FAP-targeted PET finds more recurrent medullary thyroid carcinoma lesions in retrospective study

FAP-targeted positron emission tomography using the gallium-68-labelled tracer [68Ga]BED003 detected more lesions than [18F]DOPA-PET in a small retrospective study of eight patients with recurrent or metastatic medullary thyroid carcinoma. Across 35 tracer-avid lesions, [68Ga]BED003-PET identified 34, or 97.1%, compared with 16, or 45.7%, identified by [18F]DOPA-PET, a statistically significant difference (P < 0.001). The FAP-targeted scans also produced higher uptake and tumour-to-background contrast. These results support further evaluation of [68Ga]BED003 as a complementary imaging method when [18F]DOPA-PET fails to explain elevated or rising tumour markers, but they do not establish it as a replacement for established imaging.

Medullary thyroid carcinoma is a rare neuroendocrine cancer that develops from calcitonin-producing parafollicular C cells in the thyroid and represents approximately 1–2% of thyroid cancers. Surgery remains the only potentially curative treatment, whereas recurrent or metastatic disease may be managed with local therapies in selected cases and systemic treatment when it progresses. Calcitonin and carcinoembryonic antigen are established tumour markers used during follow-up, while ultrasound, computed tomography and magnetic resonance imaging remain important diagnostic tools. [18F]DOPA-PET is an established molecular imaging reference for recurrent medullary thyroid carcinoma, although its detection depends on calcitonin levels, tumour burden, lesion size and location, and its clinical availability remains limited to relatively few specialist centres.

The study targeted fibroblast activation protein, or FAP, which is expressed in the tumour-associated stroma of medullary thyroid carcinoma. [68Ga]BED003 is a gallium-68-labelled PET tracer derived from the OncoFAP platform, developed as an ultra-high-affinity small-molecule FAP ligand. The investigators selected eight patients with histologically proven medullary thyroid carcinoma and suspected recurrence or progression. Each had elevated and/or increasing calcitonin levels that did not correspond to the findings of [18F]DOPA-PET, and all subsequently received the additional [68Ga]BED003 examination. This clinically enriched design means the cohort does not represent an unselected population of patients with recurrent medullary thyroid carcinoma.

The two PET examinations were performed a median of 33 days apart, with a range of seven to 56 days. Imaging used either combined PET/CT or PET/MRI systems. [18F]DOPA-PET acquisition began 30–60 minutes after intravenous administration of 2 MBq/kg, while [68Ga]BED003-PET acquisition began 60 ± 10 minutes after 2.5 MBq/kg. Two senior nuclear-medicine physicians and one senior radiologist reviewed the images using consensus-based clinical decision-making. Because no established tracer cutoffs were available, suspect uptake above background was compared with CT, MRI, previous or follow-up imaging and multidisciplinary assessment. Selected patients underwent biopsy or surgery for histopathological confirmation. The investigators measured maximum and mean standardized uptake values, metabolic tumour volume and maximum tumour-to-background ratios using blood pool, liver, lung and muscle reference regions.

At the patient level, [18F]DOPA-PET was positive in seven of eight patients, while [68Ga]BED003-PET was positive in six. In five patients, [68Ga]BED003-PET detected more tracer-avid lesions than [18F]DOPA-PET; another patient had the same tracer-avid lesions on both examinations. The two patients with discordant overall findings illustrated the methods’ different biological signals. One had left adrenal hyperplasia with increased [18F]DOPA uptake but no [68Ga]BED003 uptake. The other had symmetrical, low-level [68Ga]BED003 uptake in mediastinal lymph nodes but no tumour-avid findings with either tracer. Subsequent endobronchial ultrasound-guided fine-needle aspiration showed chronic inflammation from anthracosis without malignancy, demonstrating that inflammatory lymph nodes can be FAP-positive.

Across all 35 tracer-avid lesions, 15 were detected by both methods, 19 were detected only by [68Ga]BED003-PET and one only by [18F]DOPA-PET. The largest differences occurred at particular metastatic sites. [68Ga]BED003 detected all seven bone lesions, all six lung lesions and all five liver lesions, compared with four, one and zero, respectively, detected by [18F]DOPA-PET. It detected 15 of 16 lymph-node metastases, compared with 11 for [18F]DOPA-PET, and it detected the single local recurrence, which was not seen with [18F]DOPA-PET. After the patient with the longest interval between examinations was excluded, [68Ga]BED003 still detected 34 lesions compared with 15 for [18F]DOPA-PET (P < 0.001). The sensitivity analyses did not indicate that a longer interval explained the relative advantage.

