Medworks - the ultimate medical website

Positron Emission Tomography

Dr Furrat Amen

Introduction

PET has been available as a tool for the past twenty to thirty years. In the 1920s, it was hypothesised by Dirac that the electron had a opposite number and this was confirmed experimentally in 1932 by Anderson.

The first PET scanner was constructed in 1974 at Washington University after several abortive events in the 1950s and 1960s. Nowadays, these machines have thousands of small crystals, which measure two to four millimetres in diameter and are arranged in a ring around the patient (figure 1). Alternatively cheaper machines are available which contain a few block electrodes which are capable of detecting positron emitting isotopes at a reduced resolution.

Positrons are emitted from an unstable nucleus and travel for a few millimetres before hitting an eIectron. Their masses combine and form two 511 keV gamma rays which travel in opposite directions (see figure 1). These can be detected by a PET camera which recognises them using light-sensitive crystals. In this way, mathematics allow a computer to determine where this event occurred and can translate the numbers into a graphical representation.

Figure 1

The machine is programmed to rotate and allows for 'coincidence detection' of the two photons and attenuation, a feature not possible with single photon detection. Attenuation of radiotracer activity occurs because as photons pass through the body tissue, some are scattered or absorbed.

Most organic molecules can be labelled with positron emitting isotopes. Some examples of elements which may be manipulated in this way are: carbon, oxygen, nitrogen and fluorine. The most commonly used isotope is [18F] fluorodeoxyglucose (FDG) and can be useful in discovering sites of neoplasia.

This effect relies on the fact that malignant cells have increased metabolism and tend to have rapid cell proliferation with a concomitant increase in glucose metabolism. FDG may be substituted for glucose in such circumstances and can be metabolised and detected by PET imaging.

FDG is phosphorylated to FDG-6-PO4 and this is trapped in tumour cells and cannot be metabolised any more. The relative uptakes of FDG may be taken to be a good indication of growth rates and the degree of aggressiveness of tumours.

PET uses tracers which, while similar to stable isotopes in the body, do not distort the biochemical reactions being monitored. Tracers also have a short half-life which is good for reducing patient exposure to radiation. They are made on-site using cyclotrons.

A problem with the use of FDG is that by implication its uptake is competitive with that of glucose and so it may be advisable to fast the patient before attempting to use this imaging technique. Diabetes and serum hyperglycaemia may decrease FDG uptake in tumours although this might be addressed by changing to a different tracer.

Figure 2. A PET machine

Uses

The practice normally employed is to give an iv dose of 10.0 to 20.0 mCi of FDG and perform the PET scan thirty to sixty minutes later with a completion of imaging an hour after that. 'Attenuation correction' is a modification to the scan which measures a value for a particular patient's body and corrects the values of radiation detected accordingly. This allows for smaller and deeper lesions to be detected. The downside of this technology is that it takes a longer time to conduct the scan and that if the patient moves it adds noise to the image.

A local image may take fifteen to thirty minutes to take and a whole body scan may take one hour. A typical scenario might be that a patient with lung cancer might have a whole body scan to look for distant metastases and a localised and attenuated scan for assessment of local nodal disease. Such a scan might cost $2000 to $3000.

Lung cancer often presents with a nodule or non-specific opacity. It is difficult to differentiate benign from malignant lesions on plain radiographs, CT, and MRI especially before and after surgery or radiotherapy. Other problems include negative results from transbronchial or transthoracic biopsies which are not necessarily reliable and are invasive. It might be more beneficial to obtain a plain radiograph and then a PET scan to work out the metabolic activity of a lesion to differentiate the benign from the malignant.

Patients may be compared using a standardised uptake ratio (SUR) to put into perspective FDG uptake, or alternatively scans may be visually interpreted as showing malignancy by displaying increased uptake in comparison to mediastinal uptake. SUR may be thought of as the concentration of radiotracer in the tumour divided by the concentration of injected radiotracer distributed throughout the volume of the body.

