outlined using MRI and CT and the planned treatment volume to be irradiated is decided. The volume of the primary tumor detected by MRI and CT is an independent prognostic indicator [19].
| Role of FDG PET/CT | |
|---|
Staging of NPC follows the American Joint Committee on Cancer (AJCC) TNM staging manual [20]. Chen et al. [21] assessed overall TNM staging in 70 cases of NPC (20 newly diagnosed cases and 50 cases in patients who had undergone treatment) and compared the efficacy of fused FDG PET/CT and CT alone. The sensitivity, specificity, accuracy, positive predictive value, and negative predictive value of FDG PET/CT (96%, 94%, 95%, 96%, and 94%, respectively) were significantly better than those of CT (71%, 76%, 73%, 80%, and 67%; p < 0.05) [21].
Because of its superior spatial and soft-tissue contrast resolution, MRI is the imaging modality of choice to delineate the extent of the primary tumor [22, 23] (Fig. 1). A study conducted by Ng et al. [22] of 111 patients compared the efficacy of FDG PET/CT and MRI in T staging of NPC. FDG PET/CT resulted in downstaging in 17–23% of cases and upstaging in 8–10% of cases. They detected discrepancies between MRI and FDG PET findings in bony structures, such as the skull base, in 16% of patients, in the intracranial region in 14% patients, and in the parapharyngeal region in 19% of patients [22]. Underestimation of tumor mostly occurs in the skull base and cavernous sinuses and may be because of poor FDG uptake in areas of early-stage tumor invasion and because of the lower resolution of PET compared with MRI. There is also physiologically high uptake of FDG by the brain, which may obscure the tumor. Overestimation of the tumor mainly occurs in the parapharyngeal region and is the result of the inability of FDG PET/CT to resolve soft-tissue details and differentiate tumor compression of the region from tumor invasion.
 View larger version (161K) | Fig. 1A —25-year-old man who underwent staging MRI and FDG PET/CT for evaluation of recently diagnosed nasopharyngeal carcinoma (arrows). A, Anterior maximum-intensity-projection PET image (A), axial fused PET/CT image (B), and coronal (C) and axial (D) T1-weighted contrast-enhanced MR images. MR images reveal large nasopharyngeal mass with involvement of skull base and prevertebral space. Mass extends into intracranial compartment in sella turcica and cavernous sinuses. Multiple, bilateral, enlarged cervical lymph nodes are noted. FDG PET/CT image reveals large hypermetabolic nasopharyngeal mass extending into skull base and bilateral enlarged FDG-avid supraclavicular and cervical lymphadenopathy; these findings are consistent with metastatic disease. |
 View larger version (202K) | Fig. 1B —25-year-old man who underwent staging MRI and FDG PET/CT for evaluation of recently diagnosed nasopharyngeal carcinoma (arrows). B, Anterior maximum-intensity-projection PET image (A), axial fused PET/CT image (B), and coronal (C) and axial (D) T1-weighted contrast-enhanced MR images. MR images reveal large nasopharyngeal mass with involvement of skull base and prevertebral space. Mass extends into intracranial compartment in sella turcica and cavernous sinuses. Multiple, bilateral, enlarged cervical lymph nodes are noted. FDG PET/CT image reveals large hypermetabolic nasopharyngeal mass extending into skull base and bilateral enlarged FDG-avid supraclavicular and cervical lymphadenopathy; these findings are consistent with metastatic disease. |
 View larger version (356K) | Fig. 1C —25-year-old man who underwent staging MRI and FDG PET/CT for evaluation of recently diagnosed nasopharyngeal carcinoma (arrows). C, Anterior maximum-intensity-projection PET image (A), axial fused PET/CT image (B), and coronal (C) and axial (D) T1-weighted contrast-enhanced MR images. MR images reveal large nasopharyngeal mass with involvement of skull base and prevertebral space. Mass extends into intracranial compartment in sella turcica and cavernous sinuses. Multiple, bilateral, enlarged cervical lymph nodes are noted. FDG PET/CT image reveals large hypermetabolic nasopharyngeal mass extending into skull base and bilateral enlarged FDG-avid supraclavicular and cervical lymphadenopathy; these findings are consistent with metastatic disease. |
