- Email: [email protected]
The role of positron emission tomography in the evaluation of the N-positive neck
Otolaryngology– Head and Neck Surgery SEPTEMBER 2003
VOLUME 129
NUMBER 3
ORIGINAL ARTICLES The role of positron emission tomography in the evaluation of the N-positive neck MARK K. WAX,
MD,
LARRY L. MYERS,
MD,
JAYAKUMARI M. GONA,
MD,
SYED S. HUSAIN,
MD,
and HANI A. NABI,
MD, PHD,
Portland, Oregon, Dallas, Texas, and Buffalo, New York BACKGROUND: A major prognostic indicator in patients with squamous cell carcinoma of the upper aerodigestive tract is the presence or absence of cervical metastasis. Nodal involvement at different levels affects treatment. Thus identification of the degree of nodal involvement is important. Evaluation of the neck by conventional imaging modalities (computed tomography or magnetic resonance imaging) is not completely accurate. Positron emission tomography (PET) scanning as a dynamic functional assessment may allow detection of multiple metastatic nodes at different levels. PURPOSE: We sought to compare the effectiveness of PET with pathologic examination for: presence, location, and number of cervical metastases in the clinically N-positive neck. SETTING: Tertiary care academic facility.
From the Department of Otolaryngology, Oregon Health Sciences University, Department of Otolaryngology, University of Texas Southwestern, and Center for Positron Emission Tomography, Department of Nuclear Medicine, Veteran Affairs Western New York Health Care System, State University of New York at Buffalo. Presented at the Western Section of the Triologic Society Meeting, January 9, 2000, San Francisco, CA. Reprint requests: Mark K. Wax, MD, Microvascular Surgery, Department of Otolaryngology–Head and Neck Surgery, Oregon Health Sciences University, 3181 SW Sam Jackson Park Rd, PV-01, Portland, OR 97201. Copyright © 2003 by the American Academy of Otolaryngology–Head and Neck Surgery Foundation, Inc. 0194-5998/2003/$30.00 ⫹ 0 doi:10.1016/S0194-5998(03)00606-5
MATERIALS AND METHODS: From 1994 to 1997, 15 patients with clinically N-positive necks who had preoperative PET scans underwent 23 neck dissections. PET scans were correlated with the pathologic findings of the neck dissections in determining the ability to correctly identify the number and level(s) of nodal disease. RESULTS: When determining identification of the level of disease, PET demonstrated sensitivity of 81%; specificity, 99%; positive predictive value, 97%; negative predictive value, 90%; and accuracy, 92%. When evaluating the ability to correctly predict neck stage, PET demonstrated sensitivity of 86%, positive predictive value of 100%, and accuracy of 80% compared with clinical examination with sensitivity of 53% and accuracy of 53%. CONCLUSION: PET accurately identified disease in the N-positive neck. Its ability to identify multiple level disease may allow it to help predict the selectivity of neck dissection in the therapeutic protocol. (Otolaryngol Head Neck Surg 2003;129:163-7.)
A
major prognostic indicator for patients with squamous cell carcinoma of the upper aerodigestive tract is the presence or absence of cervical metastasis.1 The number, location, and histologic characteristics of these nodes influences the decision of whether to pursue ancillary treatment. Recently, there has been an increasing usage of selective neck dissection to determine nodal status. Preoperative determination of the level of nodal involvement is thus imperative in the management of these patients. Furthermore, in patients who are N positive, clinical 163
Otolaryngology– Head and Neck Surgery September 2003
164 WAX et al
Table 1. Demographic, radiologic, and pathologic data of the 15 patients who underwent PET scanning with subsequent pathologic examination of the neck dissection specimens Positive radiologic levels
Stage Patient
Clinical
Pathologic
Dissection side
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
T3 N2B TX N2A TX N2A T2 N2B T1 N2C T4 N2A T3 N2B T3 N2B T2 N1 T2 N2C T3 N2A T2 N2C T2 N1 T3 N2B T3 N1
N2B N2B N2B N2B N2C N2B N2B N2B N2B N2C N2C N2C N2C N2B N2C
Left Right Left Left Bilateral Right Left Bilateral Bilateral Bilateral Bilateral Bilateral Bilateral Right Bilateral
staging of the contralateral neck is important for treatment planning and patient counseling. Determining lymph node status solely on the basis of clinical examination is not very accurate. Falsenegative rates of up to 30% for single-node disease have been reported.2 Conventional imaging modalities, such as computed tomography (CT) and magnetic resonance imaging (MRI), have been helpful, but high false-positive/negative error rates are reported. The false-positive/negative error rate for CT is between 7.5% and 28%, and that for MR is 16%.3-5 Positron emission tomography (PET) is a functional imaging modality that has been applied to the pretreatment evaluation of extracranial head and neck neoplasms.6-8 The purpose of this study was to compare the effectiveness of PET scans with pathologic evaluation of the N-positive neck in determining the presence, location and number of cervical nodes in N-positive patients with squamous cell carcinoma (SCC) of the upper aerodigestive tract (UADT). MATERIALS AND METHODS From 1994 to 1997, 131 consecutive patients with a diagnosis of head and neck cancer underwent 171 PET scans at the Veterans Affairs Western New York Healthcare System (VAWNYHS) at Buffalo, NY. The patients received PET scans as part of an
