A RATIONAL APPROACH TO PULMONARY SCREENING IN NEWLY DIAGNOSED HEAD AND NECK CANCER

Transcription

1 A RATIONAL APPROACH TO PULMONARY SCREENING IN NEWLY DIAGNOSED HEAD AND NECK CANCER Kwok Seng Loh, FRCS,* Dale H. Brown, FRCS(C), James T. Baker, FRACS, Ralph W. Gilbert, FRCS(C), Patrick J. Gullane, FRCS(C), Jonathan C. Irish, FRCS(C) Department of Otolaryngology, University Health Network, Wharton Head & Neck Center, Princess Margaret Hospital, 610 University Avenue Canada 3-952, Toronto, Ontario M5G 2M9, Canada Accepted 14 April 2005 Published online 31 August 2005 in Wiley InterScience ( DOI: /hed Abstract: Background. The purpose of this study was to determine the detection rate of lung metastasis or a synchronous lung primary tumor in patients with newly diagnosed head and neck mucosal squamous cell carcinoma (SCC) and to determine factors that are associated with positive findings. Methods. This was a prospective cohort study of 102 patients with head and neck mucosal SCC diagnosed in a tertiary cancer center. Chest x-rays and a CT scan of the thorax were performed. An indeterminate nodule on CT scan was followed with either a repeat scan to assess progression or a CT-guided needle biopsy. Metastasis or synchronous lung primary tumor were determined by CT scan. The findings were correlated with age, sex, duration of symptoms, site of primary tumor, grade of tumor, T classification, and N classification. Results. A CT scan of the thorax showed abnormalities or suspicious nodules in 20 patients (19.6%). With either follow-up scans or CT-guided biopsy, 10 patients were eventually proven to have pulmonary metastasis and one a synchronous lung primary tumor. Of those eleven patients (10.8%), seven had normal chest x-ray. Eight (72.7%) of 11 patients with a positive CT scan had N2 or N3 disease in contrast to 32 (35.2%) of 91 patients with a normal CT scan (p =.02). Seven patients (63.6%) Correspondence to: K. S. Loh *Current address: Department of Otolaryngology, National University of Singapore, National University Hospital, 5 Lower Kent Ridge Road, Singapore entv5@nus.edu.sg. B 2005 Wiley Periodicals, Inc. with a positive CT scan had T4 disease, whereas 34 (37.4%) with a normal CT scan had T4 disease (p =.08). Primary tumors arising in the oropharynx, hypopharynx, and supraglottis had a greater risk of a positive CT scan than tumors arising in the oral cavity, glottis, or unknown sites (OR = 5.4; 95% CI, ). Age, sex, duration of symptoms, and grade of disease did not predict a positive CT scan. Conclusions. The detection rate of lung metastasis or a synchronous lung primary tumor by CT scan is 10.8%. We recommend the use of CT scans of the thorax in screening the lungs of newly diagnosed patients with T4 and/or N2 or N3 oropharyngeal, hypopharyngeal, and supraglottic SCC. A 2005 Wiley Periodicals, Inc. Head Neck 27: , 2005 Keywords: chest x-ray; CT scan; distant, metastasis; thorax In the management of head and neck mucosal squamous cell carcinoma (SCC), it has been traditional to perform a metastatic workup at initial presentation. Over the years, data in the literature suggest that the most common site of distant metastasis at the time of diagnosis in head and neck SCC is the lungs. 1,2 Furthermore, it has been reported that approximately 1% of head and neck SCCs are also associated with synchronous lung cancer. 3,4 Most centers perform screening for lung metastasis only. 990 CT Thorax in Head and Neck Cancer

2 The screening technique for the lungs is usually a chest x-ray. Indeed, the recommendation of the National Comprehensive Cancer Center Network 5 for screening of distant metastasis in head and neck mucosal SCC is a chest x-ray. This recommendation is likely based on the issue of cost-effectiveness. However, the chest x-ray has a low sensitivity in detecting the small percentage of patients with lung metastasis or synchronous lung primary tumor, and its effectiveness is questionable. The alternative imaging technique is CT. CT of the thorax is undoubtedly superior to chest x-ray in detecting metastasis and synchronous lung tumors. However, it is a much more expensive procedure, and concerns have been raised about its cost-effectiveness 6 as a screening tool. The aims of our study were to determine the rate of detection of