A study of thirty patients with solitary lung nodules showed PET to have 95% specificity and sensitivity and a highly significant difference between SURs in benign and malignant lesions. However PET is, like other imaging modalities, not 100% accurate. Some lesions which are benign may display high metabolic activity, giving rise to false positives. Some inflammatory lesions or areas of active infection may accumulate FDG eg. TB granulomas and aspergillosis.

Figure 4. The CT scan of the chest confirms the presence of a mass in the right chest wall (green arrows). Whole-body FDG-PET confirms hypermetabolism of the mass (red arrows), highly suspicious for recurrence of tumour. Note that the increased activity seen in the arm (yellow arrow) is the injection site; this should not be confused with abnormal uptake.

False negatives are also possible with some tumours which have a comparatively low metabolic activity e.g. broncholoalveolar tumours and carcinoid tumours. These also arise if a tumour is small and the resolution is four to eight millimetres, or when the malignant component is a small part of a benign lesion.

The 'Prospective Investigation Of PET in Lung Nodules' study (PIOPLIN) showed that there is a lower sensitivity for detecting lesions less than 1.5cm in size - 80% vs. 96% - although a the difference was not deemed to be significant and another study found no actual difference.

Comparing transthoracic needle aspiration with PET shows that more malignancies are missed with biopsy and PET results in more unnecessary thoracotomies. The invasive nature of needle biopsies results in 27% of patients requiring a chest drain for pneumothoracies.

PET may be helpful in staging bronchogenic carcinoma by TNM. It is limited in use for the 'T' part, being surpassed by CT. For 'N' it has been shown to have similar sensitivity to mediastinoscopy (Van Schil et al, 1989). With the 'M' part, it found 10% of patient to have distant metastases not shown on CT. It thereby upstages patients and affects management. False positives as delineated by CT, e.g. adrenal nodules, are correctly differentiated as negative by PET. Another advantage of PET is that no additional radioactivity is required for a whole body scan (Valk et al, 1995).

Therapy of bronchogenic carcinoma is associated with poor outcome. Most disease is seen with an advanced stage and a five-year survival of thirteen percent. Historically, response to chemotherapy and radiotherapy was assessed by tumour size and PET may offer advantages over this. While a partial reduction in FDG uptake is not beneficial, a normalisation of FDG uptake is associated with a good prognosis (Herbert et al, 1996).

PET is also capable of diagnosing recurrence early - a notoriously difficult objective to achieve under normal circumstances. Tissue biopsies of scar tissue, although negative, are questionable in value. It also is able to differentiate benign pleural thickening, induced by radiation, from recurrent tumour. The traps to be avoided include mistaking the increased metabolism of inflammatory change induced by radiotherapy for malignancy.

Figure 4. 40-year-old male. Fibrohistiocytoma. The CTscan of the chest confirms the presence of a mass in the right chest wall (green arrows). Whole-body FDG-PET confirms hypermetabolism of the mass (red arrows), highly suspicious for recurrence of tumour. Note that the increased activity seen in the arm (yellow arrow) is the injection site,- this should not be confused with abnormal uptake.

Cardiac imaging

PET may also be useful in coronary artery disease in assessing how limited myocardial perfusion is. It can differentiate reversibly from irreversibly injured ischaemic myocardium. This allows for quantification of the benefit of salvage therapy using angioplasty or thrombolytic therapy (see figure 3).

Figure 3. 70 year-old male with a recent anterior M1 and cardiac arrest which was successfully resuscitated. After hospital discharge, the patient was readmitted twice due to episodes of resting pain and ECG changes. Viable LAD territory, with stress-induced ischaemia.

Cardiomyopathies may be detected using PET. Regional myocardial perfusion is measured, e.g. compromised in angina. This used to be measured using invasive techniques such a using probes in the coronary sinus to measure flow by thermodilution, or using a doppler probe to measure coronary arterial blood flow. These invasive techniques do not account, however, for a possible difference between the macroscopic flow in the coronary arteries and the capillary perfusion, which may be different e.g. AV shunting.

