3′-¹⁸F-Fluoro-3′-Deoxy-l-Thymidine: A New Tracer for Staging Metastatic Melanoma?

Patients

This prospective study consisted of 10 consecutive patients with clinical stage III melanoma (locoregional disease). Patients were included from April until November 2002.

Two patients had a unknown primary and 2 patients had a primary melanoma, which was too small to assess the Clark level. All patients gave written informed consent. For inclusion, the liver and kidney functions and hematologic parameters (hemoglobin, hematocrit, erythrocytes, thrombocytes, leukocytes, and white cell count) had to be within normal limits. Pregnant patients and patients with psychiatric disorders were excluded. All screened patients could be included in the study. The Medical Ethics Committee of Groningen University Hospital approved the study protocol.

PET Studies

Synthesis of FLT was performed according to the method of Machulla et al. (19) ¹⁸F-FLT was produced by ¹⁸F-fluorination of the 4,4′-dimethoxytrityl–protected anhydrothymidine, followed by a deprotection step. After purification by reversed-phase high-performance liquid chromatography, the product was made isotonic and passed through a 0.22-μm filter. ¹⁸F-FLT was produced with a radiochemical purity of >95% and specific activity of >10 TBq/mmol. The radiochemical yield was 6.7% ± 3.7% (decay corrected).

Eight studies were performed using an ECAT EXACT HR+ (Siemens/CTI, Inc.) and 2 studies were performed on an ECAT 951/31 (Siemens/CTI, Inc.). Before PET imaging, patients were instructed to fast for at least 6 h to keep the study comparable with studies performed with ¹⁸F-FDG (20). They were also instructed to drink 1 L of water before imaging to stimulate ¹⁸F-FLT excretion from the renal calyces and stimulate subsequent voiding. Surgery followed ¹⁸F-FLT PET after a median period of 26 d (range, 7–45 d).

Sixty minutes after injection, a nonattenuation–corrected whole-body scan was acquired from crown to femur with 8 min per bed position. Because detection or exclusion of malignant lesions, rather than the quantitative determination of uptake, is the main goal of this feasibility study, only nonattenuation–corrected PET images were obtained. If the primary tumor was located under the level of the femur, the patient was scanned from crown to foot. PET images were iteratively reconstructed (ordered-subset expectation maximization) (21).

Pathologic Evaluation and Staging

The histology of all primary lesions and metastasis were evaluated according to the latest version of the American Joint Committee on Cancer (AJCC) (22). The emphasis of this classification is on tumor thickness, ulceration, and number of positive lymph nodes. Breslow thickness and Clark level of the primary lesion were evaluated. In all metastatic lesions (in-transit and lymph node metastases), tumor size and mitotic rate were measured. Tumor size was expressed in millimeters or as micrometastasis if <2 mm. Mitotic rate was expressed in number of mitoses per 2 mm² at ×400 magnification.

The staging took place according to the last version of the AJCC (22). All patients were first staged clinically, by physical examination, lactate dehydrogenase, chest radiography, and histopathologic confirmation of the locoregional lymph node(s). Next, the patients were staged on the basis of the ¹⁸F-FLT PET images only and finally were staged pathologically after surgery. The included patients with stage III melanoma had locoregional metastases in the lymph nodes of the groin or axilla.

Data Analysis

¹⁸F-FLT PET images were analyzed for uptake in malignant lesions and normal anatomic structures. Two experienced PET physicians evaluated the images independently and were aware of the original location of the primary lesion but unaware of other clinical information. They subsequently reached consensus on a lesion-by-lesion basis. The pathologist was unaware of the results of the PET images.

Because it was impossible to exactly match individual lesions on PET with the exact same lymph nodes as analyzed after resection or cytologic aspiration, it was decided to categorize all histopathologic and PET findings into relevant anatomic regions. The regions were defined as follows: superficial or deep groin, parailiac, obturator, popliteal, supraclavicular, axillary, mediastinal, skeleton, back, neck, arm, calf, and heel. PET and histologic data from these areas were correlated and sensitivity and specificity were calculated. Therefore, only regions with histopathologic confirmation were analyzed for accuracy.

Per anatomic region, all positive lymph nodes were measured in millimeters and the mitotic rate of the lymph node with the highest proliferation was calculated to estimate the detection level of ¹⁸F-FLT PET.

The results of clinical staging (before surgery), staging with ¹⁸F-FLT PET (before surgery), and pathologic staging (after surgery) were compared.

¹⁸F-FLT Distribution in Patients

Ten patients were included. Patient and primary melanoma characteristics are shown in Table 1. Patients received a median 400-MBq dose (range, 185–430 MBq) of ¹⁸F-FLT. Intense ¹⁸F-FLT uptake was observed in the skeleton, with a distribution pattern that is typical for bone marrow uptake (Fig. 1). The liver also showed avid uptake. Minor uptake is observed in intestinal structures. All other organs and tissues showed low-grade and homogeneous uptake. No activity was present in the brain.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

¹⁸F-FLT PET of 34-y-old man (patient 6) 2 mo after resection of primary melanoma on back. Metastases (large arrows) are observed in left and right axillae and left supraclavicular area. Physiologic uptake (small arrows) is observed in liver and 1 marrow—for example, in pelvis, vertebrae, ribs, and bony structures of shoulder; less intense, uniform tracer uptake is present in lungs. Uptake of tracer in brain is negligible, and no uptake is observed in mediastinum and myocardium.