Right
Left
Positive pathologic levels Right
II, III II, III
II II, III Neg Neg I, II II I Neg I, II I, II
Left
I-IV Neg I-IV Neg II I, III II I
Right
II, III II, III
II, III II, III II, III
ECS levels
III II, III Neg Neg I, II II I II I, IV I, II
Left
Pos Pos
II, III II, III II, III I-IV II, III I-IV I, II, IV II I, III, IV II I
Neg Pos
Pos Pos Neg Neg Neg Neg Pos
Pos Pos Neg Pos Neg Neg Neg Neg Neg Neg Neg
ongoing study to assess the usefulness of PET using 2-[F18]-fluoro-2-deoxy-D-glucose (FDG) in the evaluation of head and neck neoplasms. The institutional review board of VAWNYHS approved this larger study. Seventy-nine patients had previously untreated, biopsy-proved SCC of the upper aerodigestive tract. Of these 79 patients, 15 were both clinically N-positive and underwent neck dissections as part of their primary treatment. The 15 clinically N-positive patients underwent 23 neck dissections and comprise the study group. All patients had modified neck dissections with removal of levels I through V. The surgeon oriented the neck contents before being sent to pathology. The specimens were examined for the number and level(s) of positive nodes and for extracapsular spread. PET scans were correlated with the pathologic results of the neck dissections (Table 1). Statistical analysis was conducted as follows: sensitivity was calculated as the number of true-positive scans/(number of true-positive ⫹ number of falsenegative scans), specificity was calculated as truenegative scans/(true-negative scans ⫹ false-positive scans), positive predictive value (PPV) was calculated as true-positive scans/(true-positive scans ⫹ false-positive scans), negative predictive value (NPV) was calculated as true-negative scans/(truenegative scans ⫹ false-negative scans), and accu-
Otolaryngology– Head and Neck Surgery Volume 129 Number 3
WAX et al
Table 2. Location of primary tumor Site
Patients, n
(%)
Oropharynx Oral cavity Larynx Unknown primary
8 3 2 2
(54) (20) (13) (13)
racy (validity) was calculated as (true-positive scans ⫹ true-negative scans)/total. Statistical significance was assumed when P ⱕ 0.05. RESULTS Fifteen clinically N-positive patients underwent 23 neck dissections (Table 1). All patients were males with a mean age of 63.9 years (range, 43 to 83 years). Primary tumor sites are described in Table 2. Two patients were TX (14%). One patient (7%) was classified as T1, 5 patients (33%) as T2, 6 patients (40%) as T3, and 1 patient (7%) as T4. Three patients (20%) had N1 necks, whereas 12 patients (80%) had N2 necks. No patient in this study had N3 disease. Eight patients underwent bilateral neck dissections and 7 patients had a unilateral neck dissection (4 left, 3 right). We used the results of the pathologic examination of the neck contents to determine the sensitivity, specificity, PPV, NPV, and accuracy of physical examination and PET scanning. The following definitions were used: a true positive was when the PET was positive and pathology confirmed metastatic disease. A true negative was when the PET was negative and pathology confirmed no disease present. A false positive was when the PET was positive but there was no disease present on pathologic examination of the tissue. A false negative was when the PET was negative but disease was present on pathologic examination. Based on standard TNM staging, clinical examination of the neck revealed sensitivity of 53%, PPV of 100%, and accuracy of 53%. Using PET scan to stage the neck revealed sensitivity of 80%, PPV of 100%, and accuracy of 80%. It must be remembered that the study design preselected patients with clinically positive necks. Thus there were no false positives: this made the specificity and NPV 0 in both groups (Table 3). We then sought to examine the ability of PET scanning to detect and identify different levels of
165
Table 3. Comparison between physical examination and PET in determining N stage of the neck
Sensitivity Specificity PPV NPV Accuracy
PET
Physical examination
80% 0 100% 0 80%
53% 0 100% 0 53%
neck involvement. All neck specimens were oriented and identified before giving them to pathology. Each level of each side of the neck (1 to 5) was then determined to be positive or negative depending on the pathologic results. Activity on the PET scan was then correlated with the pathologic results to determine true/false positive/negative rates. Using this method, PET scan demonstrated the following: sensitivity of 81%; specificity, 99%; PPV, 97%; NPV, 90%; and accuracy, 92% (Table 4). Three patients (20%) (patients 9, 11, and 15) would have been upstaged preoperatively by PET. These patients were clinically N1 (2) or N2a. PET and pathology revealed them to be N2b (1) or N2c (2). DISCUSSION In the adult, normal-sized lymph nodes vary from 2 mm to 2 cm.2 In the neck of an average-sized adult and in the hands of