Metastases from Unknown Primary Tumor

PII S1095-0397(00)00053-4

Clinical Positron Imaging Vol. 3, No. 4, 137–141. 2000 Copyright  2000 Elsevier Science Inc. Printed in the USA. All rights reserved. 1095-0397/00 $–see front matter

ORIGINAL ARTICLE

Metastases from Unknown Primary Tumor: PET-FDG as Initial Diagnostic Procedure? Max Lonneux, MD, Abdel-Malek Reffad, MD Center of Nuclear Medicine, Cliniques Universitaires Saint-Luc, Brussels, Belgium. Purpose: To analyze the efficacy and impact on management of PET-FDG in patients with metastases from unknown primary tumor. Procedures: Retrospective analysis of 24 patients referred to the PET center for metastasis of unknown primary after a negative imaging workup. PET results were validated by means of oriented imaging, follow-up or biopsy when ethically justified. Results: PET identified the primary tumor in 13/24 (54%) of patients: breast (n ⫽ 1), lung (n ⫽ 9), colon (n ⫽ 1), stomach (n ⫽ 1) and mouth (n ⫽ 1). The false positive rate of PET was 21% (5/24). PET was shown to affect the management of 10/24 patients (42%). Conclusion: Whole body PET-FDG was more effective than conventional imaging methods in detecting unknown primary tumors. PET altered patient management in 42% of cases. PET should be performed prior to other investigations in such patients and could avoid unnecessary and often unfruitful diagnostic procedures. (Clin Pos Imag 2000;3:137–141)  2000 Elsevier Science Inc. All rights reserved. Key Words: FDG; Unknown primary tumor; Tumor staging.

Introduction

C

ancer of unknown primary origin represents 5–10% of all cancer patients and constitutes a major diagnostic challenge for the clinician. Development of immunohistochemical techniques allows the physician to clarify tumor subtypes so that a more specific treatment can be proposed, in up to 40% of cases.1 However, the identification of the primary tumor is of utmost importance for choosing the most appropriate therapeutic strategy. In the presence of a metastasis of unknown primary origin, conventional imaging procedures are poorly effective in identifying the primary tumor.2 Numerous unsuccessful radiological procedures are costly and stress-inducing for the patient. Positron emission tomography (PET) using the glucose analogue 18F-fluorodeoxyglucose has been widely used for imaging tumors.3 Recent reports have illustrated its use for the identification of unknown primary tumors.4–6 PET yields information over the entire body in a single imaging session, which is potentially more cost-effective than a “hit-or-miss” multi-device radiological workup. Usually the metastatic disease is detected because of tumor-related symptoms—such as pain—or clinical signs such as Address correspondence to: Max Lonneux, MD, Service de Me´decine Nucle´aire, Cliniques Universitaires Saint-Luc, Avenue Hippocrate 10, B-1200 Brussels, Belgium. Phone: ⫹32-2-7642580 Fax: ⫹32-2-7645408. E-mail: [email protected]

lymph node or liver enlargement. The initial assessment is limited to imaging and tissue sampling of the metastatic site. As far as the primary tumor is concerned, the most judicious place of PET within the spectrum of the imaging workup for such patients is not yet clear, since it is generally performed after a thorough negative diagnostic workup. Its high sensitivity, however, could indicate that it should be used as a first line modality. In the present study, we reviewed the medical history of patients referred to the PET center because of metastases from unknown primary tumor. Our aim was to investigate whether PET could be successful as a first-line imaging modality in such patients. We further analyzed the impact of PET results on the management of patients.

Material and Methods Patient Population We retrospectively reviewed the medical files of 24 patients addressed to our PET center for metastatic disease of unknown primary tumor. There were nine women and 15 men (mean age 59 ⫾ 10 y, range from 35 to 79 y). Metastases were located in brain (n ⫽ 7), bone (n ⫽ 2), liver (n ⫽ 3), pleura (n ⫽ 3), axillary nodes (n ⫽ 2), cervical nodes (n ⫽ 3), muscle (n ⫽ 2), eye (n ⫽ 1), meninges (n ⫽ 1). Prior to inclusion, all patients had undergone a thorough physical exami137

138 Clinical Positron Imaging, Volume 3, Number 4

nation, blood chemistries, chest X-ray and liver ultrasound. Depending on the metastatic site, additional imaging procedures had been performed prior to PET in selected patients, including breast ultrasound and mammography, brain CT or MRI, abdominal and thoracic CT, cervical ultrasound or CT, and head and neck panendoscopy. Due to the retrospective nature of our analysis, there were no specific requirements in terms of imaging procedures so that patients could enter the study population after a minimal (chest X-ray and liver ultrasound) or extensive imaging workup.