Thallium-201 and 99mTc-labelled isonitriles are single photon emitting radionuclides which have variable degrees of extraction. These differences between patients result in a mismatch between the tracer accumulation in the myocardium and myocardial perfusion. More problems are produced by the difficulties caused by trying to locate single photons. No attenuation is possible and superimposed tissues above and below the area of interest interfere.

PET with 13N-ammonia, 82Rb-chloride, and 62Cu-PTS1 is complicated by incomplete extraction which is dependent on flow. A better alternative is to use 15O-water as a tracer. This is used to investigate patients with angina, but with angiographically normal arteries. Limited perfusion in cardiac allografts may also be detected using stress induced by iv dipyridamole. Angiography is invasive in comparison and has difficulty in detecting the diffuse atheromatous change characteristic of this situation. It is also difficult for the patient to detect the symptoms normally associated with angina as the heart is denervated.

Dilated cardiomyopathy shows heterogenous diminution of regional accumulation of "C-palmitate diffusely thoughout the heart. This differentiates it from the congestive heart failure picture of segmental diminution. Hypertrophic cardiomyopathy shows a decrease of septal accumulation of FDG.

Cardiomyopathy associated with Duchenne's muscular dystrophy may be picked up by the excess accumulation of FDG in segments with decreased perfusion.

Other clinical applications

PET was initially just used for imaging the brain and heart, but now it is possible to see it being used in a variety of settings.

FDG PET can also detect thyroid cancer, breast cancer, lymphoma (see figure 7), squamous cell carcinomas of head and neck, colonic cancer, ovarian carcinoma and musculoskeletal tumours (see figure 5). It can also be used to model and quantify in vivo biochemical processes as radiotracer activity is monitored serially over time.

Figure 5. Osteosarcoma. The MRI reveals a destructive lesion of the left posterior femur (green arrow). 18FPET reveals intense uptake of 18F- in the left knee (red arrows), consistent with reactive bone. FDG-PET reveals intense uptake of FDG in both the bone and soft tissue components of the tumour The PETstudies demonstrate that metastases are not likely but that the tumour is restricted to the bone and surrounding soft tissue.

It may used to differentiate brain tumours (see figure 6) from radiation necrosis, which can appear similar on CT and MRI with contrast due to the breakdown of the blood brain barrier. Necrosis shows minimal FDG activity, with the tumour showing increase FDG uptake.

Figure 6. MRI shows a rim enhancing lesion in the right hemisphere. FDG-PET reveals a hypermetabolic rim (red arrow) consistent with highgrade astrocytoma. Note that such regions of hypermetabolism can be used as a guide for potential biopsy sites.

Figure 7. The FDG-PET study reveals the presence of hypermetabolism in the hilar lymph nodes (blue arrow), as well as in the left cervical (green arrow) and right inguinal node (red arrow). Active Hodgkin's disease was confirmed by biopsy of the left cervical node.

Preoperative testing of complex partial epilepsy that arises from one of the temporal lobes. Between seizures, the site of seizure origin will have a decrease in glucose metabolism when compared with the normal size.

Prostatic carcimoma can be difficult to differentiate from benign prostatic hypertrophy which has high FDG uptake. In the early stages prostatic carcinoma has low FDG uptake (Laubenbacher et al.) and can be masked by intense bladder activity. This may be corrected by hydrating the patient and using a Foley catheter to irrigate the bladder. An ideal tracer has yet to be found but would not be excreted into bladder and colon, and would show up only areas of prostatic carcinoma.

The future

In the future, cardiomyopathy detection will be improved using this technique and be able to assess response to treatment. Three-dimensional images of hearts, with integration of spatial and temporal indications of perfusion and regional oxygen consumption, will be possible with fluxes in metabolic pathways being shown simultaneously. PET will also be useful in delineating the causes of arrhythmias in cardiomyopathic hearts. Faster computers promise to reduce the time needed to thirty minutes for a whole body scan with better resolution coming through from new types of crystals.