 View larger version (327K) | Fig. 1D —25-year-old man who underwent staging MRI and FDG PET/CT for evaluation of recently diagnosed nasopharyngeal carcinoma (arrows). D, Anterior maximum-intensity-projection PET image (A), axial fused PET/CT image (B), and coronal (C) and axial (D) T1-weighted contrast-enhanced MR images. MR images reveal large nasopharyngeal mass with involvement of skull base and prevertebral space. Mass extends into intracranial compartment in sella turcica and cavernous sinuses. Multiple, bilateral, enlarged cervical lymph nodes are noted. FDG PET/CT image reveals large hypermetabolic nasopharyngeal mass extending into skull base and bilateral enlarged FDG-avid supraclavicular and cervical lymphadenopathy; these findings are consistent with metastatic disease. |
Chan et al. [24] assessed the relationship between T staging and metabolic parameters obtained through FDG PET in 57 patients with NPC. There was a positive correlation between T staging and metabolic tumor volume (MTV) (r = 0.504, p < 0.001), maximum standardized uptake value (SUVmax) (r = 0.516, p < 0.001), and total lesion glycolysis (TLG) (r = 0.620, p < 0.001). Multivariate analysis showed a significant association with MTV and SUVmax (R2 = 0.370, p < 0.001) [24]. These results suggest a potentially complementary role of FDG PET/ CT in the assessment of the primary tumor.
FDG PET/CT has high accuracy rates in assessing cervical nodes in patients with NPC. Assessment of LNs by MRI and CT relies on parameters—such as nodal size and enhancement—that may lead to the failure to detect small nodes. FDG PET/CT has a sensitivity of 97–100% and specificity of 73–97% in assessing cervical nodes in patients with NPC, and MRI has a sensitivity of 84–92% and specificity of 73–97% [22, 25, 26].
However, the accuracy of FDG PET/CT varies depending on the site of nodal spread, with FDG PET/CT being particularly useful in assessing lower cervical node metastases. In a study of 84 patients with primary NPC, Yen et al. [25] compared the accuracy of MRI and FDG PET/CT in detecting spread to various nodal groups. Dual-phase FDG PET/CT was performed and the sensitivity, specificity, and accuracy for the detection of spread to the lower cervical LNs of the second phase was 100%, whereas these performance values were 84%, 98%, and 90%, respectively, for MRI (p = 0.046) [25]. In their study of 111 patients with NPC, Ng et al. [22] found that MRI and FDG PET/CT findings differed with respect to cervical nodal metastases in 12 patients (10.8%). FDG PET/CT was proved to be correct in 10 of these cases; in the remaining two patients, FDG PET/CT incorrectly suggested the presence of metastases [22].
One of the reported limitations of FDG PET/CT in nodal staging is in the retropharyngeal LNs where involvement is not detected as accurately as on MRI [22, 23]; this limitation is likely because of the blooming artifact in FDG PET/CT that can result in difficulties in differentiating a primary nasopharyngeal mass from a retropharyngeal LN. In a study conducted by Ng et al. [22] FDG PET/CT had sensitivity and specificity of 88% and 94%, respectively, when compared with 96% and 100% for MRI. Of the 111 patients in that study, FDG PET/CT and MRI findings were discordant in the retropharyngeal area in 14 patients (13%) and FDG PET/CT findings were wrong in 11 of those patients [22]. Hence, reviewing the retropharyngeal nodes carefully when interpreting FDG PET/CT studies is important. Contrast-enhanced FDG PET/CT would improve the detectability of retropharyngeal LNs.