an experienced examiner, the lower limit of palpability is approximately 0.5 cm in a superficial area such as the submental and submandibular area and 1 cm in deeper areas. Many factors (size, location, consistency, and the characteristics of the neck) determine whether lymph nodes can be palpated. Differences in patient position, prior radiation treatment, or the size of the neck contribute to interobserver variation. All of this makes determination of neck status by clinical examination not very accurate. With the unreliability of clinical examination, other modalities have been investigated. CT and MRI have been used to evaluate head and neck neoplasms as well as metastatic neck disease. Both are structural imaging modalities with known limitations. On CT, the criteria for a diagnosis of metastasis such as central necrosis, size greater than 1.5 cm, or tissue plane effacement may be simulated by fatty degeneration or lipid metaplasia (usually
Otolaryngology– Head and Neck Surgery September 2003
166 WAX et al
Table 4. Effectiveness of PET scan in determining the various levels of nodal involvement By level of disease
Sensitivity Specificity PPV NPV Accuracy
79% 97% 94% 89% 90%
seen after inflammatory disease or radiation therapy), an abscessed node, or spontaneous lymph node necrosis.3-5,9,10 Less than optimal design of the neck coil, artifacts secondary to patient movement, artifacts secondary to fat-suppression techniques, and flow-related artifacts present problems that degrade MR images to some degree.11 Despite its recent increased use, MRI has added surprisingly little to the diagnostic accuracy of contrast-enhanced CT.35,9,10 Although both CT and MR allow detection of abnormally enlarged nodes or necrotic nodes, neither borderline-sized nodes without necrosis nor extracapsular spread are reliably differentiated from reactive or normal nodes in patients with head and neck cancer. PET with FDG is a functional imaging modality that uses abnormal tissue metabolism to detect neoplasms. The radioactive glucose analog FDG is metabolized in normal tissue and neoplastic tissues in proportion to the rate of tissue glucose metabolism.11 FDG is metabolically trapped in the intracellular space. This occurs more in tumors than in normal tissues and can be used to identify tumors based on accelerated glycolytic rates using PET. PET has been used successfully in the evaluation of primary head and neck neoplasms.7,8 Wahl et al in 199011 proposed that PET-FDG could be further used to detect tumor involvement of lymph nodes. Other investigators7-14 have examined the role of PET in the detection of lymph node metastasis in patients with SCC of the UADT. These studies, however, have incorporated all N stages of disease. Myers et al13 reported on a select group of patients who were clinically N0. They reported that PET was superior to CT scanning in detecting occult cervical metastasis. All 7 patients with occult disease confirmed by pathologic examination of the specimen had positive PET scans. PET was also able to detect multilevel disease in 2 of the 3
patients in whom it was present pathologically. While this demonstrated the usefuless of PET in the detection of clinically unapparent disease, it did not address the issue of the ability of PET to detect multilevel or contralateral disease. Stokel et al7 evaluated 54 patients undergoing 81 neck dissections. They compared PET with CT and ultrasound-guided fine-needle aspiration. PET revealed: sensitivity of 96%, specificity of 90%, PPV of 85%, NPV of 98%, and accuracy of 93%. Unfortunately more than 60% of the necks they report on were N0. No data are given on individual patient neck staging. The large mix of N0 and N⫹ patients makes comparison with our data impossible. Furthermore, the usefulness of PET in their N⫹ patients cannot be determined. Paulins et al reviewed 25 previously untreated patients who had PET scans with subsequent pathologic correlation. Only 6 of their patients were clinically N positive. Two of these patients demonstrated multilevel or contralateral nodal involvement on pathologic examination. Clinical examination revealed only a single node, whereas PET scan correctly identified the multilevel or contralateral nature of the disease. Our study examined only patients who had clinically N⫹ disease and underwent neck dissection with removal of levels I to V. It demonstrated that PET had an overall sensitivity of 80%, PPV of 100%, and accuracy of 80%. PET was more sensitive (80% versus 53%) and accurate (80% versus 57%) in determining the correct N stage of the patient. The small sample size precluded statistical analysis. During the period of time that this study was performed, MR images were not available at our institution. Thus we were unable to evaluate the ability of MRI to detect disease in these patients. CT scans were routinely ordered. Unfortunately, 8 patients had their scans performed elsewhere and the scans were not available for review. To consider only 7 of the 15 patients in a comparison with PET was not thought to be warranted. In our patient population, physical examination was able to grossly detect that neck disease was present. The ability of physical examination to detect disease in the contralateral neck or disease that was present in many levels was poor. Because of this, selective or modified neck dissection was