Write us an article. Ask us any questions about the websites, or give us suggestions.	Positron Emission Tomography Dr Furrat Amen Introduction PET has been available as a tool for the past twenty to thirty years. In the 1920s, it was hypothesised by Dirac that the electron had a opposite number and this was confirmed experimentally in 1932 by Anderson. The first PET scanner was constructed in 1974 at Washington University after several abortive events in the 1950s and 1960s. Nowadays, these machines have thousands of small crystals, which measure two to four millimetres in diameter and are arranged in a ring around the patient (figure 1). Alternatively cheaper machines are available which contain a few block electrodes which are capable of detecting positron emitting isotopes at a reduced resolution. Positrons are emitted from an unstable nucleus and travel for a few millimetres before hitting an eIectron. Their masses combine and form two 511 keV gamma rays which travel in opposite directions (see figure 1). These can be detected by a PET camera which recognises them using light-sensitive crystals. In this way, mathematics allow a computer to determine where this event occurred and can translate the numbers into a graphical representation. Figure 1 The machine is programmed to rotate and allows for 'coincidence detection' of the two photons and attenuation, a feature not possible with single photon detection. Attenuation of radiotracer activity occurs because as photons pass through the body tissue, some are scattered or absorbed. Most organic molecules can be labelled with positron emitting isotopes. Some examples of elements which may be manipulated in this way are: carbon, oxygen, nitrogen and fluorine. The most commonly used isotope is [18F] fluorodeoxyglucose (FDG) and can be useful in discovering sites of neoplasia. This effect relies on the fact that malignant cells have increased metabolism and tend to have rapid cell proliferation with a concomitant increase in glucose metabolism. FDG may be substituted for glucose in such circumstances and can be metabolised and detected by PET imaging. FDG is phosphorylated to FDG-6-PO4 and this is trapped in tumour cells and cannot be metabolised any more. The relative uptakes of FDG may be taken to be a good indication of growth rates and the degree of aggressiveness of tumours. PET uses tracers which, while similar to stable isotopes in the body, do not distort the biochemical reactions being monitored. Tracers also have a short half-life which is good for reducing patient exposure to radiation. They are made on-site using cyclotrons. A problem with the use of FDG is that by implication its uptake is competitive with that of glucose and so it may be advisable to fast the patient before attempting to use this imaging technique. Diabetes and serum hyperglycaemia may decrease FDG uptake in tumours although this might be addressed by changing to a different tracer. Figure 2. A PET machine Uses The practice normally employed is to give an iv dose of 10.0 to 20.0 mCi of FDG and perform the PET scan thirty to sixty minutes later with a completion of imaging an hour after that. 'Attenuation correction' is a modification to the scan which measures a value for a particular patient's body and corrects the values of radiation detected accordingly. This allows for smaller and deeper lesions to be detected. The downside of this technology is that it takes a longer time to conduct the scan and that if the patient moves it adds noise to the image. A local image may take fifteen to thirty minutes to take and a whole body scan may take one hour. A typical scenario might be that a patient with lung cancer might have a whole body scan to look for distant metastases and a localised and attenuated scan for assessment of local nodal disease. Such a scan might cost $2000 to $3000. Lung cancer often presents with a nodule or non-specific opacity. It is difficult to differentiate benign from malignant lesions on plain radiographs, CT, and MRI especially before and after surgery or radiotherapy. Other problems include negative results from transbronchial or transthoracic biopsies which are not necessarily reliable and are invasive. It might be more beneficial to obtain a plain radiograph and then a PET scan to work out the metabolic activity of a lesion to differentiate the benign from the malignant. Patients may be compared using a standardised uptake ratio (SUR) to put into perspective FDG uptake, or alternatively scans may be visually interpreted as showing malignancy by displaying increased uptake in comparison to mediastinal uptake. SUR