metastasis and synchronous lung tumors using chest x-ray and CT of the thorax and to correlate clinical parameters that are associated with positive findings. By doing so, we will then be able to rationalize the approach to detecting lung metastasis and synchronous lung tumors in newly diagnosed patients with head and neck SCC. PATIENTS AND METHODS This was a prospective cohort study of 102 patients with newly diagnosed head and neck mucosal SCC seen at the Princess Margaret Hospital over a 12-month period. Patients were included if they had primary tumors arising in the oral cavity, oropharynx, hypopharynx, and larynx. The patients initially seen with an unknown primary site with metastatic nodal disease were included. However, primary tumors originating from the nasopharynx, paranasal sinuses, parotid gland, and skin were excluded. Patients with recurrent head and neck SCC were excluded. Informed consent was obtained. All patients had a detailed history and physical examination followed by chest x-ray and subsequently CT of the thorax. These imaging investigations were performed within 4 weeks of diagnosis at our center. Patients who already had a CT of the thorax performed at the referring center were excluded to maintain uniformity of the imaging protocols and also not to subject the patient to two similar procedures. The chest x-ray was performed in an anteroposterior view. The CT of the thorax was performed using 5-mm axial sections without contrast. Staff radiologists analyzed the findings of the chest x-ray and CT of the thorax. These were reported and categorized as one of the following: normal with no evidence of metastasis, abnormal with evidence of metastasis/primary lung tumor, or indeterminate. As with the protocol at our center, the indeterminate nodules on CT of the thorax were either followed up with a repeat CT in 3 to 6 months to assess progression, or a CT-guided needle biopsy was performed. If the CT-guided biopsy was positive for malignancy or if progression was demonstrated on the follow-up CT of the thorax, the patient was determined to have metastatic disease or synchronous lung primary tumor, whichever the case may be. The CT of the thorax of these patients was termed as positive. If the follow-up CT of the thorax did not show any progression, and the CT-guided biopsy was normal, the patient was determined to have no metastasis or synchronous lung primary tumor. The CT of the thorax of these patients was termed as negative. All patients with a normal CT of the thorax were also categorized as having negative. Demographic data included age at diagnosis and the sex of the patients. The duration of symptoms was taken to be the time in months from the first symptom to the time of diagnosis. Grade of tumor was well-differentiated, moderately differentiated, or poorly differentiated SCC. The sites of primary tumor were grouped into oral cavity, oropharynx, hypopharynx, larynx, and unknown. For the group with laryngeal primary tumors, this was further divided into those with supraglottic and glottic tumors. T and N classifications were according to the Union Internationale Contre le Cancer (UICC)/American Joint Committee on Cancer (AJCC) 1997 staging system. Statistical data were analyzed with the SPSS program (version 11.0). Categorical data were analyzed by chi-square or the Fischer exact test where appropriate. RESULTS A total of 102 patients were included in this study, 80 men (78.4%) and 22 women (21.6%). The mean age was 62.9 years (range, of years). The mean duration of symptoms before diagnosis was 5.2 months. There were 46 oral cavity primary tumors (45.1%), 20 oropharynx (19.6%), nine hypopharynx (8.8%), 18 larynx (17.6%), and nine unknown primary sites (8.8%). Of the tumors in the 18 patients with tumors in the larynx, CT Thorax in Head and Neck Cancer 991