FDG PET/CT has excellent diagnostic accuracy in M staging of NPC compared with conventional imaging [27–29] (Fig. 2). Whole-body FDG PET/CT has the ability to assess both the primary tumor and the distant metastases in a single examination. A meta-analysis conducted by Chang et al. [30] elucidates the diagnostic accuracy of FDG PET and FDG PET/CT in staging of distant metastases in NPC. They analyzed eight studies with a total of 1069 patients with 172 identified distant metastases: The pooled sensitivity, specificity, positive likelihood ratio, and negative likelihood ratio were 83%, 97%, 23.38, and 0.19, respectively. A summary receiver operating characteristic (SROC) graph was plotted, and the Q index for FDG PET or FDG PET/CT in M staging of NPC was found to be 0.9307, indicating excellent performance for the modalities. The authors also analyzed the efficacy of FDG PET or FDG PET/CT by region of metastasis and reported pooled sensitivity, specificity, positive likelihood ratio, and negative likelihood ratio of 78%, 98%, 40.32, and 0.24 for bones and 53%, 100%, 135.9, and 0.47 for the liver [30]. In a location-based analysis, conducted by Ng et al. [22], FDG PET/CT was found to have the highest sensitivity for the mediastinum (100%), followed by bone (89%), lung (86%), and liver (50%).
 View larger version (292K) | Fig. 2A —30-year-old man who underwent staging FDG PET/CT for evaluation of recently diagnosed nasopharyngeal mass. A, Anterior maximum-intensity-projection PET image (A) and axial (B and C) images (top row, CT images; middle row, PET images; bottom row, fused PET/CT images). PET/CT images reveal intense FDG activity (maximum standardized uptake value = 10.2) within left parapharyngeal mass with hypermetabolic metastatic foci in cervical and mediastinal lymph nodes and within numerous lung nodules; these findings are consistent with primary nasopharyngeal malignancy with local and distant metastases. Despite systemic treatment, disease progressed, resulting in death of patient 15 months after diagnosis. |
 View larger version (129K) | Fig. 2B —30-year-old man who underwent staging FDG PET/CT for evaluation of recently diagnosed nasopharyngeal mass. B, Anterior maximum-intensity-projection PET image (A) and axial (B and C) images (top row, CT images; middle row, PET images; bottom row, fused PET/CT images). PET/CT images reveal intense FDG activity (maximum standardized uptake value = 10.2) within left parapharyngeal mass with hypermetabolic metastatic foci in cervical and mediastinal lymph nodes and within numerous lung nodules; these findings are consistent with primary nasopharyngeal malignancy with local and distant metastases. Despite systemic treatment, disease progressed, resulting in death of patient 15 months after diagnosis. |
 View larger version (252K) | Fig. 2C —30-year-old man who underwent staging FDG PET/CT for evaluation of recently diagnosed nasopharyngeal mass. C, Anterior maximum-intensity-projection PET image (A) and axial (B and C) images (top row, CT images; middle row, PET images; bottom row, fused PET/CT images). PET/CT images reveal intense FDG activity (maximum standardized uptake value = 10.2) within left parapharyngeal mass with hypermetabolic metastatic foci in cervical and mediastinal lymph nodes and within numerous lung nodules; these findings are consistent with primary nasopharyngeal malignancy with local and distant metastases. Despite systemic treatment, disease progressed, resulting in death of patient 15 months after diagnosis. |
Conventional imaging workup for distant metastases involves a combination of chest radiography, ultrasound, CT, and bone scanning. Thus, multiple imaging techniques need to be used for assessment. Moreover, the sensitivity is unacceptably low with values ranging from 25% to 67% across these studies [22, 27–29]. The skeletal system is one of the more common sites of distant metastases in NPC. Conventional skeletal scintigraphy using 99mTc-labeled bisphosphonates is a commonly used modality to detect metastatic spread to the bone. Liu et al. [28] assessed 202 patients with NPC and found that 30 (15%) had bone metastases. Both FDG PET and skeletal scintigraphy had comparable specificity of 99% and 98% (p = 0.687), respectively. However, the sensitivity was significantly higher for FDG PET (70%) than for skeletal scintigraphy (37%; p = 0.006) [28]. This significant difference illustrates the importance of FDG PET in the detection of skeletal metastatic disease to the bone.
Thus, FDG PET and FDG PET/CT have a crucial role in M staging of NPC, and because the treatment and prognosis vary widely depending on the presence of distant metastases, FDG PET and FDG PET/CT have grown to have an indispensable role in patient management. The current clinical practice is to identify patients at an increased risk of meta-static spread, especially those with high T and N stages along with abnormal metabolic parameters and EBV levels. Whole-body FDG PET/CT is then performed and the high sensitivity of the modality gives clinicians the ability to detect possible metastases confidently.