Otolaryngology– Head and Neck Surgery Volume 129 Number 3
increasingly used to stage the neck. The pathologic results than determined whether ancillary treatment was undertaken. It also influenced patient counseling. One of the purposes of this study was to evaluate the ability of PET scans to detect multilevel or contralateral neck disease. PET proved to be very accurate in determining the number of levels and whether there was contralateral disease (Table 3). In our study, 3 of 15 (20%) of patients were upstaged based on the results of their PET scans. Patient 9 had a single node on physical examination. PET, confirmed by pathologic examination, revealed multiple nodes at multiple levels. While this may be interesting information, it would not have changed the treatment plan. Patients 11 and 15 were initially staged as N1 and N2a. Both patients demonstrated contralateral disease that was confirmed by pathologic examination. This information may very well affect treatment and prognosis. There were 3 patients in whom PET failed to detect disease in either the ipsilateral or the contralateral neck. In all of these patients, the diseased node or nodes were less than 1.5 cm in diameter on pathologic examination. The ability of PET to detect metastatic nodes is to some degree dependent on the size and location of the node. It was our impression that the undetected nodes on PET were either small or microscopic deposits only. Unfortunately, due to the retrospective nature of the study we cannot comment further on this. In this study, we demonstrated that PET has acceptable statistical criteria for use in evaluating the patient with cervical metastasis. PET consistently performed better than physical examination. We recognize that our study numbers are small and our patient population was biased. Follow-up on the prognostic value of the PET scan needs to be determined. It should be remembered that PET still is inferior to CT/MRI in determining the anatomic extent of the primary tumor. It may serve as an adjunct to the investigation of patients with head and neck cancer in determining multilevel or contralateral neck disease. CONCLUSIONS We obtained PET scans of 15 patients with Npositive necks with SCC of the UADT who under-
WAX et al
167
went 23 neck dissections. PET accurately identified disease in the N-positive neck. Its ability to accurately identify multiple level disease may allow it to help predict the selectivity of neck dissection in the therapeutic program. PET appears to be a useful diagnostic aid when evaluating the N⫹ neck for the presence of cervical metastasis. REFERENCES
1. Grandi C, Alloisio M, Moglia D, et al. Prognostic significance of lymphatic spread in head and neck carcinomas: therapeutic implications. Head Neck Surg 1985;8:67-73. 2. Hendrick JW. Occult cancer with cervical lymph node metastases. In: Conley J, editor. Cancer of the head and neck. Washington: Butterworths; 1967. pp. 41-55. 3. Giron J, Chisin R, Paul JL, et al. Multimodality imaging of cervical adenopathies. Eur J Radiol 1996;21:159-66. 4. Anzai Y, Brunberg JA, Lufkin RB. Imaging of nodal metastases in the head and neck. J Magn Reson Imaging 1997;7:774-83. 5. Som PM. Detection of metastasis in cervical lymph nodes: CT and MR criteria and differential diagnosis. AJR Am J Roentgenol 1992;158:961-9. 6. Bailet JW, Abemayor E, Jabour BA, et al. Positron emission tomography: a new, precise imaging modality for detection of primary head and neck tumors and assessment of cervical adenopathy. Laryngoscope 1992;102:281-8. 7. Stokkel MPM, Broek FW, Hordjik GJ, et al. Preoperative evaluation of patients with primary head and neck cancer using dual-head 18 fluorodeoxyglucose positron emission tomography. Ann Surg 2000;231:229-34. 8. Paulus R, Sambon A, Vivengnis D, et al. 18FDG-PET for the assessment of primary head and neck tumors: clinical, computed tomography, and histopathological correlation in 38 patients. Laryngoscope 1998;108:1578-83. 9. Braams JW, Pruim J, Nikkels PG, et al. Nodal spread of squamous cell carcinoma of the oral cavity detected with PET-tyrosine, MRI and CT. J Nucl Med 1996;37:897901. 10. Adams S, Baum RP, Stuckensen T, et al. 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 1998;25:1255-60. 11. Wahl RL, Kaminski MS, Ethier SP, et al. The potential of 2-deoxy-2[18F]fluoro-D-glucose (FDG) for the detection of tumor involvement in lymph nodes. J Nucl Med 1990; 31:1831-5. 12. Valdes Olmos RA, Koops W, Loftus BM, et al. Correlative 201Tl SPECT, MRI and ex vivo 201Tl uptake in detecting and characterizing cervical lymphadenopathy in head and neck squamous cell carcinoma. J Nucl Med 1999;40:1414-9. 13. Myers LL, Wax MK, Nabi H, et al. Positron emission tomography in the evaluation of the N0 neck. Laryngoscope 1998;108:232-6. 14. Benchaou M, Lehmann W, Slosman DO, et al. The role of FDG-PET in the preoperative assessment of N-staging in head and neck cancer. Acta Otolaryngol 1996;116: 332-5.