3 Table 1. Correlation of age, sex, duration of symptoms, and grade of tumor with. Patient and tumor characteristics positive negative (n = 91) p value Mean age, y Sex, no. of patients, 10:1 3.3:1.45 male:female Mean duration of symptoms, mo Grade of tumor,.26 no. of patients (%) Well/moderately 7 (63.6%) 72 (79.1%) differentiated Poorly differentiated 4 (36.4%) 19 (20.9%) Table 3. Correlation of site of primary tumor with positive. Site of primary tumor No. of patients (%) Positive Negative (n = 91) p value Unknown 2 (18.2%) 7 (7.7%) Oral cavity 1 (9.1%) 45 (49.5%) Larynx glottis 0 9 (9.9%) Larynx supraglottis 2 (18.2%) 7 (7.7%) Oropharynx 4 (36.4%) 16 (17.6%) Hypopharynx 2 (18.2%) 7 (7.7%) Site by grouping.02 Unknown, oral, glottis 3 (27.3%) 61 (67%) Supraglottis, oropharynx, or hypopharynx 8 (72.7%) 30 (33%) nine were in the supraglottis. Fourteen patients (13.7%) had a T1 primary tumor, 18 (17.6%) T2, 20 (19.6%) T3, 41 (40.2%) T4, and 10 (9.8%) unknown. Forty-six patients (45.1%) had N0 disease, 16 (15.7%) N1, 34 (33.3%) N2, and six (5.9%) N3. A total of 20 patients (19.3%) had abnormal or indeterminate findings on CT of the thorax; of these, 11 patients (10.8%) were eventually determined to have metastasis or a lung primary tumor using either a follow-up scan or CT-guided biopsy. The rate of detection of metastasis in the lungs or a synchronous lung primary tumor using CT of the thorax was, therefore, 10.8%. Ten of these 11 patients had lung metastasis, whereas one had a synchronous lung lesion. Of these 11 patients with positive CT of the thorax, seven (63.6%) had normal chest x-ray. Of the nine patients with abnormal chest x-ray, only four (44.4%) had a positive CT scan of the thorax. Age, sex, duration of symptoms before diagnosis, and grade of tumor were not associated with a positive CT of the thorax (Table 1). Of Table 2. Correlation of T and N classification with. Classification No. of patients (%) Positive Negative (n = 91) p value T classification.08 T1, T2, T3, or unknown 4 (36.4%) 57 (62.6%) T4 7 (63.6%) 34 (37.4%) N classification.02 N0, N1 3 (27.3%) 59 (64.8%) N2, N3 8 (72.7%) 32 (35.2%) those with a positive CT of the thorax, 63.6% had T4 tumors, in contrast to 37.4% of those with a negative CT of the thorax (Table 2). This approached statistical significance ( p =.08). There were 72.7% of patients with a positive CT of the thorax with N2 or N3 disease in contrast to only 35.2% of those with a negative CT of the thorax (OR, 2.1; 95% CI, ). Table 3 shows a significantly greater proportion of tumors (72.7% vs 33%) arising in the oropharynx, hypopharynx, and supraglottic sites in patients with a positive CT of the thorax than in those with a negative CT of the thorax (OR, 5.4; 95% CI, ). DISCUSSION The effectiveness of CT of the thorax over chest x-ray in screening for metastasis or concurrent lung primary tumors is not in question. CT of the thorax is able to detect smaller lesions and certainly gives much better visualization of the lungs than does chest x-ray. The downside to the use of CT in screening the chest is cost and logistics. The cost-effectiveness of using CT of the thorax to screen the lungs in patients with newly diagnosed head and neck SCC has been called into question. Tan et al 6 questioned the clinical benefit of using CT to screen the chest, citing cost/ benefit as the main reason against its use. They advocate the use of chest x-ray. The guidelines of the National Comprehensive Cancer Network also recommend the use of chest x-ray to screen for lung metastasis. We do not disagree that performing CT of the thorax for all patients with head and neck SCC will add a significant burden to any health care system. However, a much more rational approach could be developed to identify 992 CT Thorax in Head and Neck Cancer

4 patients at risk for having metastasis and concurrent lung primary tumors develop, so that these patients can be targeted for CT of the thorax. The preceding studies do not address this issue. We are fully aware that certain clinical factors are associated with higher risks of distant metastasis developing subsequent to the diagnosis. These include T and N classifications and grade of tumor. Our hypothesis is that metastasis at diagnosis and concurrent lung primary tumors are associated with clinical variables. The aim was, therefore, to determine which clinical variables were associated with a positive CT of the thorax. We performed a prospective study to avoid the inherent bias of retrospective studies, from which most of the data in the literature were based. Our study showed that the rate of detection of lung metastasis or a synchronous lung primary tumor in patients with newly diagnosed head and neck SCC was 10.8%. We had found