PET Imaging Procedure Patients were fasted for at least six hours prior to their arrival in the PET facility. Sixty minutes after iv injection of 370 MBq 2-[fluorine-18]fluoro-2-deoxy-d-glucose (FDG), patients were positioned on the PET camera (ECAT EXACT HR, CTI, Knoxville, USA) and whole body emission scan was obtained followed by a transmission scan for subsequent attenuation correction. Total scanning time was 60 min. No bladder catheterization was used. Images were reconstructed by means of iterative processing of both emission and transmission data, following the procedure previously described.7 Images were interpreted on color monitors with simultaneous display of non attenuated and attenuated images in the transaxial, coronal and sagittal planes.

Data Analysis Verification of FDG positive sites was made by appropriate imaging procedures or biopsy when ethically justified, i.e. when a therapeutic or palliative benefit was expected from histologic confirmation (n ⫽ 13). When PET was negative, a follow-up period of at least six months was required before classifying the result as false or true negative.

Results Patient characteristics and PET results are listed in Table 1. Metastatic lesions had been surgically removed in five patients before PET (ocular metastasis, cervical node and brain in three patients). All other known tumoral sites were visible on PET images, although three brain metastases appeared as cold areas surrounded by high FDG uptake within the cerebral cortex. FDG hot spots consistent with a primary tumor were identified in 18/24 patients. In 13 cases, follow-up imaging or biopsy confirmed that the primary tumor was indeed located in breast (n ⫽ 1), lung (n ⫽ 9), colon (n ⫽ 1), stomach (n ⫽ 1) or mouth (n ⫽ 1). In patient

no. 1 presenting with a left axillary mass consistent with an atypical lobular carcinoma, PET showed moderate uptake in the retroareolar region corresponding to the primary. This lesion had not been detected by ultrasound nor mammography (Figure 1). Figure 2 illustrates a patient (no. 13) with an enlarged cervical lymph node corresponding to squamous cell carcinoma metastasis. The primary was not identified by clinical and imaging workup including endoscopy of the oropharyngeal region. PET showed a focus of high FDG uptake in the right anterior part of the oral cavity. Deep biopsies were guided by PET results, and histological analysis was positive for a squamous cell carcinoma. In nine patients, the primary tumor was located in the lungs. CT of the chest had not been performed prior to PET in two patients for whom chest X-ray was negative. In three patients, CT results were reported as “non conclusive.” In four patients, CT of the chest was negative: in two patients, the tumor was located in the lung apex and was not visible even at re-reading with the PET image available. In one patient, the lesion located in the left upper lobe became visible on CT four months after the positive PET. In the last case, the retrocardiac 1 cm tumor was indeed visible on the chest CT but was overlooked at initial reading. Interestingly, five out of seven brain metastases originated from lung primaries that were all positive on PET-FDG. In one case, PET showed multiple mediastinal lymph nodes and an adrenal metastasis but did not identify the primary tumor. Those tumor sites as well as the histologic subtype of the brain lesion were compatible with a pulmonary origin and the patient was treated as metastatic lung adenocarcinoma. PET found additional unknown metastatic sites in seven patients: bone (n ⫽ 2), lymph nodes (n ⫽ 4) and adrenal (n ⫽ 1). In five patients, the maxillary sinus, rectum, colon (n ⫽ 2) and larynx were suspected of harboring the primary tumor on the basis of PET but this was not confirmed by other modalities or subsequent follow-up. Overall the sensitivity of PET for the detection of the primary tumor was 54% (13/24). There was no false negative PET study, i.e., no patient with a negative PET for whom another imaging modality or follow-up identified the primary tumor. PET affected the therapeutic choices in 10 patients. In seven of them, the identification of the primary led to modification of the chemotherapeutic regimen. In one case, the pathologic examination of the metastasis was in favor of a mammary carcinoma. PET depicted a lung nodule that was punctured and shown to display the same histological characteristics as the metastasis. As a consequence, the chemotherapeutic regimen was modified. In one patient, PET identified the primary tumor in the anterior part of the mouth. This tumor was surgically removed and the irradiation fields were

