Defining the Role of FDG PET in Head and Neck Cancer

PII S1095-0397(99)00033-3

Clinical Positron Imaging Vol. 2, No. 6, 311–316. 1999 Copyright  2000 Elsevier Science Inc. Printed in the USA. All rights reserved. 1095-0397/99 $–see front matter

ORIGINAL ARTICLE

Defining the Role of FDG PET in Head and Neck Cancer Homer A. Macapinlac, MD, Henry W. D. Yeung, MD, Steven M. Larson, MD Nuclear Medicine Service, Department of Radiology, Memorial Sloan Kettering Cancer Center, New York, NY

Abstract The purpose of this article is to elucidate the role of 2-[18F] Fluoro-2-deoxy-D-glucose (FDG) positron emission tomography (PET) in evaluating patients with head and neck cancer. This will include background information on patient preparation and image acquisition. Normal patterns of uptake will be described in reference to computerized tomography (CT)/magnetic resonance imaging (MRI) to illustrate the relationship of physiology to the anatomic landmarks. Common clinical scenarios will be discussed, including staging, identifying recurrence, monitoring therapy, and secondary cancers. The fundamental basis of this imaging modality is the altered metabolism of tumor tissue, which includes an increase in glycolysis. FDG is a glucose analog, which is essentially trapped within tumor cells with increased glycolysis, and allows these malignant cells to be localized. Conventional methods, such as CT and MRI, are dependent on distortion of the normal anatomy in identifying the presence or absence of tumor. CT scanning for example is essentially dependent on nodal enlargement criteria for localizing lymph node metastasis and is less specific in identifying tumor in normal sized lymph nodes or in enlarged nodes not involved by tumor. Following surgery, radiation, and/or chemotherapy, the normal anatomy is distorted and these conventional methods become less specific in distinguishing recurrence from post treatment changes. FDG PET can provide more accurate information in identifying tumor before and after treatment. This improvement in specificity augments our current ability to stage the extent of disease at presentation and monitor response to therapy.1 However, the anatomic resolution of these conventional methods are superior to FDG PET, and this advantage should be used to accurately localize both the normal and abnormal findings on FDG PET.

Address correspondence to: Homer A. Macapinlac, MD, Box 77, Nuclear Medicine Service, Memorial Sloan Kettering Cancer Center, 1275 York Avenue, New York, NY 10021.

Key Words: FDG; PET; Head; Neck; Cancer.

Patient Preparation Nondiabetic patients are fasted for at least 6 hours prior to FDG injection. This is done because FDG is a glucose analog, which competes with native glucose for its accumulation and is also dependent on insulin levels. Fasting results in low serum glucose and insulin levels, which allows good tumor uptake with minimal skeletal muscle activity (See Figure 1). The patients are encouraged to drink water prior to the injection of FDG as this minimizes the concentration of the tracer in the urinary collecting system, and more may be voided prior to imaging. Diabetic patients are also asked to titrate their glucose levels to normal range at the scheduled time of injection. Our current protocol at Memorial Sloan Kettering Cancer Center (MSKCC) involves intravenous (IV) catheter placement in the antecubital fossa using aseptic technique. A blood glucose sample is obtained and then 10 mCi of FDG is injected (bolus) followed by a 10 ml saline flush. The residual activity is then measured in the dose calibrator to determine the net injected dose used for calculating standardized uptake values (SUV). The patient is then kept in a sitting or recumbent position and is asked not to talk, nor drink. This minimizes FDG uptake in the muscles of the tongue, and the vocalis muscles of the larynx. Minimizing patient movement during the uptake phase decreases background skeletal muscle activity in general. However, it is advisable to inform the patient not to perform vigorous physical activity just prior to getting the FDG injection as the muscle activity will also be higher than normal.2

PET Imaging The uptake phase is usually 45–60 minutes and the patients are imaged in a supine position and advised to try as much as possible not to move during the entire imaging procedure, which is usually no more than 30 minutes. At MSKCC, we use a dedicated PET (General 311

312 Clinical Positron Imaging, Volume 2, Number 6

Figure 1. This is a 52-year-old woman with a sarcoma of the right arm. The 2 coronal FDG PET images (left) show intense uptake in the skeletal muscles and myocardium taken following the patient’s ingestion of food (initially denied on history) approximately 30 minutes prior to FDG injection. Two coronal images (right) of the same patient following a repeat scan with a 6-hour fast prior to injection of FDG. Note the minimal uptake of FDG in the skeletal muscles and myocardium.