in our study that 19.6% of the CTs of the thorax were either abnormal or suspicious of metastasis. This, at first glance, may be similar to the data in the literature, which suggest that this rate ranges from 16% to 19% However, we need to be cautious in the interpretation of the detection rates using CT of the thorax in these studies. In these studies, the figures stated were for all abnormal CTs of the thorax, and no attempt had been made to ascertain those that may have been indeterminate or suspicious. To determine the nature of these indeterminate or suspicious CTs of the thorax, we embarked on a policy of CTguided needle biopsy or follow-up scans. Indeed, we found that nine of these suspicious CT scans did not have malignancy on CT-guided biopsy or progression on follow-up scans. Therefore, although there was a rate of 19.6% abnormal or suspicious CTs of the thorax, the actual rate of clinical lung metastasis or a synchronous lung primary tumor was lower, at 10.8%. Ideally, it would be even more accurate if we had open biopsies of all abnormalities seen on the CT of the thorax, but this was obviously not practical and certainly not ethical. Hence, the next best solution was either a follow-up scan or, when it was possible and safe, a CT-guided needle biopsy. Our data, therefore, indicate that using CT of the thorax to screen the lungs in patients with newly diagnosed head and neck SCC, approximately 20% will show an abnormality but only 10.8% will actually be truly representative of lung metastasis or a synchronous lung primary tumor. The rate of 10.8% of detecting lung metastasis or a synchronous lung primary tumor in a cohort of patients with newly diagnosed head and neck SCC is significantly high. This would indicate that patients with newly diagnosed head and neck SCC should be screened for lung metastasis and a synchronous lung primary tumor. Interestingly, chest x-ray was able to detect only 36.4% of these abnormal CTs of the thoraxes. Our study illustrates the poor reliability of chest x-ray as a screening tool for the lungs compared with CT of the thorax. Other screening methods for the lung include bronchoscopy, bronchial washings, and PET scans. Bronchoscopy and bronchial washings are both invasive methods, and their routine use in all cases of head and neck SCC has not been recommended. 12 Positron emission tomography [PET] scans have been reported to be effective in detecting lung metastasis or synchronous lung primary tumors, 13 but, at least for now, it remains even more costly than CT and not universally available. On the other hand, Keyes et al 14 reported that using (F-18) fluorodeoxy-d-glucose (FDG)-PET gave a case-finding yield of only 2%. In their study, FDG-PET was not superior to other imaging modalities for detecting lesions in the lungs in patients with newly diagnosed head and neck cancer. Hence, the overall costeffectiveness remains unanswered. Taking these factors into consideration, we believe that CT of the thorax should be the method of choice in screening the lungs in patients with newly diagnosed head and neck SCC. The clinical parameters associated with a positive CT of the thorax were N2/N3 disease and primary tumor originating from the oropharynx, hypopharynx, or supraglottis. Almost 73% of patients with a positive CT of the thorax had N2 or N3 disease in contrast to 35% of those with a negative CT of the thorax. This was statistically significant, and the risk of a positive CT of the thorax in those with N2/N3 disease was 2.1 times more than in those with N0/N1 disease. Our data also showed that primary tumors arising in the oropharynx, hypopharynx, and supraglottis were five times more likely than those originating from the glottis and oral cavity to have a positive CT of the thorax. The implication here is that any head and neck mucosal SCC arising from sites in the oropharynx, hypopharynx, and supraglottis or patients with N2/N3 disease should have a CT of the thorax to screen the lungs. It was noteworthy that T classification did not correlate with a positive CT of the thorax. We did, however, CT Thorax in Head and Neck Cancer 993