PET-FDG to Diagnose Matastases from Unknown Tumor / Lonneux and Reffad 139

Table 1. Characteristics of patients with metastases from unknown primary tumor (n ⫽ 24) Patient

Localization of metastasis

Histology*

PET suggested primary site

Confirmation

PET status

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

axillary lymph node cerebellum eye cervical lymph node axillary lymph node bone brain brain brain brain pleura bone cervical node liver meninges muscle abdominal wall brain liver cervical and supraclavicular lymph nodes liver pleura brain pleura

ADC ADC ADC SCC ADC ADC ADC ADC ADC ADC ADC ADC SCC ADC PDC ADC ADC ADC unknown unknown ADC ADC ADC PDC

breast none lung maxillary sinus rectum lung lung lung lung none none lung mouth transverse colon none lung caecum, peritoneum lung right colon none stomach larynx lung colon

surgery no primary found biopsy no primary found no primary found biopsy biopsy biopsy biopsy no primary found no primary found biopsy biopsy no primary found no primary found biopsy no primary found biopsy coloscopy, biopsy no primary found biopsy no primary found biopsy no primary found

TP TN TP FP FP TP TP TP TP TN TN TP TP FP TN TP FP TP TP TN TP FP TP FP

* ADC ⫽ adenocarcinoma; SCC ⫽ squamous–cell carcinoma; PDC ⫽ poorly differentiated carcinoma.

adjusted to include the site of primary as well as the metastatic lymph nodes. In one patient with liver metastases, PET showed the primary located in the right colon, which was surgically removed. In the three remaining patients, PET results, although accurate, did not significantly alter the treatment.

Discussion Our data show that whole-body PET-FDG is effective in identifying the primary tumor in patients with metastatic disease and previous non contributive workup. In our series, PET identified the primary tumor in 54% (13/24) of cases. Moreover, there was no false negative

Figure 1. Non attenuated PET-FDG image obtained in a woman presenting with a left axillary mass corresponding to metastatic lymph nodes (Left: coronal. Center: sagittal. Right: transaxial). PET showed the nodal metastases (open arrow) as well as a small lesion located in the retroareolar region of the left breast (arrow) that corresponded to a lobular carcinoma.

140 Clinical Positron Imaging, Volume 3, Number 4

Figure 2. High FDG uptake in a right cervical adenopathy corresponding to squamous cell carcinoma metastasis (open arrow). PET also showed a hot spot located in the base of mouth (arrow). Deep biopsies were performed in the PET positive region and confirmed the presence of a squamous cell carcinoma that was subsequently removed and irradiated.

PET, that is no patient for whom PET did not detect a primary tumor that was discovered at a later stage by other procedures. The main advantage of PET is that it yields information over the entire body. If performed early in their workup, PET could avoid numerous diagnostic explorations of patients who often have a poor prognosis. Indeed, PET has been reported more sensitive for detecting tumors than conventional imaging procedures, and our data as well as reports from other groups support the statement that PET should be performed as first line procedure.4–6 In our study, the false positive rate of PET was 21% (5/24). Three out of five false positive findings were observed in colon and rectum. Physiological colonic uptake can be high and heterogeneous and mimic a tumoral lesion.8 It has been reported that this uptake could be reduced by appropriate colon preparation.9 This is not used in our center because of the discomfort caused to the patient by a procedure that has not been shown to significantly increase the diagnostic value of PET. In the two remaining cases, moderate uptake was observed in the maxillary sinus and vocal chord respectively. Careful examination of these regions was negative. False positive PET findings in the head and neck region is a common problem since benign inflammation or structural asymetry can be misleading. Semiquantitative indexes do not necesseraly help because inflammation may concentrate FDG at the same level as tumor.8 In patients with unknown disease PET images are cautiously interpreted; nevertheless, we think that physiological or artificial uptake in the digestive tract or head and neck region make images difficult to interpret, and reduce the specificity of PET-FDG. We noticed that PET affected the management of 10/24 patients (42%). Changes in chemotherapeutic regimens were the most frequently observed. This is of importance since toxicity of chemotherapy is not

