FAPI PET/CT Shows False Positives After Liver Radiation

Two case reports published in EJNMMI Reports document instances where [68Ga]Ga-FAPI positron emission tomography/computed tomography (PET/CT) produced false-positive results in patients with hepatocellular carcinoma (HCC) following stereotactic body radiation therapy (SBRT). The study, authored by M. Shivacharan, Sumeet Suresh Malapure, Kirthan Chatra, and Shirley Lewis, highlights a critical interpretative limitation of this molecular imaging technique in the post-treatment surveillance of liver malignancies. In both reported cases, increased FAPI uptake within the irradiated liver segments was initially concerning for tumour recurrence but was ultimately attributed to post-radiation fibroinflammatory changes, confirmed through correlative magnetic resonance imaging (MRI) and clinical assessment.

Fibroblast activation protein (FAP) is a type II transmembrane serine protease selectively overexpressed on cancer-associated fibroblasts, which constitute a major component of the tumour microenvironment in HCC. FAP expression is minimal in normal adult liver parenchyma but is upregulated in activated fibroblasts associated with chronic liver disease, cirrhosis, and hepatic malignancies. This biological profile has led to the development of radiolabelled FAP inhibitors, or FAPI, for PET imaging, which enables the indirect visualisation of the tumour stroma rather than the tumour cells themselves. The high affinity of FAPI for activated fibroblasts makes it a promising tool for HCC detection, often outperforming traditional [18F]FDG PET/CT, which is insensitive for many well- or moderately differentiated tumours with high glucose-6-phosphatase activity.

Comparative studies have demonstrated that [68Ga]Ga-FAPI PET/CT exhibits markedly higher sensitivity than conventional [18F]FDG PET/CT for detecting HCC lesions. In a study by Wang et al., [68Ga]Ga-FAPI-04 PET detected 85.7% of HCC lesions versus 57.1% for [18F]FDG, a disparity that was especially large in small and low-grade tumours. Meta-analyses report pooled sensitivities around 94–98% for FAPI versus roughly 56–63% for FDG in liver tumours. In HCC-specific analyses, the advantage is even more striking, with FAPI sensitivity reaching approximately 98.5% compared to 60.9% for FDG. This superior detection stems from the high avidity of HCC stroma for FAP ligands combined with the low physiological liver background uptake of [68Ga]Ga-FAPI.

Despite its high sensitivity, FAPI PET/CT is not perfectly tumour-specific. Any process that activates fibroblasts, such as wound healing, fibrosis, or inflammation, may yield FAPI uptake. Ablative therapies like transarterial chemoembolization, radiofrequency ablation, or stereotactic radiotherapy induce local inflammation and scar formation, which can appear as increased uptake on FAPI imaging. Prior reports have observed FAPI uptake at prior treatment sites that MRI showed to be benign post-treatment fibrosis. Therefore, following locoregional therapy or in benign fibrotic conditions, an increased FAPI signal must be interpreted with caution, as it may reflect healing tissue rather than active cancer.

The first case involved a 68-year-old man with biopsy-proven oligometastatic HCC, alcohol-related liver disease, type 2 diabetes mellitus, and hypertension. He presented for routine post-SBRT surveillance after receiving 50 Gy in five fractions to the liver and 30 Gy in three fractions to the left iliac crest. His treatment history also included over 25 cycles of systemic Atezolizumab plus Bevacizumab therapy. Baseline [68Ga]Ga-FAPI PET/CT revealed two hepatic lesions in segments V/VI and IV, along with bone metastasis. In June 2025, a follow-up PET/CT scan showed an interval increase in FAPI avidity in the previously SBRT-treated liver segment VI, a finding initially concerning for tumour recurrence.

However, a concurrently performed MRI of the abdomen in June 2025 revealed a T1 hyperintense, T2 hypointense non-enhancing lesion with capsular retraction in segment VI. These features were consistent with a near-complete response under LI-RADS Treatment Response criteria, specifically classified as non-viable. The MRI findings indicated post-radiation fibrosis, with no arterial enhancement to suggest viable tumour. Tumour markers showed a mildly rising trend in Alpha-fetoprotein (AFP) from 12.6 to 35 ng/mL, but the patient remained clinically stable with no new complaints. The focal FAPI avidity was ultimately attributed to post-radiation fibrosis, consistent with a false-positive uptake.