FDG PET and FDG PET/CT have the potential to change management in patients with NPC when compared with conventional imaging because of their superior ability to detect nodal and distant metastases [31, 32]. Law et al. [31] conducted a study to assess the utility of FDG PET in the management of NPC. Of 48 patients with NPC who underwent imaging before starting treatment, the management strategy was changed in 33% of patients as a result of the findings on FDG PET/CT. These changes in management strategies were made because of new findings related to metastases (8%) and nodal spread (25%) [31]. Gordin et al. [32] conducted a study to assess the impact of FDG PET/CT in the management of NPC. Of the 33 patients with NPC they studied, FDG PET/CT altered the management in 57%. It eliminated the need for previously planned diagnostic procedures in 33% of the patients, induced a change in the planned therapeutic approach in 15% of the patients, and guided biopsy to a specific metabolically active area that was in an edematous region in 9% of the patients. Of the five patients (15%) with FDG PET/CT findings that led to a change in therapeutic approach, FDG PET/CT newly detected distant metastases in three patients, which changed the management from curative to palliative. One patient had a lymph node that was negative on fine-needle aspiration cytology but positive on FDG PET. Neck dissection was performed and the node was found to be malignant. Negative FDG PET/CT findings obviated neck dissection in one patient with clinically suspected nodal metastases [32]. FDG PET/CT plays a valuable role in the systemic staging of patients with NPC.
Radiotherapy is the mainstay in the treatment of localized NPC. However, multimodality treatment with a combination of chemotherapy and radiotherapy is needed for locally advanced NPC. The current recommendations are to provide radiotherapy only for T1 tumors without nodal or distant metastases and to provide concurrent chemoradiotherapy to any patient with evidence of nodal spread and T stage greater than T1 [33]. A regimen of concurrent chemoradiotherapy with cisplatin followed by adjuvant chemotherapy with cisplatin and 5-fluorouracil has been established as the standard for locally advanced disease in a phase III randomized study [34]. Intensity-modulated radiotherapy treatment (IMRT) allows improved target coverage, particularly in the skull base, while sparing adjacent areas from late radiation injury including the parotid glands to limit xerostomia [35]. The current practice is to deliver a total radiation dose of approximately 70 Gy to the GTV with a margin to create a clinical and planning target volume [36]. A critical factor in treatment planning is accurately delineating the GTV that needs to be irradiated.
FDG PET/CT is increasingly used in radiotherapy planning of NPC (Fig. 3). Hung et al. [37] assessed the correlation between tumor volumes delineated by CT and FDG PET/CT and its relation to overall survival (OS) in 32 patients with NPC. The summation-of-area technique was used to delineate the tumor volume in CT, and Hung et al. referred to this method as “ VolumeCT.” FDG PET/CT volumes were similarly obtained using three methods: threshold SUV ≥ 2.5, which they referred to as “ Volume2.5”; ≥ 40% of SUVmax, “Volume40”; and ≥ 50% of SUVmax, “Volume50.” The average volumes calculated by VolumeCT, Volume2.5, Volume40, and Volume50 were 16.48 ± 12.46 cm3, 25.87 ± 16.96 cm3, 13.66 ± 6.90 cm3, and 8.25 ± 4.52 cm3, respectively. When compared with VolumeCT, Volume2.5 was significantly larger (p = 0.0003) and Volume50 was significantly smaller (p = 0.0006). The correlation between VolumeCT and Volume2.5 was good (r = 0.64, p = 0.0001), and the correlation between VolumeCT and Volume50 was poor (r = 0.23, p = 0.20). No significant difference was detected between VolumeCT and Volume40 (p = 0.24), but the correlation was poor (r = 0.15, p = 0.39) [37]. Thus, it may be reasonable to include regions with an SUV ≥ 2.5 or ≥ 50% of SUVmax to assist in determining the contours of the GTV for radiotherapy planning. However, clinical examination findings and integration of other important standard imaging modalities such as MRI and CT should also be carefully used to determine the contours of the primary disease.