5 appreciate that the correlation between a positive CT of the thorax and T4 disease approached statistical significance. Other clinical parameters such as age, sex, duration of symptoms before diagnosis, and grade of disease did not correlate significantly with a positive CT of the thorax. Ong et al 8 reported that nodal disease and primary site were not associated with a positive CT of the thorax and recommended that CT of the thorax be performed in all patients with head and neck cancers. Our study provides evidence that we could avoid performing CT of the thorax in those with N0/N1 disease and when the tumors did not arise in the oropharynx, hypopharynx, and supraglottis. There are drawbacks in this study. Ideally, we would like to have pathologic proof of metastasis in all our patients with a positive CT of the thorax. In some of these patients, the lesion may not be in a location to have a safe biopsy, and the next best solution is to monitor the lesion to confirm it is enlarging. This notwithstanding, our study demonstrates that the use of CT of the thorax in screening the lungs of patients with head and neck SCC can be rationalized. We showed that patients with primary tumors originating in the oropharynx, hypopharynx, or supraglottis and all patients with N2 N3 disease should be screened for metastasis in the chest or a concurrent lung primary tumor. On the basis of our data, we also recommend that patients with T4 disease should have a CT of the thorax. Other patients have a much lower risk, and we recommend that there is no need to screen them or at most a chest x-ray will suffice. By following this set of guidelines, we believe that the use of CT of the thorax will be more cost-effective. CONCLUSION By use of CT of the thorax as the screening modality, the rate of detection of lung metastasis or synchronous lung primary tumors in patients with newly diagnosed head and neck SCC is 10.8%. Patients with head and neck SCC with primary tumor sites in the oropharynx, hypopharynx, or supraglottis and those with N2/N3 disease should have a CT of the thorax to screen the lungs. REFERENCES 1. Probert JC, Thompson RW, Bagshaw MA. Patterns of spread of distant metastasis in head and neck cancer. Cancer 1974;33: Zbaren P, Lehmann W. Frequency and sites of distant metastasis in head and neck squamous cell carcinoma. Arch Otolaryngol Head Neck Surg 1987;113: Kuriakose MA, Loree TR, Rubenfeld A, et al. Simultaneously presenting head and neck and lung cancer: a diagnostic and treatment dilemma. Laryngoscope 2002; 112: Erkal HS, Mendenhall WM, Amdur RJ, Villaret DB, Stringer SP. Synchronous and metachronous carcinomas of the head and neck mucosal sites. J Clin Oncol 2001; 19: Pfister DG, Ang K, Brockstein B, et al, for the National Comprehensive Cancer Network. NCCN practice guidelines for head and neck cancer. Oncology (Huntingt) 2000; 14: Tan LKS, Greener CC, Seikaly H, Rassekh CH, Calhoun KH. Role of screening chest computer tomography in patients with advanced head and neck cancer. Otolaryngol Head Neck Surg 1999;120: Mercader VP, Gatenby RA, Mohr RM, Fisher MS, Caroline DF. CT surveillance of the thorax in patients with squamous cell carcinoma of the head and neck: a preliminary experience. J Comput Assist Tomogr 1997;21: Ong TK, Kerawal CJ, Martin IC, Stafford FW. The role of thorax imaging in staging head and neck squamous cell carcinoma. J Craniomaxillafac Surg 1999;27: de Bree R, Deurloo EE, Snow GB, Leemans CR. Screening for distant metastases in patients with head and neck cancer. Laryngoscope 2000;110: Reiner B, Siegel E, Sawyer R, Brocato RM, Maroney M, Hooper F. The impact of routine CT of the chest on the diagnosis and management of newly diagnosed squamous cell carcinoma of the head and neck. AJR Am J Roentgenol 1997;169: Houghton DJ, Hughes ML, Gervey C, et al. Role of chest CT scanning in the management of patients presenting with head and neck cancer. Head Neck 1998;20: Benninger MS, Enrique RR, Nichols RD. Symptomdirected selective endoscopy and cost containment for evaluation of head and neck cancer. Head Neck 1993;15: Teknos TN, Rosenthal EL, Lee D, Taylor R, Marn CS. Positron emission tomography in the evaluation of stage III and IV head and neck cancer. Head Neck 2001; 23: Keyes JW Jr, Chen MTM, Watson NE, Greven KM, McGuirt WF, Williams DW III. FDG PET evaluation of head and neck cancer: value of imaging the thorax. Head Neck 2000;22: CT Thorax in Head and Neck Cancer