VOLUME 129
NUMBER 3
ORIGINAL ARTICLES The role of positron emission tomography in the evaluation of the N-positive neck MARK K. WAX,
MD,
LARRY L. MYERS,
MD,
JAYAKUMARI M. GONA,
MD,
SYED S. HUSAIN,
MD,
and HANI A. NABI,
MD, PHD,
Portland, Oregon, Dallas, Texas, and Buffalo, New York BACKGROUND: A major prognostic indicator in patients with squamous cell carcinoma of the upper aerodigestive tract is the presence or absence of cervical metastasis. Nodal involvement at different levels affects treatment. Thus identification of the degree of nodal involvement is important. Evaluation of the neck by conventional imaging modalities (computed tomography or magnetic resonance imaging) is not completely accurate. Positron emission tomography (PET) scanning as a dynamic functional assessment may allow detection of multiple metastatic nodes at different levels. PURPOSE: We sought to compare the effectiveness of PET with pathologic examination for: presence, location, and number of cervical metastases in the clinically N-positive neck. SETTING: Tertiary care academic facility.
From the Department of Otolaryngology, Oregon Health Sciences University, Department of Otolaryngology, University of Texas Southwestern, and Center for Positron Emission Tomography, Department of Nuclear Medicine, Veteran Affairs Western New York Health Care System, State University of New York at Buffalo. Presented at the Western Section of the Triologic Society Meeting, January 9, 2000, San Francisco, CA. Reprint requests: Mark K. Wax, MD, Microvascular Surgery, Department of Otolaryngology–Head and Neck Surgery, Oregon Health Sciences University, 3181 SW Sam Jackson Park Rd, PV-01, Portland, OR 97201. Copyright © 2003 by the American Academy of Otolaryngology–Head and Neck Surgery Foundation, Inc. 0194-5998/2003/$30.00 ⫹ 0 doi:10.1016/S0194-5998(03)00606-5
MATERIALS AND METHODS: From 1994 to 1997, 15 patients with clinically N-positive necks who had preoperative PET scans underwent 23 neck dissections. PET scans were correlated with the pathologic findings of the neck dissections in determining the ability to correctly identify the number and level(s) of nodal disease. RESULTS: When determining identification of the level of disease, PET demonstrated sensitivity of 81%; specificity, 99%; positive predictive value, 97%; negative predictive value, 90%; and accuracy, 92%. When evaluating the ability to correctly predict neck stage, PET demonstrated sensitivity of 86%, positive predictive value of 100%, and accuracy of 80% compared with clinical examination with sensitivity of 53% and accuracy of 53%. CONCLUSION: PET accurately identified disease in the N-positive neck. Its ability to identify multiple level disease may allow it to help predict the selectivity of neck dissection in the therapeutic protocol. (Otolaryngol Head Neck Surg 2003;129:163-7.)
A
major prognostic indicator for patients with squamous cell carcinoma of the upper aerodigestive tract is the presence or absence of cervical metastasis.1 The number, location, and histologic characteristics of these nodes influences the decision of whether to pursue ancillary treatment. Recently, there has been an increasing usage of selective neck dissection to determine nodal status. Preoperative determination of the level of nodal involvement is thus imperative in the management of these patients. Furthermore, in patients who are N positive, clinical 163
Otolaryngology– Head and Neck Surgery September 2003
164 WAX et al
Table 1. Demographic, radiologic, and pathologic data of the 15 patients who underwent PET scanning with subsequent pathologic examination of the neck dissection specimens Positive radiologic levels
Stage Patient
Clinical
Pathologic
Dissection side
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
T3 N2B TX N2A TX N2A T2 N2B T1 N2C T4 N2A T3 N2B T3 N2B T2 N1 T2 N2C T3 N2A T2 N2C T2 N1 T3 N2B T3 N1
N2B N2B N2B N2B N2C N2B N2B N2B N2B N2C N2C N2C N2C N2B N2C
Left Right Left Left Bilateral Right Left Bilateral Bilateral Bilateral Bilateral Bilateral Bilateral Right Bilateral
staging of the contralateral neck is important for treatment planning and patient counseling. Determining lymph node status solely on the basis of clinical examination is not very accurate. Falsenegative rates of up to 30% for single-node disease have been reported.2 Conventional imaging modalities, such as computed tomography (CT) and magnetic resonance imaging (MRI), have been helpful, but high false-positive/negative error rates are reported. The false-positive/negative error rate for CT is between 7.5% and 28%, and that for MR is 16%.3-5 Positron emission tomography (PET) is a functional imaging modality that has been applied to the pretreatment evaluation of extracranial head and neck neoplasms.6-8 The purpose of this study was to compare the effectiveness of PET scans with pathologic evaluation of the N-positive neck in determining the presence, location and number of cervical nodes in N-positive patients with squamous cell carcinoma (SCC) of the upper aerodigestive tract (UADT). MATERIALS AND METHODS From 1994 to 1997, 131 consecutive patients with a diagnosis of head and neck cancer underwent 171 PET scans at the Veterans Affairs Western New York Healthcare System (VAWNYHS) at Buffalo, NY. The patients received PET scans as part of an
