Introduction
The intensity of diagnostic procedures involved in primary cancer staging remains a controversial issue. The presence of distant metastases should alter the selection of the therapy, and also has an impact on quality of life when there is no prospect of cure. The prognosis of patients with distant metastases from primary head and neck squamous cell carcinoma (SCC) is generally poor, with less than two years survival after diagnosis.Reference Calhoun, Fulmer, Weiss and Hokanson1, Reference Merino, Lindberg and Fletcher2
Only a limited number of studies have evaluated the incidence of distant metastases at the time of initial presentation. Distant metastases are noted in 7–17 per cent of head and neck SCC patients at presentation.Reference Bhatia and Bahadur3–Reference Dennington, Carter and Meyers7 The overall incidence of clinically detected distant metastases from head and neck SCC ranges from 4 to 24 per cent. The lung is the most common site (52–91 per cent), followed by bone (19–36 per cent) and liver (6–20 per cent). Autopsy studies demonstrated the incidence of distant metastases to be as high as 57 per cent. The incidence of distant metastases is directly related to the tumour stage, particularly in patients with advanced node (N) negative stage disease.Reference Ferlito, Shaha, Silver, Rinaldo and Mondin8
A variety of diagnostic techniques are established, routine practice for the initial staging of patients with head and neck SCC. These include: chest X-ray or computed tomography (CT) for evaluation of the lungs; bone scanning and plasma bone-specific alkaline phosphatase levels for bone investigation; and abdominal ultrasonography, CT or magnetic resonance imaging (MRI) plus serum liver function tests for liver evaluation. The role of whole body [F-18]-fluorodeoxyglucose positron emission tomography in combination with CT scanning is still under investigation. The choice of initial staging examinations is influenced by tumour stage, the patient's general condition, the probability of a change in treatment if distant metastases are detected, institutional guidelines and socioeconomic factors.
The purpose of this study was to evaluate the role of bone scanning, computed tomography of the thorax and abdominal ultrasonography in the initial staging of patients presenting with untreated, advanced head and neck SCC.
Materials and methods
One hundred and sixty-three patients (40 women and 123 men) with head and neck SCC were scheduled for major surgery and underwent screening for distant metastases between January 2000 and December 2004. Patients' mean age was 57 years and ranged from 25 to 90 years. Twenty-nine patients had stage II head and neck SCC, 36 had stage III and 98 had stage IV. Primary tumour sites included the oral cavity (n = 18), oropharynx (n = 62), hypopharynx (n = 25), larynx (n = 38), paranasal sinuses (n = 5) and nasopharynx (n = 15).
All patients with histologically confirmed head and neck SCC underwent screening for distant metastases.
Computed tomography of the head and neck region and thorax was performed in all 163 patients, in order to evaluate primary locoregional tumour extension as well as lung and mediastinal status. Spiral CT scans were obtained with a fourth-generation Siemens Somaton Plus machine (Siemens AG, Erlangen, Germany) after intravenous administration of contrast medium (Ultravist 370; Schering AG, Berlin, Germany). Continuous axial scanning planes were used at 125-mm slice thickness without interslice gap. Radiological criteria were: for lung metastases, multiple, smooth and mostly peripherally located lesions; for bronchogenic carcinoma, solitary, speculated and mostly centrally located lesions; and for mediastinal lymph node metastases, a diameter of more than 10 mm.
In addition, all patients underwent bone scanning, according to the bone scanning procedure guidelines of the Austrian Society of Nuclear Medicine. Bone scans were acquired with a dedicated full-ring scanner, one hour after intravenous injection of 550 MBq Tc-99 m diphosphonate. When no explanation for a scintigraphic abnormality could be found (e.g. osteolysis, osteoarthritis or degenerative changes), plain film images were taken.
In order to evaluate the possibility of liver metastases, 159 patients received an abdominal ultrasound.
Statistical analyses were accomplished using SPSS® software. Correlation analyses were calculated using the chi-square test.
Results
Distant metastases were found in nine patients (5.52 per cent). All of these patients had advanced locoregional disease and stage IV tumour. The exact distribution is shown in Table I.
Table I Results of screening for distant metastases
Pt no = patient number; SCC = squamous cell carcinoma; TNM = tumour–node–metastasis; M = male; F = female
Computed tomography of the thorax revealed lung metastases in six patients (3.68 per cent). Mediastinal lymph node metastases were not detected in any patient.
Abdominal ultrasonography showed metastatic lesions in one patient without any sign of primary abdominal tumour. The other ultrasound examinations showed haemangioma, cysts and liver steatosis as additional diagnoses.
Bone scanning detected abnormalities in 80 patients. For further clarification, spot films and plain X-ray films of different anatomical regions (i.e. hip, shoulder, leg, rib and spine) were made. These additional examinations revealed two cases of metastases in the cervical spine, also detected by computed tomography. In one patient, multiple metastases in the lumbar spine were incidentally detected by the orthopaedic surgeons; this patient's bone scan had been falsely negative for this anatomical region. In total, bone metastases were found in 1.84 per cent of our patients.
Statistical analysis yielded no significant correlation between advanced locoregional disease and the occurrence of distant metastases. This fact may have been influenced by the high number of stage IV tumours in the cohort.