Electric, Milwaukee, WI) Advance scanner.3 Our standard imaging acquisition parameters are 6-minute emission scans with single slice overlap at every bed position. The scanning starts from the base of the skull (with our external anatomic landmark for technologists being the top of the ear) and proceeds caudally for 3 bed positions, which will essentially include both lung fields. We rou-

tinely perform 4-minute transmission scans over the same areas for attenuation correction.

Image Analysis The vendor-provided software allows transmission scans to be displayed (which appear like a crude CT)

Figure 2. This is a 37-year-old woman with thyroid cancer post completion thyroidectomy and left vocal cord paralysis. Attenuationcorrected transaxial FDG PET scans (left to right, top to bottom) from the base of the skull to the larynx. First row shows extraocular muscle uptake, and the uptake in the adenoids. The second row shows uptake in the palatine tonsils. Third row shows uptake in the digastric muscles in the floor of the mouth. Fourth row shows uptake in the right vocalis muscle of the larynx, and note the cervical spinal cord uptake posteriorly.

PET in Head Neck Cancer / Macapinlac et al. 313

Figure 3. Same patient as in Figure 2 with FDG PET scans showing mediastinal uptake (short arrow) and right vocalis muscle uptake (arrow) correlating with the patient’s left vocal cord paralysis. Corresponding transmissions scans are routinely used as a guide to localize uptake, with good delineation of the lungs and airway.

alongside the attenuation-corrected scans to allow anatomic localization of activity. The images should be reviewed in all standard planes and volumetric projection if necessary. In selected cases when tumor localization is difficult, the transmission study can be co-registered with the most recent MRI or CT and to eventually allow reslicing of the attenuation-corrected scans to planes identical to the MRI or CT for accurate localization. The MSKCC co-registration method was previously described4 and in summary consists of converting CT or MRI scans into a common file format (minc.*). These datasets are then transferred directly to a HPApollo 9000 Model 735 (Hewlett Packard, Palo Alto, CA) and PV Wave (Visual Numerics Inc., Boulder, CO) based on a house-developed program originally done by Pelizzari and Chen.5 The PV wave program is used to generate resliced PET images that correspond to appropriate original CT or MRI transaxial slices. Sagittal and coronal registration are also possible with this method. Regions of interest are manually drawn along the border of the identified abnormality and in body scans, lesions are considered indicative of malignancy if the SUV max is ⬎3.5 and benign if ⬍3.5. The standard

uptake value is calculated based on the following formula: SUV ⫽

decay corrected dose (mCi)/(ml) tumor injected dose (mCi)/body weight (gm)

Normal Patterns of Uptake and Variants Familiarization with the normal distribution of FDG in untreated as well as treated patients is important. Likewise, the availability of the MRI or CT scan is fundamentally important to the proper interpretation of these scans. There is normal high activity in neural tissues, which include the cerebellar hemispheres, temporal lobes of the brain, and mild activity is consistently identified in the cervical spinal cord. Variable muscle activity is noted but is usually identified in the vocalis muscles of the larynx, extraocular muscles, and the surface of the tongue (saliva). Normal lymphoid activity is also present, eg, Waldeyer’s ring and thymic activity. This uptake pattern appears most pronounced in younger patients ⬍40 years of age. Normal vocalis muscle uptake is usually clearly delineated enough to detect vocal cord paralysis (See Figures 2 and 3). What is important is to adhere to the patient preparation procedures so

314 Clinical Positron Imaging, Volume 2, Number 6

Figure 4. Patient with known left floor of the mouth cancer was referred for staging. FDG PET scans show uptake in the primary tumor and ipsilateral local nodes (arrow). The patient had resection of the primary with a modified left neck dissection revealing only the local nodal metastasis and absence of any distal nodal involvement.