Right
Left
Positive pathologic levels Right
II, III II, III
II II, III Neg Neg I, II II I Neg I, II I, II
Left
I-IV Neg I-IV Neg II I, III II I
Right
II, III II, III
II, III II, III II, III
ECS levels
III II, III Neg Neg I, II II I II I, IV I, II
Left
Pos Pos
II, III II, III II, III I-IV II, III I-IV I, II, IV II I, III, IV II I
Neg Pos
Pos Pos Neg Neg Neg Neg Pos
Pos Pos Neg Pos Neg Neg Neg Neg Neg Neg Neg
ongoing study to assess the usefulness of PET using 2-[F18]-fluoro-2-deoxy-D-glucose (FDG) in the evaluation of head and neck neoplasms. The institutional review board of VAWNYHS approved this larger study. Seventy-nine patients had previously untreated, biopsy-proved SCC of the upper aerodigestive tract. Of these 79 patients, 15 were both clinically N-positive and underwent neck dissections as part of their primary treatment. The 15 clinically N-positive patients underwent 23 neck dissections and comprise the study group. All patients had modified neck dissections with removal of levels I through V. The surgeon oriented the neck contents before being sent to pathology. The specimens were examined for the number and level(s) of positive nodes and for extracapsular spread. PET scans were correlated with the pathologic results of the neck dissections (Table 1). Statistical analysis was conducted as follows: sensitivity was calculated as the number of true-positive scans/(number of true-positive ⫹ number of falsenegative scans), specificity was calculated as truenegative scans/(true-negative scans ⫹ false-positive scans), positive predictive value (PPV) was calculated as true-positive scans/(true-positive scans ⫹ false-positive scans), negative predictive value (NPV) was calculated as true-negative scans/(truenegative scans ⫹ false-negative scans), and accu-
Otolaryngology– Head and Neck Surgery Volume 129 Number 3
WAX et al
Table 2. Location of primary tumor Site
Patients, n
(%)
Oropharynx Oral cavity Larynx Unknown primary
8 3 2 2
(54) (20) (13) (13)
racy (validity) was calculated as (true-positive scans ⫹ true-negative scans)/total. Statistical significance was assumed when P ⱕ 0.05. RESULTS Fifteen clinically N-positive patients underwent 23 neck dissections (Table 1). All patients were males with a mean age of 63.9 years (range, 43 to 83 years). Primary tumor sites are described in Table 2. Two patients were TX (14%). One patient (7%) was classified as T1, 5 patients (33%) as T2, 6 patients (40%) as T3, and 1 patient (7%) as T4. Three patients (20%) had N1 necks, whereas 12 patients (80%) had N2 necks. No patient in this study had N3 disease. Eight patients underwent bilateral neck dissections and 7 patients had a unilateral neck dissection (4 left, 3 right). We used the results of the pathologic examination of the neck contents to determine the sensitivity, specificity, PPV, NPV, and accuracy of physical examination and PET scanning. The following definitions were used: a true positive was when the PET was positive and pathology confirmed metastatic disease. A true negative was when the PET was negative and pathology confirmed no disease present. A false positive was when the PET was positive but there was no disease present on pathologic examination of the tissue. A false negative was when the PET was negative but disease was present on pathologic examination. Based on standard TNM staging, clinical examination of the neck revealed sensitivity of 53%, PPV of 100%, and accuracy of 53%. Using PET scan to stage the neck revealed sensitivity of 80%, PPV of 100%, and accuracy of 80%. It must be remembered that the study design preselected patients with clinically positive necks. Thus there were no false positives: this made the specificity and NPV 0 in both groups (Table 3). We then sought to examine the ability of PET scanning to detect and identify different levels of
165
Table 3. Comparison between physical examination and PET in determining N stage of the neck
Sensitivity Specificity PPV NPV Accuracy
PET
Physical examination
80% 0 100% 0 80%
53% 0 100% 0 53%
neck involvement. All neck specimens were oriented and identified before giving them to pathology. Each level of each side of the neck (1 to 5) was then determined to be positive or negative depending on the pathologic results. Activity on the PET scan was then correlated with the pathologic results to determine true/false positive/negative rates. Using this method, PET scan demonstrated the following: sensitivity of 81%; specificity, 99%; PPV, 97%; NPV, 90%; and accuracy, 92% (Table 4). Three patients (20%) (patients 9, 11, and 15) would have been upstaged preoperatively by PET. These patients were clinically N1 (2) or N2a. PET and pathology revealed them to be N2b (1) or N2c (2). DISCUSSION In the adult, normal-sized lymph nodes vary from 2 mm to 2 cm.2 In the neck of an average-sized adult and in the hands of an experienced examiner, the lower limit of palpability is approximately 0.5 cm