In all patients, ≥1 abnormal lesions were found using ¹⁸F-FLT PET. Patient-based sensitivity therefore was 100%.

Region-Based Analysis

Twenty-two true-positive, 3 true-negative, 3 false-negative, and 2 false-positive regions were observed, resulting in a pathologically proven sensitivity of 88% (Table 2). There were 3 true-negative and 2 false-positive lesions.

Two of the 3 false-negative results were caused by multiple cutaneous and subcutaneous malignant lesions (satellite or primary lesions) located on the back of 2 patients, with diameters ranging from 1 to 10 mm and mitotic rate ranging from 5 to 16 mitoses per 2 mm² (Table 3). The other false-negative result was caused by a micrometastasis in a lymph node in the groin.

One of the 2 false-positive lesions was located in the groin, but no positive lymph nodes were found after resection of the groin area. The other false-positive lesion was located in the heel. This was the location of the primary melanoma, which was resected in total 4 y earlier and is still clinically negative.

The 3 true-negative regions, which displayed very little or no ¹⁸F-FLT uptake, were 1 region with negative lymph nodes after resection and 2 resected benign skin lesions, 1 located on the calf and 1 on the arm.

At the lesion level, analysis of tumor size and mitotic rate of metastatic lymph nodes revealed that the smallest detected lesion consisted of 3 closely adjacent lymph nodes, with a micrometastasis (<2 mm) in each lymph node. The detected lesion, with the lowest mitotic activity that was still detected, had a mitotic rate of 9 mitoses per 2 mm².

In 2 of the above-mentioned true-positive lesions, ¹⁸F-FLT PET detected malignancy, which was initially missed with conventional staging techniques. In both patients, this had therapeutic consequences because either dissection or radiation therapy was now indicated, but this had no influence on staging. These lesions were found in patients 1 and 2. In patient 1, a lesion in the fossa poplitea was indicated as malignant by ¹⁸F-FLT PET and was confirmed on US-guided aspiration cytology. The patient underwent a groin dissection and a resection of the popliteal lesion. In the supraclavicular region of patient 2, an additional lesion was detected, which was confirmed by US-guided aspiration cytology. This patient received radiation therapy for locoregional control.

Additional Lesions

Eleven previously unknown lesions were present on ¹⁸F-FLT PET (Table 3). These were not histopathologically confirmed but were analyzed with conventional staging techniques. Five lesions were true-positive when compared with CT or physical examination. Two mediastinal and 2 paraaortal lesions were confirmed as metastases (>1 cm); 1 lesion in the supraclavicular lesion was confirmed on physical examination.

Four of these 11 unknown lesions were true-negative based on a completely normal CT scan or clinical follow-up. One mediastinal lesion and 1 paraaortal lesion were negative on CT (<1 cm). Two lesions in the head and neck area were confirmed to be clinically negative as well during clinical follow-up.

Finally, 2 lesions detected by ¹⁸F-FLT PET and interpreted as malignant could not be evaluated. The lesion in patient 2, which was interpreted as benign, was located in a lumbar vertebra, which was below the level of the CT scan. The bone scan, which was performed 2.5 mo later, showed multiple lesions in the spine. It can be assumed that these lesions were already present at the time of the ¹⁸F-FLT PET. The remaining lesion in patient 5 was located in the area of popliteal lymph nodes of the involved leg next to a vascular prosthesis. However, recently this patient developed brain metastases, for which he was palliatively treated. During a follow-up of period of 3 mo, this lesion remained clinically negative. Although the above-mentioned 2 lesions were not confirmed by histopathology, it can be assumed that these 2 lesions were malignant because these patients developed disseminated disease within 12 wk.

There were 4 lesions, which were missed on ¹⁸F-FLT PET, without histopathologic evaluation (Table 3). These 4 lesions were false-negative on ¹⁸F-FLT PET when compared with CT. Three of these lesions were interpreted as mediastinal lymph node metastases (>1 cm) on CT and 1 was interpreted as a bone metastasis in a thoracic vertebra on CT.

Effect of ¹⁸F-FLT PET on Staging

¹⁸F-FLT PET detected the extent of the locoregional disease correctly in all patients (Table 4). The clinical (before surgery), ¹⁸F-FLT PET, and pathologic (after surgery) staging were compared. All patients were clinical stage III. However, 1 patient was upstaged to stage IV both by CT and by ¹⁸F-FLT PET, as both modalities detected mediastinal metastases. Another patient was upstaged by ¹⁸F-FLT PET as compared with the clinical presurgical staging, which was confirmed by CT. The detected metastases, which caused the upstaging in the second patient, were located in paraaortal (lymphatic) region.