 View larger version (299K) | Fig. 3A —Radiation therapy planning. A, Axial fused FDG PET/CT image shows nasopharyngeal mass and target volumes. Red volume represents nasopharyngeal lesion that is FDG-avid on PET. Aqua contour represents planning target volume, which is optimized to be treated to 70 Gy. Pink and purple contours outline parotid glands as organs at risk; objective is to spare parotid glands by using mean dose of 26–30 Gy. Bilateral hemi-necks will also be treated: 70 Gy to gross disease and 56–60 Gy to elective nodal areas. |
 View larger version (316K) | Fig. 3B —Radiation therapy planning. B, Intensity-modulated radiation therapy (IMRT) plans shown on planning CT. Red isodose line represents 70-Gy isodose, and aqua line represents 60-Gy isodose line. Upper IMRT field is matched to low anterior 3D conformal photon field. Spinal cord, bilateral parotid glands, and oral cavity are spared with sharp dose gradients between planning target volumes and adjacent organs at risk. In C, A = anterior. In D, L = lateral. |
 View larger version (264K) | Fig. 3C —Radiation therapy planning. C, Intensity-modulated radiation therapy (IMRT) plans shown on planning CT. Red isodose line represents 70-Gy isodose, and aqua line represents 60-Gy isodose line. Upper IMRT field is matched to low anterior 3D conformal photon field. Spinal cord, bilateral parotid glands, and oral cavity are spared with sharp dose gradients between planning target volumes and adjacent organs at risk. In C, A = anterior. In D, L = lateral. |
 View larger version (250K) | Fig. 3D —Radiation therapy planning. D, Intensity-modulated radiation therapy (IMRT) plans shown on planning CT. Red isodose line represents 70-Gy isodose, and aqua line represents 60-Gy isodose line. Upper IMRT field is matched to low anterior 3D conformal photon field. Spinal cord, bilateral parotid glands, and oral cavity are spared with sharp dose gradients between planning target volumes and adjacent organs at risk. In C, A = anterior. In D, L = lateral. |
FDG PET has the potential to assess the response of the tumor early during treatment (Fig. 4). The SUVmax corresponds well with response of tumor to treatment, with greater decreases in SUVmax correlating well with better survival [38–40]. Xie et al. [40] assessed metabolic response of NPC to radiotherapy by stratifying treated patients into two groups. Of 58 patients in the study with posttreatment scans, those who showed a decrease in SUVmax of more than 2.5 were classified as metabolic complete responders and the others as metabolic partial responders. The timing of posttreatment scans ranged between 1 and 5 months with a median of 2.1 months. Five-year OS and disease-free survival (DFS) were 74% and 65%, respectively, in metabolic complete responders and 46% and 38% in metabolic partial responders (p = 0.027 and p = 0.018) [40]. Therefore, FDG PET can accurately assess response to therapy, which gives prognostic information for patients.
 View larger version (283K) | Fig. 4A —Therapy assessment in 49-year-old man with nasopharyngeal mass. A, Axial fused PET/CT image obtained as part of initial FDG PET/CT study for staging evaluation of newly diagnosed nasopharyngeal mass reveals hypermetabolic nasopharyngeal mass (arrow) (maximum standardized uptake value = 5.28). Patient underwent concurrent chemoradiation therapy. |
 View larger version (281K) | Fig. 4B —Therapy assessment in 49-year-old man with nasopharyngeal mass. B, Axial fused PET/CT image obtained 9 months after initial staging study (A) to assess response to therapy shows good treatment response with interval resolution of FDG activity (arrow). |
| Prognosis | |
|---|
The metabolic information about the tumor derived from FDG PET scans has been found to be extremely valuable in predicting prognosis in patients. SUVmax, TLG, and MTV are some of the parameters that have been shown to correlate with survival. SUVmax is a parameter that has been researched extensively in prognostication of NPC cases. TLG is the product of the mean SUV and the MTV. Thus, TLG, as a single value, incorporates both the SUV and MTV of the tumor and is potentially a better prognostic marker. A summary of the studies assessing the value of SUVmax in predicting prognosis in NPC is outlined in Table 1 and those showing the value of TLG and MTV in assessing prognosis are given in Table 2. Hung et al. [41] conducted a study with 371 patients to assess the prognostic value of SUVmax obtained on FDG PET before patients started undergoing IMRT. Using receiver operating characteristic analysis, they determined an optimal SUVmax value of 9.3 for the primary tumor and 7.4 for the nodes. Keeping theses values as the standard, 5-year distant metastases–free survival (DMFS) and OS were found to be significantly lower for patients with primary tumor above the cutoff (84% and 77%, respectively) than those with tumor below the cutoff (91% and 86%) (p = 0.045 and p = 0.002). When the SUVmax of the LNs was assessed, the 5-year DMFS was 82% for patients with values above the cutoff and 94% for those below it (p = 0.001). The difference in the OS between the two groups was found not significant [41].
