Positron emission tomography (PET) is a relatively new imaging technique that is increasingly used in oncology. It involves introducing into the body a biologically relevant tracer molecule that contains a positron-emitting radioactive isotope. Positrons are “anti-electrons” and when one encounters an electron both particles destroy each other, emitting two gamma rays in opposite directions that can be detected. As positron-electron encounters occur essentially as soon as the positrons are emitted, the location of the tracer molecule can be determined from these gamma rays. In principle, any molecule can be labeled allowing the metabolism of any biological process of interest to be studied. This technique is much more sensitive than other imaging methods, including MRI, in detecting regions of metabolically active tissue.

A number of novel applications of PET imaging were discussed at the National Cancer Research Institute’s recent conference, held this year in early October, in Birmingham. This annual event is the largest international cancer research meeting held regularly in the UK. One of the 30 parallel sessions was devoted specifically to this topic, and PET based techniques also featured strongly in sessions on invasion and metastasis, radiation oncology, and brain tumours.

Sibylle Ziegler (Technical University of Munich, Germany) started off the dedicated PET session with a general introduction to the technique, the instrumentation involved, and its uses in oncology. It is a high-resolution technique: clinical PET machines can resolve distances down to 5-7 mm, and research machines down to 1-2 mm. It is also versatile, and can be used to detect different tissues and molecular processes depending on the radio-labelled tracer molecule used. Tumours, which take up more glucose than normal tissue, are commonly visualised using the glucose analogue flurodeoxyglucose (FDG) labelled with the positron emitter 18F. This isotope is the most widely used in PET because of its relatively long half life of just under two hours. Isotopes with much shorter half-lives can only be used if the radionuclides can be produced directly at the site of a Nuclear Medicine facility, as their radioactivity will decay before they can be transported even a short distance. Recently, PET has been combined with computer tomography (CT), enabling PET’s functional imaging to be combined with anatomical information in a single scan. Combining PET with magnetic resonance is a very active research field but the only clinical facilities are a few small bore scanners dedicated to brain imaging only.

Stefano Fanti (University of Bologna, Italy) gave a useful summary of PET techniques in oncology that use radio-tracers other than the standard FDG. These are still uncommon, but a number of applications that do not involve glucose metabolism have been developed. These frequently involve choline, labelled with either 18F or 11C (although the latter isotope has a half-life of only about 20 minutes, so radio-tracers that use it must be synthesised on-site). One PET application using radio-labelled choline that is becoming relatively common is in the diagnosis and staging of prostate cancer. An alternative tracer is also needed for imaging the central nervous system, as sugar analogues cannot penetrate the blood-brain barrier. Amino acids and their analogues, such as 18F-labelled fluoro-tyrosine and 11C-labelled methionine, can be used to image brain tumours, as can the catecholamine, dopamine.

Many tumours are hypoxic, that is, they have low concentrations of oxygen, and the degree of hypoxia can be related to prognosis, particularly to response to radiotherapy. John Humm (Memorial Sloan-Kettering Cancer Centre, New York USA) is exploring methods for measuring tumour hypoxia using PET. “Hypoxia is usually measured by pushing oxygen pressure probes into patients. This, in contrast, is a non-invasive technique”, he says. His most recently published work is a trial of PET imaging with the radio-tracer 18F-fluoromisonidazole, which is taken up most intensely by hypoxic tissue, to measure the extent of hypoxia in head and neck tumours. Thirteen patients were scanned twice, three days apart, and seven were found to have stable regions of high uptake that are characteristic of chronic hypoxia. Such images could potentially be used for “dose painting” i.e. delivering a higher radiation dose to the radioresistant regions of the tumor in a course of radiotherapy.

Clinical trials have shown that PET imaging using amino acids as the tracers can be more sensitive and specific in the diagnosis of brain tumours, particularly gliomas, than other imaging methods such as MRI. Anca Grosu (University of Freiburg, Germany) presented a summary of the literature in this area. She evaluated studies of the role of PET using either 11C-methionine (MET-PET) or 18F-fluoroethyl tyrosine (FET-PET) in the diagnosis of brain gliomas. Imaging amino acid metabolism with PET could typically diagnose and locate malignant glioma with sensitivity and specificity of over 85%. Furthermore, and in agreement with Humm’s results with head and neck cancers, PET imaging has also been found effective for investigating the extent of hypoxia in brain tumours. Fluorine-labelled imidazoles are most often used for this application but 60Cu-labelled methylthiosemi-carbazone has also been proposed, although this radionuclide is another of those with a half-life of less than half an hour.

The NCRI has a website dedicated to PET in cancer research and clinical care at http://ncri-pet.org.uk/.