may be thought of as the concentration of radiotracer in the tumour divided by the concentration of injected radiotracer distributed throughout the volume of the body. A study of thirty patients with solitary lung nodules showed PET to have 95% specificity and sensitivity and a highly significant difference between SURs in benign and malignant lesions. However PET is, like other imaging modalities, not 100% accurate. Some lesions which are benign may display high metabolic activity, giving rise to false positives. Some inflammatory lesions or areas of active infection may accumulate FDG eg. TB granulomas and aspergillosis. Figure 4. The CT scan of the chest confirms the presence of a mass in the right chest wall (green arrows). Whole-body FDG-PET confirms hypermetabolism of the mass (red arrows), highly suspicious for recurrence of tumour. Note that the increased activity seen in the arm (yellow arrow) is the injection site; this should not be confused with abnormal uptake. False negatives are also possible with some tumours which have a comparatively low metabolic activity e.g. broncholoalveolar tumours and carcinoid tumours. These also arise if a tumour is small and the resolution is four to eight millimetres, or when the malignant component is a small part of a benign lesion. The 'Prospective Investigation Of PET in Lung Nodules' study (PIOPLIN) showed that there is a lower sensitivity for detecting lesions less than 1.5cm in size - 80% vs. 96% - although a the difference was not deemed to be significant and another study found no actual difference. Comparing transthoracic needle aspiration with PET shows that more malignancies are missed with biopsy and PET results in more unnecessary thoracotomies. The invasive nature of needle biopsies results in 27% of patients requiring a chest drain for pneumothoracies. PET may be helpful in staging bronchogenic carcinoma by TNM. It is limited in use for the 'T' part, being surpassed by CT. For 'N' it has been shown to have similar sensitivity to mediastinoscopy (Van Schil et al, 1989). With the 'M' part, it found 10% of patient to have distant metastases not shown on CT. It thereby upstages patients and affects management. False positives as delineated by CT, e.g. adrenal nodules, are correctly differentiated as negative by PET. Another advantage of PET is that no additional radioactivity is required for a whole body scan (Valk et al, 1995). Therapy of bronchogenic carcinoma is associated with poor outcome. Most disease is seen with an advanced stage and a five-year survival of thirteen percent. Historically, response to chemotherapy and radiotherapy was assessed by tumour size and PET may offer advantages over this. While a partial reduction in FDG uptake is not beneficial, a normalisation of FDG uptake is associated with a good prognosis (Herbert et al, 1996). PET is also capable of diagnosing recurrence early - a notoriously difficult objective to achieve under normal circumstances. Tissue biopsies of scar tissue, although negative, are questionable in value. It also is able to differentiate benign pleural thickening, induced by radiation, from recurrent tumour. The traps to be avoided include mistaking the increased metabolism of inflammatory change induced by radiotherapy for malignancy. Figure 4. 40-year-old male. Fibrohistiocytoma. The CTscan of the chest confirms the presence of a mass in the right chest wall (green arrows). Whole-body FDG-PET confirms hypermetabolism of the mass (red arrows), highly suspicious for recurrence of tumour. Note that the increased activity seen in the arm (yellow arrow) is the injection site,- this should not be confused with abnormal uptake. Cardiac imaging PET may also be useful in coronary artery disease in assessing how limited myocardial perfusion is. It can differentiate reversibly from irreversibly injured ischaemic myocardium. This allows for quantification of the benefit of salvage therapy using angioplasty or thrombolytic therapy (see figure 3). Figure 3. 