Chan et al. [42] assessed the value of various prognostic factors, including TLG, in 165 patients with NPC. They used 330 g as the cutoff value for TLG and determined 5-year OS and DFS of 55% and 44.4%, respectively, for patients with values above the cutoff and 86.1% and 79.3% for those with TLG values below it (p < 0.001). Multivariate Cox proportional hazards regression showed that TLG was an independent prognostic factor for OS (hazard ratio [HR] = 3.497, p = 0.002) [42].
Another prognostic factor that has been investigated recently is the heterogeneity of FDG uptake by a tumor, and the results indicate that it has the potential to play an important role. Huang et al. [43] studied 40 cases of NPC to assess the role of intratumoral heterogeneity. Heterogeneity of intratumoral FDG uptake is represented by the heterogeneity factor, which was defined as a derivative of the volume threshold function of the tumor. The heterogeneity factor significantly correlated to SUVmax (r = –0.372, p = 0.018), tumor volume (r = –0.983, p < 0.001), and T stage (r = –0.457, p = 0.003). A log-rank test showed that patients with lower heterogeneity factors (i.e., more heterogeneous tumor) had a significantly lower DFS (55%) than those with higher heterogeneity factors (89%) when using –0.24 as the threshold (p < 0.05) [43].
Plasma concentrations of EBV DNA correlate well with tumor burden and prognosis of NPC [44]. Recent evidence suggests that metabolic parameters obtained from FDG PET scans correlate well with EBV DNA levels, which lends further weight to using FDG PET/CT for determining prognosis. Chang et al. [45] assessed the utility of plasma EBV levels and metabolic parameters in 108 patients. Linear regression analysis found an association between total TLG and plasma EBV DNA levels (R2 = 0.589, p < 0.05). Multivariate analysis revealed a significant association between OS and TLG (HR = 4.911, p = 0.045) and EBV DNA (HR = 4.721, p = 0.035) [45]. Thus, metabolic information obtained from the tumor by FDG PET correlates well with plasma EBV DNA levels and contributes to prognostication for patients with NPC.
| Follow-Up | |
|---|
Identifying residual disease and recurrence and differentiating them from postradiation changes are of paramount importance in the follow-up period. FDG PET is the diagnostic modality of choice for these purposes (Figs. 5 and 6). Liu et al. [46] conducted a meta-analysis that compared the diagnostic capability of FDG PET in detecting residual disease and recurrence of NPC with that of CT and MRI. A total of 1813 patients met their inclusion criteria. The pooled sensitivity, specificity, and diagnostic odds ratio for FDG PET were 95%, 90%, and 96.51, respectively. The sensitivity, specificity, and diagnostic odds ratio for CT were 76%, 59%, and 7.01, respectively, and those for MRI were 78%, 76% and 10.33 (p < 0.001). SROC curves were plotted, with the area under the curve (AUC) for PET being 0.9696. The Q index for FDG PET (0.92) was significantly higher than those for CT (0.72, p < 0.001) and MRI (0.76, p < 0.01) [46]. The superior diagnostic accuracy of FDG PET in detecting recurrences illustrates the important role that it can play in the surveillance of NPC. FDG PET, with its ability to assess the metabolic function of regions, can help differentiate a recurrence from postradiotherapy changes. Fibrotic tissue is expected to have minimal or no uptake of FDG, whereas tumors can be expected to be FDG-avid [46].