that normal background activity variation is minimized. However, normal muscle activity is high in patients who are tense. Usually with experience, one can read around these areas because they are essentially in the shape and form of known muscle groups, and with experience in knowing where nodal activity is usually expected. However, there are situations wherein the muscle activity is so intense and nodular that it becomes difficult to make a distinction. Our group has given oral diazepam (5–10 mg) in order to minimize normal muscle activity. This is not done routinely, but if the referring physician advises us that the patient is tense or claustrophobic, then premedication is done for these patients. FDG PET imaging can detect acute inflammatory processes, which can exceed the uptake of tumor. Chronic granulomatous conditions, such as sarcoidosis,6 can mimic tumor if active. The nuclear medicine physician can optimize his or her interpretation of the scan by taking an adequate history and physical exam. The nuclear medicine physician has to determine if the patient has any evidence or history of an active infection, as we need to know the chronology and

extent of surgery and radiation. Acute post radiation effects are usually manifested as more superficial skin activity. Whereas a more remote radiation therapy, ⬎2 weeks is usually noted as increased muscle activity around the neck area, particularly in larynx carcinoma patients undergoing radiation therapy. We perform PET scans usually 6–8 weeks following radiation therapy to minimize background activity and 2 months or more following chemotherapy to minimize marrow activity, which can substantially diminish our ability to detect bony metastasis.

FDG PET in Head and Neck Cancer Head and neck cancers are found in increasing frequency (5% of all malignancies) and are clearly associated with the use of tobacco and/or excess alcohol ingestion.7 The most important adverse prognostic factors usually include increasing T stage and N stage, which are based on inspection and palpation when possible, and both by indirect mirror examination and direct endoscopy when necessary. Clearly, the anatomic imaging modalities supplement the ability to

PET in Head Neck Cancer / Macapinlac et al. 315

Figure 5. Baseline CT scan of a patient with base of tongue cancer and persistent left cervical adenopathy following chemotherapy and radiation. The patient had a FDG PET scan, that showed no abnormal uptake in the neck, which was verified by the negative modified neck dissection. Repeat scans after neck dissection document the absence of any residual tumor in the neck.

stage the extent of disease, particularly the appropriate nodal drainage areas of the neck. The treatment options are dependent on the extent of disease and have evolved to become multimodality, ie, combining surgery, radiotherapy, in early stage disease and chemotherapy being added for advanced disease. The ability of FDG PET to improve the accuracy of the pretreatment evaluation is documented in several prospective studies with relative sensitivities and specificities better than physical examination and or CT/ MRI (See Figure 4). Recently, a prospective study of 60 patients who had histologically proven squamous cell CA of the head and neck was done to compare preoperative FDG PET with CT, MRI, and sonography.8 The resected neck specimens revealed a total of 1284 lymph nodes, 117 of which showed metastatic involvement. Their study indicated FDG PET correctly identified lymph node metastases with a sensitivity of 90% and a specificity of 94%. CT and MRI showed nodal metastases with a sensitivity of 82% (specificity 85%) and 80% (specificity 79%), respectively. This study indicates FDG PET is the procedure with the highest sensitivity and specificity for detecting lymph node metastases of head and neck. A prospective study by Wong9 (31 patients with primary disease and 23 patients with suspected recurrence) demonstrated that FDG PET was more accurate than CT and MRI for identifying both primary and recurrent tumors as well as metastatic lesions in the neck. The potential incremental advantage of PET over conventional imaging appears to be more promising in the posttreatment setting when previous surgery, radiation, or chemotherapy results in an alteration of the anatomy, which diminishes the specificity and sensitivity of CT or MRI. This was essentially the conclusion

Figure 6. Fifty-nine year-old man with biopsy proven right cervical nodal metastasis was referred for identification of possible primary after the CT and MRI failed to identify the primary tumor. The patient had 2 previous biopsies; first the nasopharynx, and then Rosenmuller’s fossa bilaterally. The FDG PET scan showed uptake in the posterior nasopharynx (arrow) seen on the sagittal scan. The tumor was confirmed at surgery wherein the tumor was found under a normalappearing posterior nasopharyngeal mucosa.