70 year-old male with a recent anterior M1 and cardiac arrest which was successfully resuscitated. After hospital discharge, the patient was readmitted twice due to episodes of resting pain and ECG changes. Viable LAD territory, with stress-induced ischaemia. Cardiomyopathies may be detected using PET. Regional myocardial perfusion is measured, e.g. compromised in angina. This used to be measured using invasive techniques such a using probes in the coronary sinus to measure flow by thermodilution, or using a doppler probe to measure coronary arterial blood flow. These invasive techniques do not account, however, for a possible difference between the macroscopic flow in the coronary arteries and the capillary perfusion, which may be different e.g. AV shunting. Thallium-201 and 99mTc-labelled isonitriles are single photon emitting radionuclides which have variable degrees of extraction. These differences between patients result in a mismatch between the tracer accumulation in the myocardium and myocardial perfusion. More problems are produced by the difficulties caused by trying to locate single photons. No attenuation is possible and superimposed tissues above and below the area of interest interfere. PET with 13N-ammonia, 82Rb-chloride, and 62Cu-PTS1 is complicated by incomplete extraction which is dependent on flow. A better alternative is to use 15O-water as a tracer. This is used to investigate patients with angina, but with angiographically normal arteries. Limited perfusion in cardiac allografts may also be detected using stress induced by iv dipyridamole. Angiography is invasive in comparison and has difficulty in detecting the diffuse atheromatous change characteristic of this situation. It is also difficult for the patient to detect the symptoms normally associated with angina as the heart is denervated. Dilated cardiomyopathy shows heterogenous diminution of regional accumulation of "C-palmitate diffusely thoughout the heart. This differentiates it from the congestive heart failure picture of segmental diminution. Hypertrophic cardiomyopathy shows a decrease of septal accumulation of FDG. Cardiomyopathy associated with Duchenne's muscular dystrophy may be picked up by the excess accumulation of FDG in segments with decreased perfusion. Other clinical applications PET was initially just used for imaging the brain and heart, but now it is possible to see it being used in a variety of settings. FDG PET can also detect thyroid cancer, breast cancer, lymphoma (see figure 7), squamous cell carcinomas of head and neck, colonic cancer, ovarian carcinoma and musculoskeletal tumours (see figure 5). It can also be used to model and quantify in vivo biochemical processes as radiotracer activity is monitored serially over time. Figure 5. Osteosarcoma. The MRI reveals a destructive lesion of the left posterior femur (green arrow). 18FPET reveals intense uptake of 18F- in the left knee (red arrows), consistent with reactive bone. FDG-PET reveals intense uptake of FDG in both the bone and soft tissue components of the tumour The PETstudies demonstrate that metastases are not likely but that the tumour is restricted to the bone and surrounding soft tissue. It may used to differentiate brain tumours (see figure 6) from radiation necrosis, which can appear similar on CT and MRI with contrast due to the breakdown of the blood brain barrier. Necrosis shows minimal FDG activity, with the tumour showing increase FDG uptake. Figure 6. MRI shows a rim enhancing lesion in the right hemisphere. FDG-PET reveals a hypermetabolic rim (red arrow) consistent with highgrade astrocytoma. Note that such regions of hypermetabolism can be used as a guide for potential biopsy sites. Figure 7. The FDG-PET study reveals the presence of hypermetabolism in the hilar lymph nodes (blue arrow), as well as in the left cervical (green arrow) and right inguinal node (red arrow). Active Hodgkin's disease was confirmed by biopsy of the left cervical node. Preoperative testing of complex partial epilepsy that arises from one of the temporal lobes. Between seizures, the site of seizure origin will have a decrease in glucose metabolism when compared with the normal size. Prostatic carcimoma can be difficult to differentiate from benign prostatic hypertrophy which has high FDG uptake. In the early stages prostatic carcinoma has low FDG uptake (Laubenbacher et al.) and can be masked by intense bladder activity. This may be corrected by hydrating the patient and using a Foley catheter to irrigate the bladder. An ideal tracer has yet to be found but would not be excreted into bladder and colon, and would show up only areas of prostatic carcinoma. The future In the future, cardiomyopathy detection will be improved using this technique and be able to assess response to treatment. Three-dimensional images of hearts, with integration of spatial and temporal indications of perfusion and regional oxygen consumption, will be possible with fluxes in metabolic pathways being shown simultaneously. PET will also be useful in delineating the causes of arrhythmias in cardiomyopathic hearts. Faster computers promise to reduce the time needed to thirty minutes for a whole body scan with better resolution coming through from new types of crystals.