The quantitative measurements followed the same pattern. Median maximum standardized uptake value was 6.62 for [68Ga]BED003 and 2.10 for [18F]DOPA, while median mean uptake was 3.90 and 1.50, respectively; both comparisons had Bonferroni-adjusted P values below 0.001. Median metabolic tumour volume was 1.38 cm3 with [68Ga]BED003 and 1.06 cm3 with [18F]DOPA, with a Bonferroni-adjusted P value of 0.0033. All four tumour-to-background measurements were significantly higher with [68Ga]BED003. Median liver TBRmax was 9.83 versus 0.94, lung TBRmax was 10.78 versus 8.50, muscle TBRmax was 5.41 versus 3.22 and blood-pool TBRmax was 6.02 versus 2.23; each comparison had a Bonferroni-adjusted P value below 0.001. No significant scanner effect was found after adjustment for multiple comparisons.

Histopathological confirmation was available for five of the eight patients and supplied both confirmations and warnings about tracer interpretation. A liver lesion missed by [18F]DOPA-PET but detected by [68Ga]BED003 proved to be a medullary thyroid carcinoma metastasis, while a separate [18F]DOPA-positive adrenal lesion was found to be a pheochromocytoma rather than metastatic disease. Biopsy or surgery also confirmed medullary thyroid carcinoma in a local recurrence, a hilar lymph node and a liver metastasis visible only with [68Ga]BED003. Conversely, an [18F]DOPA-positive infraclavicular lymph-node metastasis was missed by [68Ga]BED003, and [68Ga]BED003-positive hilar nodes in another patient proved inflammatory. The [68Ga]BED003 findings contributed to subsequent management decisions in three patients and prompted additional diagnostic clarification in a fourth, although the study did not measure effects on survival or long-term clinical outcomes.

Several biological and technical factors may contribute to the observed differences, although the study cannot establish which mechanism was responsible. Physiological hepatic and biliary activity can reduce the conspicuity of [18F]DOPA-positive liver lesions, whereas [68Ga]BED003 generally produced low liver background and high contrast. Small pulmonary metastases are susceptible to respiratory motion and partial-volume effects, which may lower measured uptake and make low-contrast [18F]DOPA lesions harder to distinguish. FAP, however, is not specific to cancer. Hepatic fibrosis, inflammation, post-treatment changes and benign tumours may show tracer uptake; in this cohort, intensely [68Ga]BED003-positive liver lesions in one patient were classified as most consistent with haemangiomas based on MRI appearance and long-term stability. The false-positive lymph nodes further show that [68Ga]BED003 positivity cannot automatically be equated with malignancy, while a negative scan cannot reliably exclude every lesion in every patient.

FAP imaging may have theranostic implications, but uptake in this cohort did not meet a proposed selection threshold for some available short-retention FAPI therapeutic compounds: a maximum standardized uptake value of at least 10 in more than half of a patient’s lesions. Five of 35 lesions reached an SUVmax of at least 10, occurring in three of seven patients with metastatic lesions, but none met that patient-level criterion. The study was limited by its single-center retrospective design, eight-person sample, selection of patients with inconclusive [18F]DOPA-PET findings, absence of histopathological confirmation for every lesion and fixed order of imaging. Incorporation and observer bias could not be fully excluded. No early [18F]DOPA acquisition, FAP immunohistochemistry or lesion-based dosimetry was available. The results therefore provide preliminary evidence for [68Ga]BED003-PET as a complementary option in selected recurrent or metastatic cases and call for prospective multicenter validation.

Subject of Research: Paired PET imaging for recurrent or metastatic medullary thyroid carcinoma

Article Title: FAP-targeted [68Ga]BED003-PET compared with [18F]DOPA-PET in patients with recurrent medullary thyroid carcinoma

Article References: Ventura, D., Noto, B., Backhaus, P., Schindler, P., Rahbar, K., Schäfers, M., Riemann, B., & Roll, W. (2026). FAP-targeted [68Ga]BED003-PET compared with [18F]DOPA-PET in patients with recurrent medullary thyroid carcinoma. Frontiers in Nuclear Medicine, 6, 1949036. https://doi.org/10.3389/fnume.2026.1949036

Image Credits: AI Generated

DOI: 10.3389/fnume.2026.1949036

Keywords: medullary thyroid carcinoma, FAP-targeted PET, [68Ga]BED003, [18F]DOPA-PET, OncoFAP, molecular imaging, cancer recurrence, tumour-to-background ratio, theranostics, FAP-targeted, BED003-PET, compared