 View larger version (221K) | Fig. 5A —Recurrent local disease in 60-year-old man with history of nasopharyngeal carcinoma who underwent restaging FDG PET/CT. A, Anterior maximum-intensity-projection PET image (A), CT image (top row, B), PET image (middle row, B), and fused PET/CT image (bottom row, B) obtained for follow-up after selective neck dissection and radiotherapy. PET/CT image reveals no evidence of FDG-avid recurrent disease. |
 View larger version (119K) | Fig. 5B —Recurrent local disease in 60-year-old man with history of nasopharyngeal carcinoma who underwent restaging FDG PET/CT. B, Anterior maximum-intensity-projection PET image (A), CT image (top row, B), PET image (middle row, B), and fused PET/CT image (bottom row, B) obtained for follow-up after selective neck dissection and radiotherapy. PET/CT image reveals no evidence of FDG-avid recurrent disease. |
 View larger version (243K) | Fig. 5C —Recurrent local disease in 60-year-old man with history of nasopharyngeal carcinoma who underwent restaging FDG PET/CT. C, Anterior maximum-intensity-projection PET image (C), CT image (top row, D), PET image (middle row, D), and axial fused PET/CT image (bottom row, D) obtained at restaging FDG PET/CT 1.5 years after A and B reveal metabolically active (maximum standardized uptake value = 12.8) left oropharyngeal thickening (arrows) extending to posterior nasopharyngeal wall; these findings are consistent with recurrent malignancy. |
 View larger version (119K) | Fig. 5D —Recurrent local disease in 60-year-old man with history of nasopharyngeal carcinoma who underwent restaging FDG PET/CT. D, Anterior maximum-intensity-projection PET image (C), CT image (top row, D), PET image (middle row, D), and axial fused PET/CT image (bottom row, D) obtained at restaging FDG PET/CT 1.5 years after A and B reveal metabolically active (maximum standardized uptake value = 12.8) left oropharyngeal thickening (arrows) extending to posterior nasopharyngeal wall; these findings are consistent with recurrent malignancy. |
 View larger version (254K) | Fig. 6A —Recurrent metastatic disease in 62-year-old woman with history of nasopharyngeal squamous cell carcinoma treated with chemoradiation. Patient underwent restaging FDG PET/CT study for evaluation of recent-onset neck and scalp masses and hip pain. A, Anterior maximum-intensity-projection PET image (A) and axial fused PET/CT images (B). PET/CT images reveal numerous foci of intense FDG activity involving osseous axial and proximal appendicular skeleton, cranium, lungs, mediastinal and abdominal lymph nodes, and right hepatic lobe; these findings are consistent with widespread metastatic involvement. |
 View larger version (213K) | Fig. 6B —Recurrent metastatic disease in 62-year-old woman with history of nasopharyngeal squamous cell carcinoma treated with chemoradiation. Patient underwent restaging FDG PET/CT study for evaluation of recent-onset neck and scalp masses and hip pain. B, Anterior maximum-intensity-projection PET image (A) and axial fused PET/CT images (B). PET/CT images reveal numerous foci of intense FDG activity involving osseous axial and proximal appendicular skeleton, cranium, lungs, mediastinal and abdominal lymph nodes, and right hepatic lobe; these findings are consistent with widespread metastatic involvement. |
However, these findings are not uniform, and variations in the accuracy of detection of residual disease and recurrence occur between MRI and FDG PET in different situations. In a study conducted by Chan et al. [47] including 146 patients, the efficacy of FDG PET to detect recurrent NPC was compared with that of MRI after stratification according to the initial T stage of the disease. In patients with an initial T4 stage, FDG PET had a significantly higher specificity for recurrence than MRI (96% vs 63%, respectively; p = 0.04). In patients with an initial T stage of T1–T3 disease, FDG PET and MRI had a similar specificity; however, on further analysis, specificity was 72% for FDG PET in patients with T1 or T2 disease who had undergone intracavitary brachytherapy as opposed to 98% for MRI in those who were treated with other modalities (p = 0.003). The authors hypothesized that the severe focal inflammation due to uneven distribution of brachytherapy dose mimics malignancy at FDG PET [47].