in a superficial area such as the submental and submandibular area and 1 cm in deeper areas. Many factors (size, location, consistency, and the characteristics of the neck) determine whether lymph nodes can be palpated. Differences in patient position, prior radiation treatment, or the size of the neck contribute to interobserver variation. All of this makes determination of neck status by clinical examination not very accurate. With the unreliability of clinical examination, other modalities have been investigated. CT and MRI have been used to evaluate head and neck neoplasms as well as metastatic neck disease. Both are structural imaging modalities with known limitations. On CT, the criteria for a diagnosis of metastasis such as central necrosis, size greater than 1.5 cm, or tissue plane effacement may be simulated by fatty degeneration or lipid metaplasia (usually
Otolaryngology– Head and Neck Surgery September 2003
166 WAX et al
Table 4. Effectiveness of PET scan in determining the various levels of nodal involvement By level of disease
Sensitivity Specificity PPV NPV Accuracy
79% 97% 94% 89% 90%
seen after inflammatory disease or radiation therapy), an abscessed node, or spontaneous lymph node necrosis.3-5,9,10 Less than optimal design of the neck coil, artifacts secondary to patient movement, artifacts secondary to fat-suppression techniques, and flow-related artifacts present problems that degrade MR images to some degree.11 Despite its recent increased use, MRI has added surprisingly little to the diagnostic accuracy of contrast-enhanced CT.35,9,10 Although both CT and MR allow detection of abnormally enlarged nodes or necrotic nodes, neither borderline-sized nodes without necrosis nor extracapsular spread are reliably differentiated from reactive or normal nodes in patients with head and neck cancer. PET with FDG is a functional imaging modality that uses abnormal tissue metabolism to detect neoplasms. The radioactive glucose analog FDG is metabolized in normal tissue and neoplastic tissues in proportion to the rate of tissue glucose metabolism.11 FDG is metabolically trapped in the intracellular space. This occurs more in tumors than in normal tissues and can be used to identify tumors based on accelerated glycolytic rates using PET. PET has been used successfully in the evaluation of primary head and neck neoplasms.7,8 Wahl et al in 199011 proposed that PET-FDG could be further used to detect tumor involvement of lymph nodes. Other investigators7-14 have examined the role of PET in the detection of lymph node metastasis in patients with SCC of the UADT. These studies, however, have incorporated all N stages of disease. Myers et al13 reported on a select group of patients who were clinically N0. They reported that PET was superior to CT scanning in detecting occult cervical metastasis. All 7 patients with occult disease confirmed by pathologic examination of the specimen had positive PET scans. PET was also able to detect multilevel disease in 2 of the 3
patients in whom it was present pathologically. While this demonstrated the usefuless of PET in the detection of clinically unapparent disease, it did not address the issue of the ability of PET to detect multilevel or contralateral disease. Stokel et al7 evaluated 54 patients undergoing 81 neck dissections. They compared PET with CT and ultrasound-guided fine-needle aspiration. PET revealed: sensitivity of 96%, specificity of 90%, PPV of 85%, NPV of 98%, and accuracy of 93%. Unfortunately more than 60% of the necks they report on were N0. No data are given on individual patient neck staging. The large mix of N0 and N⫹ patients makes comparison with our data impossible. Furthermore, the usefulness of PET in their N⫹ patients cannot be determined. Paulins et al reviewed 25 previously untreated patients who had PET scans with subsequent pathologic correlation. Only 6 of their patients were clinically N positive. Two of these patients demonstrated multilevel or contralateral nodal involvement on pathologic examination. Clinical examination revealed only a single node, whereas PET scan correctly identified the multilevel or contralateral nature of the disease. Our study examined only patients who had clinically N⫹ disease and underwent neck dissection with removal of levels I to V. It demonstrated that PET had an overall sensitivity of 80%, PPV of 100%, and accuracy of 80%. PET was more sensitive (80% versus 53%) and accurate (80% versus 57%) in determining the correct N stage of the patient. The small sample size precluded statistical analysis. During the period of time that this study was performed, MR images were not available at our institution. Thus we were unable to evaluate the ability of MRI to detect disease in these patients. CT scans were routinely ordered. Unfortunately, 8 patients had their scans performed elsewhere and the scans were not available for review. To consider only 7 of the 15 patients in a comparison with PET was not thought to be warranted. In our patient population, physical examination was able to grossly detect that neck disease was present. The ability of physical examination to detect disease in the contralateral neck or disease that was present in many levels was poor. Because of this, selective or modified neck dissection was