The second case involved a 77-year-old man with alcohol-related chronic liver disease, type 2 diabetes mellitus, hypertension, and alcohol-associated cognitive impairment. He presented in February 2023 with a radiologically diagnosed HCC, classified as BCLC Stage A. Cross-sectional imaging demonstrated a dominant 2.5 × 2.6 cm arterially enhancing lesion in the right hepatic lobe and a 10 mm lesion in segments VII/VIII, against a background of liver cirrhosis and portal hypertension. Following multidisciplinary tumour board discussion, the patient underwent SBRT with curative intent, receiving 40 Gy in five fractions from March 20 to 30, 2023. The treatment was well tolerated, with only mild constipation reported.

At eleven months post-SBRT in January 2024, the patient’s serum AFP had risen to 260 ng/mL. An MRI abdomen performed in late January 2024 demonstrated an interval reduction in the size and number of hepatic lesions, consistent with post-radiation change, without a morphologically defined residual or recurrent lesion. Given the discordance between the rising AFP and the favourable MRI findings, the patient underwent whole-body [68Ga]Ga-FAPI PET/CT to evaluate for extrahepatic progression. The PET/CT demonstrated ill-defined, intensely FAPI-avid lesions in hepatic segments VII/VIII, measuring 7.5 × 6 × 5.5 cm, with a maximum standardised uptake value (SUVmax) of 7.83. These lesions corresponded to the previously irradiated field.

A second, mildly FAPI-avid subcapsular hypodensity was noted in segment VIII without significant enhancement. No distant metastases or skeletal lesions were identified. The primary diagnostic challenge was the convergence of a rising AFP, an intensely FAPI-avid lesion in the irradiated field, and a second indeterminate lesion. However, the concurrently performed MRI demonstrated an ill-defined T2/STIR hyperintense signal alteration in segment VII without diffusion restriction and without a discrete morphological lesion. These findings were inconsistent with viable tumour. The mild delayed washout on PET/CT, while a feature that can overlap with malignant imaging characteristics, was interpreted in the context of the discordant MRI and the known behaviour of FAPI in post-radiation fibroinflammatory tissue.

Multidisciplinary tumour board review in February 2024 concluded that the imaging findings most likely represented post-radiation fibroinflammatory change in the treated field. The board recommended medical oncology and palliative medicine follow-up without further intervention. At the 14-month post-SBRT review in April 2024, the FAPI uptake was confirmed as a false-positive post-radiation change, with no recurrence or residual viable disease identified. It should be noted that the patient’s overall clinical course was complicated by progressive alcohol-related cognitive impairment and declining functional status, with an ECOG Performance Status of 2. The false-positive interpretation remains based on the evidence available at the index evaluation.

These cases illustrate that SBRT-induced hepatic fibrosis can lead to false-positive FAPI uptake, a previously unreported challenge in interpreting FAPI PET/CT for post-radiation HCC. In most SBRT responders, AFP levels decline substantially after treatment, with a median decline around 96.5% at three months in recent cohorts. Persistent arterial phase hyperenhancement alone does not suggest recurrence, as post-SBRT radiologic response evolves slowly. Standard response criteria such as mRECIST should be applied cautiously in the early post-treatment period, as in-field progression rates are low and true recurrence is better distinguished from expected post-radiation changes at later follow-up.

Structural imaging with contrast-enhanced MRI or CT remains the standard for detecting recurrence and characterising indeterminate lesions. FAPI PET/CT should complement, not replace, conventional imaging. In these cases, MRI was crucial for adjudicating the FAPI findings. To the authors’ knowledge, this report is among the first to document FAPI avidity after hepatic SBRT that correlates with MRI-proven post-radiation fibrosis. The findings highlight the need for prospective studies to define the clinical role of FAPI PET/CT in post-treatment surveillance, ensuring that isolated FAPI avidity in a treated field is not taken as definitive evidence of recurrence without concordant anatomical imaging and biochemical trends.

Subject of Research: False-positive [68Ga]Ga-FAPI PET/CT uptake in hepatocellular carcinoma following stereotactic body radiation therapy

Article Title: [68Ga] Ga-FAPI positron emission tomography/computed tomography uptake following liver stereotactic body radiation therapy in hepatocellular carcinoma: two case reports

Article References: Shivacharan, M., Malapure, S. S., Chatra, K., & Lewis, S. (2026). [68Ga] Ga-FAPI positron emission tomography/computed tomography uptake following liver stereotactic body radiation therapy in hepatocellular carcinoma: two case reports. EJNMMI Reports, 10(1), Article 48. https://doi.org/10.1186/s41824-026-00315-6

Image Credits: AI Generated

DOI: 10.1186/s41824-026-00315-6

Keywords: hepatocellular carcinoma, FAPI PET/CT, stereotactic body radiation therapy, false-positive uptake, post-radiation fibrosis, liver cancer surveillance, molecular imaging, MRI correlation, Ga-FAPI, positron, emission, tomography