The Clinical Dilemma of Incidental Findings on the Low-Resolution CT Images from SPECT/CT MPI Studies

DEFINITIONS AND FREQUENCY OF INCIDENTAL FINDINGS

An incidental finding can be defined as an unsuspected finding that is not related to the clinical reason for performing the diagnostic test (8,9). Examples of incidental findings that can be identified on attenuation-correction CT images of the chest include coronary vessel calcification, vascular anomalies, valve replacement, pacemakers, atelectasis, effusion, consolidation, lobar collapse, nodules, masses, pulmonary metastases, ground glass opacities, and aortic aneurysm (20).

Incidental findings may or may not already be known from previous diagnostic tests. Findings that are already known to the clinician are likely to have a management strategy in place and therefore often do not require further investigation. Previously unknown incidental findings are termed new incidental findings and are likely to fall into 1 of 3 categories: clinically significant, clinically insignificant, or indeterminate. Clinically significant incidental findings are highly suspected of having an underlying pathology that may negatively affect patient wellbeing, requiring further investigation (9).

The frequency of incidental findings on attenuation-correction CT images is particularly high within the thorax (10), possibly because of the inherent contrast resolution and low attenuation of the x-ray beam through this area. Lung cancer shares comorbidities with heart disease—the clinical reason that these patients are being scanned. It is therefore not surprising that extracardiac pathology is frequently detected.

The high number of incidental findings identified on attenuation-correction CT images causes a dilemma. The acquisition is intended for attenuation correction alone and not for evaluation, and the relatively low resolution of the images means that characterization of findings is often not possible. Additionally, unlike diagnostic-quality CT images of the chest and abdomen, attenuation-correction images are likely to have breathing artifacts because patients are not required to hold their breath. The craniocaudal range of the acquisition is also limited to the cardiac area only; detection of incidental findings is therefore also restricted.

INCIDENTAL FINDINGS ON SPECT/CT MPI

The number of incidental findings on the attenuation-correction CT images produced during SPECT/CT MPI is noteworthy, and the percentage of clinically significant and indeterminate findings can often be as high as 10% (10).

A high proportion of findings are pulmonary nodules (11). Although most lung nodules are benign, they have the potential to develop into lung cancer (12). Attenuation-correction CT images, which often reveal lung nodules as well as other pathologies, were never intended for radiologic reporting but there is growing evidence to suggest that they should be (10).

Lung cancer is the leading cause of death in most countries, including the United States (12,13). Early lung cancer is often asymptomatic, with symptoms usually arising only when the disease is at an advanced stage. The prognosis for lung cancer is consequently poor (13). However, the prognosis for non–small cell lung cancer improves significantly if it is detected at an early stage when surgical resection is possible (12,13). This is not the case with small cell lung cancer. In the management of pulmonary nodules, the recommendations that have been developed from lung cancer screening trials suggest a necessary balance between early intervention to reduce mortality from lung cancer and the risk of early morbidity and mortality from intervention due to false-positive findings and overdiagnosis (11).

The high prevalence of lung nodules on CT of the chest is of particular concern during SPECT/CT MPI because any such findings will be incidental and may signpost a patient to an alternative or additional-care pathway. Although there is an argument for early treatment of lung cancer, the detection of lung nodules on attenuation-correction CT images does not necessarily result in reduced mortality from lung cancer. A high proportion of lung nodules are ultimately found to be benign, and the patient may not benefit, and may actually be harmed, from follow-up of these nodules.

This sentiment is echoed by the outcomes of lung cancer screening trials; for a screening program to be effective, mortality must be reduced and the benefits to the patient must outweigh the risks (12). The National Lung Screening Trial has demonstrated that it is possible to reduce mortality from lung cancer using low-dose CT screening in place of chest radiography and sputum tests (14). However, in doing so, there is a high rate of false-positive findings and overdiagnosis, as well as an increased radiation burden to the patient (15). Overdiagnosis of indolent tumors that would not have become symptomatic during the patient’s lifetime have cost implications as well as raising patient anxiety and morbidity in the same way that false-positive findings might (16). Consequently, several organizations have decided not to implement screening programs. Where screening programs have been implemented, they are focused specifically on the population at high risk (17).

These considerations add to the dilemma of whether clinicians should review the attenuation-correction images. Reviewing them and providing a report may possibly enable early diagnosis of pathology and potentially a better prognosis. However, identifying incidental findings may increase risk to the patient without necessarily providing any benefit. Further diagnostic examinations are often associated with an increase in ionizing radiation dose and possibly invasive procedures that may carry physical risk and cause psychologic harm to the patient (9,18,19). Furthermore, only a small percentage of these incidental findings are significant at definitive diagnosis (9,20,21).

CONTRAST ISSUES

There is often poor inherent contrast resolution between organs and soft tissues within the body due to relatively small differences in density. The use of contrast agents artificially improves contrast resolution between soft-tissue structures. The iodination of the contrast agent increases the density within the organs and so will affect the extent to which the x-ray beam is attenuated. Different organs take up contrast at different rates, resulting in an increase in contrast resolution on the CT images. Similarly, pathologic tissue can demonstrate selective uptake of contrast agents; tumors that are highly vascular tend to enhance avidly whereas ischemic tissue tends not to enhance. This selective uptake improves demonstration of pathology in relation to normal tissue and aids characterization of the pathology. For this reason, it is unusual to perform diagnostic CT of the chest, abdomen, and pelvis without the use of an intravenous contrast agent.