negligible. Identification of the primary tumor allows patients and providers to choose the most effective and cost-effective treatment for each patient. Incidentally, the knowledge of the primary can also lead to enrollment of patients in appropriate therapeutic trials. Besides the aforementioned effect on therapy, knowing the primary tumor adds prognosis information, which can be essential for the patient. No conclusion can be drawn from this retrospective study about the cost-effectiveness of using PET as firstline imaging procedure in patients with unknown primary tumor. PET is considered an expensive technique and is not widely available. However, the cost of a whole body PET examination has to be compared to the cumulative costs of the multiple imaging and endoscopic procedures that are usually performed. Moreover, from the patient’s point of view, a one-day noninvasive procedure is usually preferable to repeated visits to the hospital, especially if the single procedure is more effective, as shown in this study. The availability of PET remains a problem since its low-cost alternatives have not proven as effective for detecting tumor lesions, especially below the diaphragm.10,11 However, the number of clinical PET centers is rapidly increasing, which should make this technology available to the great majority of patients. In conclusion, whole body PET-FDG has a superior diagnostic performance compared to combined conventional diagnostic procedures in patients with metastases from unknown primary tumor. By detecting the primary in up to half the patients with previous negative workup, PET affects the patient management in 42% of cases. Our results support the idea that PET should be performed early in the workup of such patients. PET could thereby allow patients to avoid time-consuming and often unfruitful diagnostic procedures.

PET-FDG to Diagnose Matastases from Unknown Tumor / Lonneux and Reffad 141

References 1. Hainworth, J.; Greco, F. Management of patients with cancer of unknown primary sites. Oncology-Huntington 14:563–574; 2000. 2. Nystrom, J.; Wiener, J.; Wolf, R.; Bateman, J.; Violoa, M. Identifying the primary site in metastatic cancer of unknown origin: inadequacy of roentgraphic procedures. JAMA 241:381–383; 1979. 3. Delbeke, D. Oncological applications of FDG PET imaging. J. Nucl. Med. 40:1706–1715; 1999. 4. Kole, A.C.; Nieweg, O.E.; Pruim, J.; Hoekstra, H.J.; Koops, H.S.; Roodenburg, J.L. et al. Detection of unknown occult primary tumors using positron emission tomography. Cancer 82:1160–1166; 1998. 5. Lassen, U.; Daugaard, G.; Eigtved, A.; Damgaard, K.; Friberg, L. 18F-FDG whole body positron emission tomography (PET) in patients with unknown primary tumours (UPT). Eur. J. Cancer 35:1076–1082; 1999. 6. Bohuslavizki, K.; Klutmann, S.; Kro¨ger, S.; Sonnemann, U.; Buchert, R.; Werner, J. et al. FDG PET detection of unknown primary tumors. J. Nucl. Med. 41:816–822; 2000.

7. Lonneux, M.; Borbath, I; Bol, A; Coppens, A; Sibomana, M; Bausart, R. et al. Attenuation correction in whole body FDG oncological studies: the role of statistical reconstruction. Eur. J. Nucl. Med. 6:591–598; 1999. 8. Engel, H.; Steinert, H; Buck, A; Berthold, T; Huch Bo¨ni, R; von Schultess, G. Whole-body PET: physiological and artifactual fluorodeoxyglucose accumulations. J. Nucl. Med. 37:441–446; 1996. 9. Miraldi, F.; Vesselle, H.; Faulhaber, P.; Adler, L.; Leisure, G. Elimination of artifactual accumulation of FDG in PET imaging of colorectal cancer. Clin. Nucl. Med. 23:3–7; 1998. 10. Lonneux, M.; Delval, D.; Bausart, R.; Moens, R.; Willockx, R; Van Mael, P. et al. Can dual-headed 18F-FDG SPET imaging reliably supersede PET in clinical oncology? A comparative study in lung and gastrointestinal tract cancer. Nucl. Med. Comm. 19:1047–1054; 1998. 11. Shreve, P.D.; Steventon, R.S.; Deters, E.C.; Kison, P.V.; Gross, M.D.; Wahl, R.L. Oncologic diagnosis with 2-[fluorine-18]fluoro-2-deoxy-D-glucose imaging: dual-head coincidence gamma camera versus positron emission tomographic scanner. Radiology 207:431–437; 1998.