Discussion
The high prevalence of additional malignant tumours in patients with newly diagnosed head and neck SCC, and the relatively poor detection rate of such tumours by conventional chest radiography, support the routine inclusion of chest CT in primary staging. In one study, only 29 per cent of 66 malignant tumours detected upon chest CT scanning of 189 patients had been detected by chest radiography.Reference Reiner, Siegel, Sawyer, Brocato, Maroney and Hooper9
Halpern et al. recommended an investigative chest CT prior to initiating any definitive therapy.Reference Halpern10 In this study, nine of 24 patients (37.5 per cent) with advanced head and neck SCC and negative plain chest X-rays had positive findings on their chest CT scans. Chest CT scanning has been proposed to be the most important staging diagnostic technique in all patients presenting with head and neck SCC,Reference Warner and Cox11, Reference Ong, Kerawala, Martin and Stafford12 and also in patients with three or more lymph node metastases, bilateral lymph node metastases, lymph nodes of 6 cm or larger, low jugular lymph node metastases, locoregional tumour recurrence, or second primary tumours.Reference de Bree, Deurloo, Snow and Leemans5 In a study of 25 patients, chest CT detected two further suspicious, false positive lesions missed on chest radiograph, whereas in 20 patients neither chest X-ray nor CT showed any evidence of pulmonary metastases.Reference Tan, Greener, Seikaly, Rassekh and Calhoun13 Nilssen et al. reported two patients (of 57) with synchronous tumours with SCC on chest CT.Reference Nilssen, Murthy, McClymont and Denholm14 In both cases, the lesions were identified on chest X-ray prior to scanning. Another study reported that staging CT of the thorax had a very low yield in 44 nasopharyngeal carcinoma patients with neck metastases.Reference Leung, Cheung, Teo and Lam15
In the current study, bone scan screening revealed hitherto undetected metastases in two patients. The two with bone metastases were also identified on neck and chest CT scans, and no one presented lung or liver metastases. In contrast to our findings, de Bree et al. found lung metastases in all four of their patients with bone metastases detected by bone scanning.Reference de Bree, Deurloo, Snow and Leemans5 Because bone metastases are most likely to be found in the vertebrae and ribs, performing chest CT scanning as screening may detect distant metastases at these sites as well. Routine bone scanning for primary staging of head and neck SCC is not recommended,Reference Jäckel and Rausch6, Reference Ampil, Wood, Chin, Hoasjoe, Aarstad and Hilton16–Reference Wolfe, Rowe and Lowry21 and should be reserved for patients with advanced primary tumours, with regional node metastases or with clinical or laboratory evidence of bone involvement.Reference Belson, Lehman, Chobanian and Malin22 In the event of an elevated alkaline phosphatase level, which has a high specificity (98 per cent) but low sensitivity (20 per cent) for the detection of distant bone metastases, a bone scan is recommended.Reference Wolfe, Rowe and Lowry21, Reference Troell and Terris23 De Bree et al. did not find any significant differences in the biochemical test results of patients with and without bone metastases.Reference de Bree, Deurloo, Snow and Leemans5
• At the first presentation of even advanced locoregional head and neck squamous carcinoma, distant metastases are rare
• Additional CT of thorax in addition to chest x-ray appears to be the only relevant investigation for the initial staging of patients with advanced head and neck squamous cell carcinoma
• Other diagnostic techniques such as abdominal ultrasonography and isotope bone scanning do not provide any further useful information
The liver is the third most common site of distant metastases in patients with head and neck SCC; however, liver metastases occur rarely in the absence of other distant metastases, particularly lung metastases. Routine use of ultrasonography or CT and MRI in the absence of other distant metastases is generally not recommended;Reference de Bree, Deurloo, Snow and Leemans5, Reference Jäckel and Rausch6, Reference Nilssen, Murthy, McClymont and Denholm14, Reference Dost, Schrader and Talanow19, Reference Belson, Lehman, Chobanian and Malin22, Reference Wernecke, Rummeny, Bongartz, Vassallo, Kivelitz and Wiesmann24 however, some findings support examination of the liver in cases with elevated liver function test results.Reference Wolfe, Rowe and Lowry21, Reference Troell and Terris23 De Bree et al. found no significant difference between the liver function test results of patients with and without liver metastases.Reference de Bree, Deurloo, Snow and Leemans5 In our study, only one patient had liver metastases; this patient's liver function test results were normal at initial staging, and the lung was affected as well.
The incidence of distant metastases in patients presenting with advanced head and neck SCC has been reported as 1–17 per cent (Table II). An advanced tumour (T) stage and/or N stage and a high histological grade are associated with the occurrence of distant metastases.Reference Calhoun, Fulmer, Weiss and Hokanson1, Reference Garavello, Ciardo, Spreafico and Gaini25 The incidence of distant metastases is also influenced by the location of the primary tumour. Primary tumours of advanced T stage in the hypopharynx, oropharynx and oral cavity are associated with the highest incidence of distant metastases.Reference Ferlito, Shaha, Silver, Rinaldo and Mondin8 Roland found different types of gradings in his selected patients. Less differentiated tumors seem to metastasize more than well differentiated tumours.Reference Jäckel and Rausch6 In a 1992 series of 2007 patients with histologically proven, graded head and neck SCC, distant metastases were found initially in 3.4 per cent of patients with poor differentiated tumours, compared with 1.8 per cent with well differentiated tumours.Reference Roland, Caslin, Nash and Stell26 Patients with four or more clinical neck lymph node metastases or low jugular lymph node metastases had the highest incidence of distant metastases.Reference de Bree, Deurloo, Snow and Leemans5
Table II Incidence of distant metastases at presentation: published studies
SCC = squamous cell carcinoma; ACC = adenoid cystic carcinoma; RMS = rhabdomyosarcoma; MM = malignant melanoma
Conclusion
At the first presentation of even advanced locoregional head and neck SCC, distant metastases were rare. An additional CT of thorax in addition to chest x-ray seems to be the only relevant investigation for the initial staging procedure in patients with advanced head and neck SCC. Other diagnostic techniques, such as abdominal ultrasonography and bone scanning, do not provide any further relevant information.