The time to scanning is an important factor in reducing the number of false-positive findings of NPC recurrence with FDG PET. Inflammation in the early postradiotherapy period actively takes up glucose, leading to false-positive results on FDG PET. Therefore, the current practice is to wait 3–4 months after treatment before the first follow-up FDG PET study. For a meta-analysis, Liu et al. [46] chose studies only if FDG PET had been performed 3–4 months after radiotherapy, which accounts for the high accuracy rates seen.
Wang et al. [48] assessed a surveillance strategy including a combination of plasma EBV DNA levels and FDG PET. After treatment of NPC, plasma EBV DNA levels were analyzed every 3–6 months and FDG PET was performed when EBV DNA levels were positive. Two hundred forty-five patients with previously treated NPC underwent plasma EBV DNA monitoring, and 36 patients who had positive EBV DNA levels and five patients who had no detectable EBV DNA levels but had clinical features strongly suggestive of recurrence underwent FDG PET. Plasma EBV DNA levels detected NPC recurrence with 100% accuracy. The sensitivity, specificity, and accuracy of FDG PET by visual interpretation were 82%, 77%, and 79%, respectively. The authors recommended that EBV DNA level assessment be used to identify patients with recurrence and then FDG PET be performed to identify the site and characteristics of the lesion. This recommendation is a cost-effective strategy for surveillance after therapy [48].
After confirmation of recurrence, reirradiation is a potential treatment option. FDG PET/CT has a role to play in delineating GTV in recurrent cases. Because the radiation field must be limited to only gross disease with a minimal margin, accurately defining the tumor volume is very critical in the reirradiation setting. Zheng et al. [49] compared the GTV delineated by CT and FDG PET/CT in 39 cases of recurrent NPC. The average size of GTV was 15.9 cm3 on CT, whereas it was 13.7 cm3 on FDG PET/CT. Geographic misses of tumor were more frequent with CT than with FDG PET/CT for both gross tumor (18% vs 10%, respectively) and microscopic tumor (51% vs 33%) [49].
Recurrent tumors tend to be resistant to radiotherapy; therefore, surgical salvage therapy with nasopharyngectomy in resectable cases is the treatment option of choice in these cases. SUVmax determined on FDG PET and plasma EBV values have been proved to be useful in determining the prognosis of such cases treated by surgery. Chan et al. [50] conducted a study involving 64 patients with recurrent NPC treated with nasopharyngectomy. They discovered that patients with curative surgery had significantly lower preoperative log plasma EBV DNA values (2.2 vs 3.4, respectively; p = 0.013) and mean SUVmax (4.3 vs 6.9, p = 0.009) than patients with recurrence [50]. Overall, FDG PET plays a role in the long-term follow-up and surveillance of NPC and aids in the management of recurrent cases.
Hybrid PET/MRI has been introduced in clinical practice recently. It allows clinicians to combine PET—a sensitive tool for locoregional and distant metastases staging, therapy assessment, and follow-up—with MRI, which provides detailed depictions of tissue planes that can reveal local spread of primary tumor at the skull base and intracranial involvement. Although no published studies are currently available, we anticipate this modality will play a cohesive role in the management of NPC.
| Conclusion | |
|---|
FDG PET and FDG PET/CT have a growing role in the diagnosis and management of NPC. MRI and FDG PET have complementary roles, with MRI contributing to T staging and FDG PET having greater efficacy for N and M staging. FDG PET can provide the GTV that needs to be irradiated, and the accurate changes to staging that result from FDG PET compared with conventional imaging enable the treating oncologists to select the appropriate mode of treatment. The various metabolic parameters from FDG PET scans obtained before treatment have been shown to provide valuable prognostic information. FDG PET/CT is also useful in assessing the response to treatment. A combination of EBV DNA levels and FDG PET can effectively monitor patients during follow-up to detect recurrence and can help in planning treatment and assessing prognosis in recurrent cases. With the introduction of clinical PET/MRI, we anticipate PET/MRI will play a significant role in the management of NPC in the future by combining the benefits of PET and MRI.
WEB
This is a web exclusive article.
| References | |
|---|
