Who Benefits From Surveillance Imaging?

This is the authors' abstract. We don't add key points for this paper.

Routine imaging in remission is performed for patients with many types of cancer in an attempt to detect early, asymptomatic relapse. In patients with lymphoma in the United States, imaging is often performed as routine practice, frequently required in clinical trials, and recommended in treatment guidelines. This is all in the absence of proof that these images improve overall survival or provide other benefits. In Journal of Clinical Oncology (JCO), Voss et al report a trial that included routine surveillance imaging for children and adolescents with Hodgkin’s lymphoma in remission. Most relapses were detected by changes in symptoms, physical findings, or laboratory results; there was no evidence that detection of relapse by routine imaging changed survival. This report by Voss et al addresses an issue that is equally pertinent to adult patients with lymphoma. Because many patients with lymphoma relapsing after complete remission have a chance for a durable second remission with salvage therapy (ie, usually autologous hematopoietic stem-cell transplantation), it is reasonable to consider the possibility that early detection of relapse in asymptomatic patients might increase the chance for cure. To achieve this goal, screening for early relapse with surveillance imaging would need to be accurate and not associated with excessive toxicity or cost. Although the sensitivity and specificity of the imaging technique affect its accuracy, the largest impact is related to the prevalence of the condition in the population being screened. There are many reports of the sensitivity and specificity of computed tomography (CT) scanning in detecting lymphoma. Although these range from well less than 50% to 100%, a reasonable average would be a sensitivity of approximately 60% to 65% and specificity of approximately 90% to 95%. It is more difficult to be certain of the sensitivity and specificity for positron emission tomography (PET)/CT scanning, although it seems that the sensitivity is higher, and the specificity is lower. The most important variable is the prevalence of lymphoma in the population being screened. A study by Radford et al in patients with Hodgkin’s lymphoma in remission found that relapse was detected during one in 68 follow-up visits to an oncologist. In our clinic, the frequency of detecting relapse for patients with diffuse large B-cell lymphoma in remission is between one in 40 to 50 visits. Thus, applying the data from the Radford et al study, the chance that an abnormality on a routine surveillance CT scan would represent lymphoma would be approximately 10%, and applying the data from Nebraska for diffuse large B-cell lymphoma, the chance would be approximately 20%. Of course, these calculations almost certainly overestimate the utility of these images, because in most patients, relapse is detected by obvious changes in clinical status—often in patients returning to clinic between planned visits. The potential value of routine surveillance CT scanning in aggressive non-Hodgkin’s lymphoma, indolent non-Hodgkin’s lymphoma, and Hodgkin’s lymphoma has been evaluated in numerous reports. A report by Liedtke et al found that patients with relapsed diffuse large B-cell lymphoma detected by surveillance CT scan were likely to have low-volume disease, and the lymphoma was likely to be chemosensitive. However, they found only a nonsignificant trend toward better survival. Studies of Hodgkin’s lymphoma, including that reported by Voss et al, have not shown improved treatment outcome with routine imaging using CT scans. There are fewer reports on using surveillance CT scans for follicular lymphoma in remission. This lymphoma typically grows slowly, and asymptomatic relapse might be present for an extended time. Gerlinger et al described 71 patients who achieved complete remission from relapsed follicular lymphoma through autotransplantation and were observed with annual surveillance imaging. Surveillance imaging detected relapse in 16 patients, and routine clinical evaluation found relapse in 18 patients. The relapses detected by surveillance imaging did not require therapy for a longer time compared with the clinical relapses, but overall survival was not affected. The impact of surveillance PET/CT scans on identifying early relapse in patients with lymphoma has also been studied. In general, surveillance PET/CT scanning to detect relapse has a higher falsepositive rate than CT scans and has not had an impact on survival. A report from Denmark described 53 patients with Hodgkin’s lymphoma who underwent 127 surveillance PET/CT scans. The positive predictive value was 19%, and the negative predictive value was 100%. Lambert et al reported using PET/CT scans to screen for early relapse in patients after allogeneic transplantation and found that PET/CT scans in this setting detected relapse earlier than CT scans and allowed for early administration of donor lymphocyte infusions. There are several potential risks of routine surveillance imaging in patients with lymphoma. These include the possibility of inducing a second cancer because of the radiation dose, the risk of biopsies necessary to document relapse after a positive image, and potential anxiety and fear associated with these images, which could lower quality of life. The risk of causing cancer through medical imaging has recently been highlighted; reports suggested that medical imaging might be a significant cause of cancer in the United States. This seems to be a particular concern in young patients. Shenoy et al tried to calculate the risk of subsequent cancers after surveillance images in patients of different ages and sexes. The anticipated incidence of second cancers JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 30 NUMBER 21 JULY 2