The use of intravenous contrast agents with SPECT/CT is not without problems. Contrast agents will attenuate γ-rays in the same way that they attenuate x-rays. The result is degradation of the quality of the SPECT images. Contraindications relating to the use of iodinated intravenous contrast agents also exist.

It is because of the acquisition parameters used, along with the lack of an intravenous contrast agent to enhance contrast resolution, that attenuation-correction CT images are not considered diagnostic.

IMAGE-QUALITY ISSUES

The quality of attenuation-correction CT images is directly related to how the images have been acquired, typically with a low tube current (mA) and a wide slice. The resulting low-resolution CT images enable attenuation correction to be performed. The long acquisition time results in motion artifacts from breathing. The nature of incidental findings can be difficult to determine, resulting in a high number of false-positive findings.

As technology has progressed, multidetector CT scanners have been utilized in some hybrid systems. Multidetector CT offers increased technologic capabilities with the potential to produce images superior in quality to the attenuation-correction images typically associated with earlier systems. The potential to acquire narrow slices, reduce imaging time, and increase the tube current has provided the option of improving image quality to a level that, in some cases, is comparable to that of diagnostic CT. Having this option has inevitably led to differences in the quality of the attenuation-correction images acquired in different departments, related not only to the capability of the scanner but also to the way the parameters have been optimized (20). Ethical considerations arise when CT image quality is improved in this way, because there will be an inevitable increase in ionizing radiation dose to the patient in exchange for an uncertain gain.

PSYCHOLOGIC EFFECTS

False-positive incidental findings or findings that result in overdiagnosis can increase patient anxiety. Although diagnosis of malignancies at an early stage can potentially reduce patient mortality, only a very small percentage of findings on attenuation-correction CT images have been found to be malignant or detrimental to the patient at definitive diagnosis (28).

There is only limited literature surrounding the psychologic effects on patients who learn of incidental SPECT/CT findings. There is, however, an abundance of literature relating to the psychologic effect of a recent diagnosis of cancer, or the wait for a definitive diagnosis, and some parallels can be drawn to patients with incidental findings.

Patients who have incidental findings on a screening examination or are being imaged because of suspected cancer will usually have a care pathway and support structure in place, typically involving nursing, medical, and other professionals who are available to help patients and their families. This is often not the case for patients who have incidental findings on SPECT/CT. Such patients will receive limited support because of the unexpected nature of the findings.

Pulmonary nodules are one of the most frequent incidental findings on SPECT/CT of the thorax (19,29). These nodules can be single or multiple and can be of unknown significance. Most are benign, but if indeterminate the patient will require surveillance that may continue for 2–3 y (30). Almost all patients who are told they have a pulmonary nodule assume they have cancer (16). Just raising the possibility of cancer can be threatening (17) and can lead to distress until, and possibly beyond, definitive diagnosis (18).

Pulmonary nodules are of particular clinical relevance (8) because they do have the potential to become malignant, although most will be benign (31). Further management of pulmonary nodules is based on their size, with the likelihood of malignancy increasing with the size of the nodule (32,33). The indeterminate nature of pulmonary nodules frequently leads to follow-up CT imaging to monitor any change in the size of the nodule. If its size remains stable over a period of 2 y, it is considered benign. This lengthy period of repeated imaging inevitably increases the radiation burden to the patient along with the anxiety level. Biopsy of pulmonary nodules, which would give a more definitive answer, is often not possible because of their small size and location within the lung (32). Despite reassurance, the patients may believe they have cancer in the absence of a definitive diagnosis.

Lung cancer screening involves harm as well as benefit, and in some circumstances the harm can outweigh the benefit (33). This also applies to patients with incidental findings on attenuation-correction CT images. The harm from overtreatment of false-positive findings should be kept in mind when considering the benefits of reviewing these images (34). Although patients may gain some reassurance from a negative lung cancer screening, a negative attenuation-correction CT scan from SPECT/CT MPI cannot be equally reassuring because of the limited range of chest MPI. A normal attenuation-correction scan does not exclude pathology that falls outside the limited range of the scan. In addition, early detection of lung cancer does not necessarily mean an improved outcome. Approximately 25%–30% of patients will present with potentially curable disease (35), but lung cancers develop quickly and metastasize early, often leading to a poor prognosis (19). Metastases often are present at the time of initial presentation. This frequently poor prognosis also needs to be balanced against the high rate of benign pathology detected on attenuation-correction CT (19).

The psychologic effects on patients who are diagnosed with false-positive findings can discourage patients from attending screening or diagnostic procedures in the future (33). If they have had a significant misdiagnosis, they are likely to lose trust in diagnostic procedures. Psychologic effects can be transient or more persistent. At initial diagnosis the patient is likely to experience anxiety, but over time this tends to develop into depression. Both anxiety and depression can persist before and after treatment (35).

FINANCIAL ISSUES

The financial impact of routinely reporting the findings of attenuation-correction CT, as well as the time required, needs to be considered. A radiologic report takes time to construct and, in the case of clinically significant findings, is likely to affect future management of the patient. The financial impact is justifiable when there is a net patient benefit, but such does not appear to be the case for attenuation-correction CT images, with their high yield of false-positive findings. Similarly, the follow-up examinations that may be needed to reach a definitive diagnosis have a cost and require time yet will not necessarily benefit the patient and may, in fact, be a potential cause of harm (36). Thus, the cost-effectiveness of reporting attenuation-correction CT findings is brought into question, along with whether such reporting will benefit or harm the patient.

DISCLOSURE

No potential conflict of interest relevant to this article was reported.