316 Clinical Positron Imaging, Volume 2, Number 6

of a review of several prospective series by McGuirt10 in that FDG PET scanning is comparable to conventional imaging of head and neck cancers in detecting primary and metastatic disease with its greatest role being in the evaluation of the postradiotherapy patient. Anzai11 showed that in a small series of 12 patients with suspected recurrence, comparing FDG PET with MRI or CT, FDG PET yielded a sensitivity and specificity of 88% and 100%, respectively. Whereas, MRI and/or CT, demonstrated a sensitivity and specificity of 25% and 75%, respectively. Similar findings were seen by Lapela12 in 15 patients with suspected recurrence of head and neck cancers. Their results showed a sensitivity of 88% and a specificity of 86%. A study by Lowe13 demonstrated promising results in the accuracy of FDG PET in classifying response to chemotherapy in 26 patients with advanced head and neck cancer. They used tissue biopsies to document complete response or residual disease (See Figure 5). The sensitivity and specificity of PET for residual cancer after therapy was 90% and 83%, respectively. Two patients had initial negative biopsies and PET showed persistent disease. Pathology review and a repeat biopsy led to confirmation of the PET results in these cases, giving a sensitivity of 90% for initial tissue biopsy. Another area of interest is the detection of an occult primary tumor in patients with known cervical metastases. Braams14 showed that in 13 patients studied, FDG PET identified the primary tumor in 30 % of patients who had undergone conventional imaging. Our experience at MSKCC is similar in identifying primary tumors (See Figure 6), as FDG PET provides useful information that can lead to appropriate treatment and serve as a guide for biopsy to obtain proper diagnosis. The extent of the PET literature found for head and neck cancer is substantial and provides an adequate basis for its more widespread use clinically. FDG PET imaging also allows better monitoring of treatment response and its clinical impact will be greater in the future because as novel approaches to therapy are developed, the biologic response parameters that PET provide appear to be the most promising among the current imaging modalities. Supported in Part by DOE#DE-FG02-86ER60407 and the Laurent and Alberta Gerschel Foundation.

References 1. Larson, S.M. Positron emission tomography in oncology and allied disease. In: De Vita, V.T., Hellman, S., Rosenberg, S.A., eds. Cancer, principles and practice of oncology, Vol. 3. Philadelphia: J.B. Lippincott Publishing; 1989:1–12. 2. Engel, H.; Steinert, H.; Buck, A.; Berthold, T.; Huch Boni, R.A.; von Schulthess, G.K. Whole-body PET: physiological and artifactual fluorodeoxyglucose accumulations. J. Nucl. Med. 37:441–6; 1996. 3. DeGrado, T.R.; Turkington, T.G.; Williams, J.J.; et al. Performance characteristics of a whole body PET scanner. J. Nucl. Med. 35:1398–406; 1994. 4. Scott, A.M.; Macapinlac, H.A.; Zhang, J.J.; et al. Clinical applications of fusion imaging in oncology. Nucl. Med. Biol. 21:775–84; 1994. 5. Pelizzari, C.A.; Chen, G.T.Y.; Spelbring, D.R.; Weischelbaum, R.R.; Chen, C.T. Accurate three dimensional registration of CT, PET and/or MRI images of the brain. J. Comput. Assist. Tomogr. 14:20–6; 1989. 6. Lewis, P.J.; Salama, A. Uptake of fluorine-18-fluorodeoxyglucose in sarcoidosis. J. Nucl. Med. 35:1647–9; 1994. 7. Spitz, M.R. Epidemiology and risk factors for head and neck cancer. Semin. Oncol. 21:281–8; 1994. 8. Adams, S.; Baum, R.P.; Stuckensen, T.; Bitter, K.; Hor, G. Prospective comparison of 18F-FDG PET with conventional imaging modalities (CT, MRI, US) in lymph node staging of head and neck cancer. Eur. J. Nucl. Med. 25:1255–60; 1998. 9. Wong, W.L.; Chevretton, E.B.; McGurk, M.; et al. A prospective study of PET-FDG imaging for the assessment of head and neck squamous cell carcinoma. Clin. Otolaryngol. 22:209–14; 1997. 10. McGuirt, W.F.; Greven, K.; Williams, D. III. PET scanning in head and neck oncology: a review. Head Neck 20: 208–15; 1998. 11. Anzai, Y.; Carroll, W.R.; Quint, D.J.; et al. Recurrence of head and neck cancer after surgery or irradiation: prospective comparison of 2-deoxy-2-[F-18]fluoro-D-glucose PET and MR imaging diagnoses. Radiology 200:135–41; 1996. 12. Lapela, M.; Grenman, R.; Kurki, T.; et al. Head and neck cancer: detection of recurrence with PET and 2-[F-18]fluoro-2-deoxy-D-glucose. Radiology 197:205–11; 1995. 13. Lowe, V.J.; Dunphy, F.R.; Varvares, M.; et al. Evaluation of chemotherapy response in patients with advanced head and neck cancer using [F-18] fluorodeoxyglucose positron emission tomography. Head Neck 19:666–74; 1997. 14. Braams, J.W.; Pruim, J.; Kole, A.C.; et al. Detection of unknown primary head and neck tumors by positron emission tomography. Int. J. Oral Maxillofac. Surg. 26:112– 5; 1997.