Oct 18, 2018
This JCO Podcast provides observations and commentary on the JCO article “Impact of Surveillance Imaging Modality on Survival After Recurrence in Patients with Favorable Histology Wilms Tumor: A Report from the Children's Oncology Group” by Mullen et al. My name is Dr. Meredith Irwin, and I am an oncologist and Professor at the Hospital for Sick Children and University of Toronto in Toronto, Canada. My oncologic specialty is pediatric solid tumors.
Wilms tumor is the most common pediatric kidney tumor. With current therapies, the 5-year overall survival for newly diagnosed patients is 90%. For the 15% who relapse, most commonly in the lung or abdomen, the cure rates still often exceed 50%. The likelihood for cure is based on a number of risk factors, including histology, stage and previous therapies. Thus, similar to many pediatric cancer patients, those with Wilms undergo surveillance imaging during and following completion of upfront therapy in order to potentially discover recurrences sooner with the hope that early identification of lower disease burden will translate to higher salvage rates. However, to date, several small studies have failed to demonstrate that early detection of recurrence by surveillance imaging is associated with better prognosis for pediatric cancers including medulloblastoma, lymphoma and neuroblastoma. Moreover, with increasing awareness of the potential risks from imaging-associated ionizing radiation exposures and associated financial costs, there is significant interest in generating evidence to determine optimal surveillance strategies and limit cumulative exposures.
To address these issues for Wilms tumor, the study by Mullen et al was designed to determine whether, compared to relapses detected by surveillance with chest x-ray and ultrasound, those detected with cross-sectional CT imaging were associated with improved survival rates. To do this, four authors retrospectively reviewed flowsheets and imaging reports for patients who recurred on the 5th National Wilms Tumor Study (NWTS-5) between 1995 and 2002 to determine whether their relapses were identified by signs/symptoms or scheduled surveillance imaging. If by imaging, they recorded the modality- CT, ultrasound and/or chest x-ray. The study cohort included the 281 patients who recurred following initial diagnosis of favorable histology unilateral Wilms. All patients underwent imaging at specified time intervals, but the modalities varied.
The 5-year overall survival for this cohort post-relapse was 67%, which is similar to other studies. 48% of recurrences were detected with routine surveillance by chest x-ray or ultrasound, 25% by surveillance with CT, and 25% presented with symptoms between scheduled imaging, most commonly pain or abdominal mass. Certain characteristics differed among these groups. The CT-detected patients were more likely to be stage 4, and in the symptoms group, more relapses were extrapulmonary and occurred post-treatment. The authors asked whether the specific method of detection of relapse impacted prognosis. Although recurrences detected by symptoms were associated with an inferior 5-year survival of 55%, versus 76% for the imaging-detected group, there were no differences in survival for patients based on the imaging modality used for detection at any point during or following therapy. There was also no significant advantage to CT over u/s or x-ray for abdominal and chest relapses, respectively. Although a higher number of foci detected at recurrence and larger relapse size correlated with inferior outcomes, analyses failed to find a significant advantage for CT use in the subgroups of lung-only relapses or advanced-stage patients.
The authors also examined financial burden and radiation exposures. To determine the economic impact, the authors predicted costs based on US Medicaid/Medicare reimbursement rates. For patients with stage III disease, more than 200 imaging studies (with estimated costs between $20,000 and $45,000) were required to identify one relapse. For stage IV, between 158 and 190 studies costing more than $15,000 were needed. Estimated radiation exposures from these surveillance protocols varied between 9.4 and 83 mSv, depending on the total number of CTs. Although it is difficult to quantify how cumulative radiation doses may impact future cancer risks, expert groups have estimated that there is a 1 in 1,000 increased risk of future cancer deaths for every 10mSv exposure at age 10, and these risks are higher for patients exposed at younger ages. Other potential negative consequences of surveillance imaging not examined by these authors include the associated psychological stress for families, which is termed “scanxiety,” and possible adverse neurological effects of anesthetic agents for young children.
The significance of this report is that it is the first study to compare the utility and cost of different imaging methods for the identification of Wilms tumor recurrences. In comparison to x-rays or ultrasounds, there was no advantage to using CT, which is costlier and resulted in higher radiation exposures. It is possible that an advantage to CT was not identified in part because only the best prognosis patients, those with favourable histology, were included. The authors excluded those with bilateral disease, who often have germline predisposition syndromes, and patients with anaplastic histology, which has a higher relapse rate
Ideally, a prospective RCT design would be performed to determine whether cross-sectional imaging would be superior to ultrasounds/x-ray combinations. However, given the overall excellent survival for Wilms tumor, the number of patients required to detect a significant survival advantage would likely be very large. In addition, given the cost and potential late effects of radiation exposures, engagement from oncologists, radiologists and families might be challenging.
In conclusion, the findings in this manuscript support the recommendations by Mullen and colleagues that surveillance post completion of therapy for favorable histology unilateral Wilms does not need to include CT scans and can instead be based on symptoms and ultrasound and/or x-rays. These authors did not specify whether duration of surveillance could be limited; however, a recent publication in Lancet Oncology by Brok and colleagues supported consideration of a cut-off of two years. Their SIOP study reported that the detection of one relapse 2-5 years post-therapy required 500 scans. Importantly, it remains possible that for rare patient subsets with anaplastic histologies or other biomarkers of unfavorable disease, which may include 1q LOH, surveillance protocols might need to be adjusted. This may be similar to strategies for other tumors, such as neuroblastoma, where, in comparison to patients with high-risk disease, those with low or intermediate-risk disease with survival rates of more than 80-90% are increasingly undergoing surveillance regimens without frequent CT scans, which, in part, is determined based on clinical and biological risk factors. These approaches to risk stratification of imaging are increasingly being used. Thus, clinicians may begin to modify surveillance based on the risk of recurrence for a particular subgroup of patients, similar to precision medicine approaches for treatment. However, it will be important to perform studies similar to the accompanying manuscript to generate evidence to inform and support precision surveillance imaging recommendations that are specific to different tumor types.
This concludes this JCO podcast. Thank you for listening.