For any PET exam, a trade-off exists between radiation dose exposure to a patient and the diagnostic quality of the acquired images. Because it is unethical to subject a patient to multiple exams of different tracer doses, lead author Sergios Gatidis, of the Department of Radiology's diagnostic and interventional section, and colleagues developed a method for simulating low-dose PET images. The resulting images enable direct comparison of different tracer doses in a single patient.1

In earlier studies, the authors showed that radiation exposure could be reduced by up to 70% by replacing PET/CT with PET/MRI using the tracer 18F-FDG (FDG). In a study of 18 paediatric cancer patients with solid tumours, they determined that PET/MRI and PET/CT achieved similar lesion detection rates, with MRI data from areas of soft tissue providing additional diagnostic information that could not be seen on CT.2 A subsequent study showed that FDG-PET/MRI without contrast was technically feasible, and at least equivalent to FDG-PET/CT, for tumour staging and characterization in patients under the age of six.3

In this latest study, the authors assessed the feasibility of reducing radiation exposure from the PET portion of the PET/MRI scan.

Simulation study

The simulation methodology allows reconstruction of PET images with equivalent properties to PET images measured after injection of lower tracer doses. The use of randomized subsampling of PET data enables more realistic clinical dose simulations incorporating dynamic measurement effects, such as tracer uptake for longer measurements, decay of short half-life tracers and/or patient movement.

The researchers simulated different tracer activities (0.25 to 2.5 MBq/kg) using data from 30 whole-body FDG-PET/MRI exams of 24 paediatric patients diagnosed with or suspected of having solid tumour malignancies. A paediatric radiologist and a nuclear medicine physician analysed the PET data independently. They assessed the visual detectability of physiologic FDG uptake in 19 anatomic structures and the detectability of potentially pathologic focal FDG uptake in other parts of the body, classifying all findings considered relevant to an oncologic diagnosis. The readers specified the lowest level of tracer activity simulated that had sufficient image quality. False positive findings were also calculated.

Study results showed that 88% of data sets with a simulated activity of 1.5 MBq/kg FDG, using point spread function (PSF) reconstruction, had sufficient image quality for accurate interpretation. The use of PSF modelling significantly improved the detection of focal FDG-avid lesions. Lesion detectability and SUV quantification were not relevantly impaired down to simulated tracer activities of 1.5 MBq/kg FDG.

Since current international guidelines for FDG application in paediatric oncology recommend tracer activities of 3 to 5 MBq/kg FDG, this represents a reduction of radiation dose of more than 50%. However, the authors note that this may not be feasible for all scans ordered, for example, assessing response to treatment of Hodgkin's lymphoma. Radiologists need to determine optimal tracer levels for each individual patient, based on patient characteristics and the clinical question to be answered.

Gatidis and colleagues point out that even greater reduction might be possible with longer image acquisition time. The original PET image acquisition took four minutes per bed position, instead of the more commonly used two to three minutes. They suggest that it is worth investigating a five minute per position acquisition to determine whether tracer activity could be lowered further, noting that the additional exam time needed would not exceed the time to perform the MRI scan.

Benefit to cancer survivors

Multiple studies have shown that children who survive paediatric cancer into adulthood and middle age have a higher risk of developing additional cancers or serious medical conditions, as a result of cumulative radiation exposure from radiotherapy and/or routine imaging follow-ups.

A 2008 study estimating cumulative radiation dose from PET/CT scans ordered for diagnosis, staging and follow-up in 78 cancer patients aged one year to 18 determined that the average cumulative radiation dose was 64.4 mSv for CT (with a range of 2.7 to 326 mSv) and 14.5 mSv for PET (with a range of 2.8 to 73 mSv). This did not include dose from radiotherapy, which almost half of paediatric cancer patients receive. Reducing radiation exposure from any source is of critical importance.

Related stories

• Cardiac PET/MRI protocol cuts radiation dose
• PET/MRI: ideal for brain tumour imaging
• PET/MRI matches PET/CT with less dose
• Reducing FDG activity cuts PET/MRI dose