Otolaryngology– Head and Neck Surgery Volume 129 Number 3
increasingly used to stage the neck. The pathologic results than determined whether ancillary treatment was undertaken. It also influenced patient counseling. One of the purposes of this study was to evaluate the ability of PET scans to detect multilevel or contralateral neck disease. PET proved to be very accurate in determining the number of levels and whether there was contralateral disease (Table 3). In our study, 3 of 15 (20%) of patients were upstaged based on the results of their PET scans. Patient 9 had a single node on physical examination. PET, confirmed by pathologic examination, revealed multiple nodes at multiple levels. While this may be interesting information, it would not have changed the treatment plan. Patients 11 and 15 were initially staged as N1 and N2a. Both patients demonstrated contralateral disease that was confirmed by pathologic examination. This information may very well affect treatment and prognosis. There were 3 patients in whom PET failed to detect disease in either the ipsilateral or the contralateral neck. In all of these patients, the diseased node or nodes were less than 1.5 cm in diameter on pathologic examination. The ability of PET to detect metastatic nodes is to some degree dependent on the size and location of the node. It was our impression that the undetected nodes on PET were either small or microscopic deposits only. Unfortunately, due to the retrospective nature of the study we cannot comment further on this. In this study, we demonstrated that PET has acceptable statistical criteria for use in evaluating the patient with cervical metastasis. PET consistently performed better than physical examination. We recognize that our study numbers are small and our patient population was biased. Follow-up on the prognostic value of the PET scan needs to be determined. It should be remembered that PET still is inferior to CT/MRI in determining the anatomic extent of the primary tumor. It may serve as an adjunct to the investigation of patients with head and neck cancer in determining multilevel or contralateral neck disease. CONCLUSIONS We obtained PET scans of 15 patients with Npositive necks with SCC of the UADT who under-
WAX et al
167
went 23 neck dissections. PET accurately identified disease in the N-positive neck. Its ability to accurately identify multiple level disease may allow it to help predict the selectivity of neck dissection in the therapeutic program. PET appears to be a useful diagnostic aid when evaluating the N⫹ neck for the presence of cervical metastasis. REFERENCES
1. Grandi C, Alloisio M, Moglia D, et al. Prognostic significance of lymphatic spread in head and neck carcinomas: therapeutic implications. Head Neck Surg 1985;8:67-73. 2. Hendrick JW. Occult cancer with cervical lymph node metastases. In: Conley J, editor. Cancer of the head and neck. Washington: Butterworths; 1967. pp. 41-55. 3. Giron J, Chisin R, Paul JL, et al. Multimodality imaging of cervical adenopathies. Eur J Radiol 1996;21:159-66. 4. Anzai Y, Brunberg JA, Lufkin RB. Imaging of nodal metastases in the head and neck. J Magn Reson Imaging 1997;7:774-83. 5. Som PM. Detection of metastasis in cervical lymph nodes: CT and MR criteria and differential diagnosis. AJR Am J Roentgenol 1992;158:961-9. 6. Bailet JW, Abemayor E, Jabour BA, et al. Positron emission tomography: a new, precise imaging modality for detection of primary head and neck tumors and assessment of cervical adenopathy. Laryngoscope 1992;102:281-8. 7. Stokkel MPM, Broek FW, Hordjik GJ, et al. Preoperative evaluation of patients with primary head and neck cancer using dual-head 18 fluorodeoxyglucose positron emission tomography. Ann Surg 2000;231:229-34. 8. Paulus R, Sambon A, Vivengnis D, et al. 18FDG-PET for the assessment of primary head and neck tumors: clinical, computed tomography, and histopathological correlation in 38 patients. Laryngoscope 1998;108:1578-83. 9. Braams JW, Pruim J, Nikkels PG, et al. Nodal spread of squamous cell carcinoma of the oral cavity detected with PET-tyrosine, MRI and CT. J Nucl Med 1996;37:897901. 10. Adams S, Baum RP, Stuckensen T, et al. 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 1998;25:1255-60. 11. Wahl RL, Kaminski MS, Ethier SP, et al. The potential of 2-deoxy-2[18F]fluoro-D-glucose (FDG) for the detection of tumor involvement in lymph nodes. J Nucl Med 1990; 31:1831-5. 12. Valdes Olmos RA, Koops W, Loftus BM, et al. Correlative 201Tl SPECT, MRI and ex vivo 201Tl uptake in detecting and characterizing cervical lymphadenopathy in head and neck squamous cell carcinoma. J Nucl Med 1999;40:1414-9. 13. Myers LL, Wax MK, Nabi H, et al. Positron emission tomography in the evaluation of the N0 neck. Laryngoscope 1998;108:232-6. 14. Benchaou M, Lehmann W, Slosman DO, et al. The role of FDG-PET in the preoperative assessment of N-staging in head and neck cancer. Acta Otolaryngol 1996;116: 332-5.