# [18F]FMISO PET in metastatic neuroendocrine tumours: a pilot study

## TL;DR

This pilot study explores whether hypoxia explains the 'DONUT' phenomenon in neuroendocrine tumors using [18F]FMISO PET scans.

## Contribution

The study introduces dynamic [18F]FMISO imaging to evaluate hypoxia in DONUT lesions of metastatic neuroendocrine tumors.

## Key findings

- Dynamic [18F]FMISO imaging showed increased uptake in DONUT lesions for 8 out of 10 patients.
- Only one patient showed [18F]FMISO uptake greater than normal liver on delayed static imaging.
- Dynamic imaging is suggested as more effective than static imaging for detecting hypoxia in these lesions.

## Abstract

The phenomenon of peripheral [68Ga]DOTATATE avidity without central avidity (which we have termed a “DONUT") has been observed in neuroendocrine neoplasm (NEN) lesions. There has been speculation as to whether this is due to hypoxia, de-differentiated disease or other causes. The presence of hypoxia may have prognostic and therapeutic implications, and was evaluated in these lesions using the PET hypoxia imaging biomarker [18F]FMISO.

Prospective pilot study in patients with metastatic NENs with at least one DONUT lesion (central [68Ga]DOTATATE non-avidity). [18F]FDG and [18F]FMISO scans were acquired within 60 days of the [68Ga]DOTATATE PET/CT. [18F]FMISO scans were acquired as a dynamic scan over 20 mins from injection with a delayed image at 2 hours. The dynamic acquisition was analysed quantitatively using a graphical approach yielding parametric images of Influx Rate Constant and Volume of Distribution. [18F]FMISO uptake within the identified DONUT hole on the 2 hr delayed scan was qualitatively scored by two experienced nuclear medicine physicians as: 0 (no uptake), 1 (uptake less than normal liver), 2 (uptake equal to normal liver), or 3 (uptake greater than normal liver).

Ten patients were enrolled with primary sites including pancreas (n=3), small bowel (n=3), rectum (n=2), duodenum (n=1) and lung (n=1). Six subjects were scored 1, three subjects were scored 2, and one subject was scored 3. All lesions evaluated were located in the liver. Quantitative [18F]FMISO parametric imaging showed evidence of increased uptake rate (Ki) in the photopenic areas of the DONUT lesions in 8/10 subjects. Surrounding uptake rate in normal liver was extremely low. In the qualitative delayed image assessment, only one subject demonstrated [18F]FMISO uptake greater than surrounding normal liver (small bowel primary, G2).

Only one of ten patients with DONUT lesions demonstrated increased [18F]FMISO uptake rate on delayed static imaging. In contrast, dynamic imaging demonstrated increased [18F]FMISO uptake rate in the region of [68Ga]DOTATATE photopenia on 8 of 10 patients. Future research using [18F]FMISO in NEN patients should incorporate dynamic imaging.