Home > Magazine > Exclusive Coverage

Frequent Imaging for Survivors of DLBCL: Does It Improve Survival?

Oct 19, 2015

By Shira Klapper, Senior Writer/Editor

A study out of Denmark and Sweden assessed whether a follow-up protocol including serial imaging made a difference in overall survival (OS) among young survivors (age 18 to 65of diffuse large B-cell lymphoma (DLBCL) who had achieved complete response (CR). The study looked at two cohorts of patients—one from Denmark, where the medical standard is to send patients for serial imaging as part of the two-year, post-CR follow-up care, and the other from Sweden, where the tradition is to send survivors for imaging only after doctors suspect a tumor. The study, “Routine Imaging for Diffuse Large B-Cell Lymphoma in First Complete Remission Does Not Improve Post-Treatment Survival: A Danish–Swedish Population-Based Study,” published online, ahead of print, October 5, 2015 in the Journal of Clinical Oncology (JCO), found no significant difference in survival between the 525 Danish patients and the 696 Swedish patients; In both cohorts, the three-year OS rate was 92%.

Serial imaging as a standard of care for survivors of DLBCL is controversial: According to the article in JCO, there is no conclusive evidence showing that OS is higher among survivors who undergo serial imaging to detect tumors, compared to those whose follow-up care includes check-ups, clinical examination, and blood tests. In addition, imaging exposes all survivors to the potential harmful effects of radiation when less than 20% of survivors treated with combined immunotherapy and chemotherapy will relapse.

“In many institutions, including those in Denmark, we tend to use a lot of routine imaging because we want to detect relapse early,” said study coauthor Tarec Christoffer El-Galaly, MD. “This study is significant because it confirms what many researchers believe right now, which is that when DLBCL is in remission, inclusion of imaging in the follow-up protocol does not seem to improve survival compared to a follow-up protocol relying on check-ups and blood tests.”

The study also found that lower OS was correlated with the following characteristics: age greater than 60 years, elevated lactate dehydrogenase, B symptoms (fever, night sweats, and weight loss), and an Eastern Cooperative Oncology Group performance status of 2 or greater.

Comparing cohorts from two different countries: a methodology that helped avoid selection biases

Dr. El-Galaly said that a previous retrospective study on which he worked showed that survivors of DLBCL who receive serial imaging have a higher OS compared with survivors who receive clinical care only. However, he explained, since that study was retrospective and not a randomized controlled trial, it was not able to control for selection biases, that is, the possibility that differences other than serial imaging might explain variations in OS. For example, in what is known as lead-time bias, patients who received serial imaging might have had their tumors identified at an earlier stage than those not undergoing serial imaging, but for both groups, the time from the onset of cancer until death is the same.

“It’s just a matter of when you start counting survival,” said Dr. El-Galaly.

Yet another bias that affects retrospective studies is length-time bias. In this bias, patients with more aggressive cancers will have shorter asymptomatic disease periods because their cancer is rapidly progressing, which leaves only a short window of opportunity to detect preclinical relapse through routine imaging. Indolent cancers, on the other hand, “are more likely to be detected by random imaging as disease in these patients will go unnoticed and asymptomatic for longer periods,” said Dr. El-Galaly. “So, you may not be comparing the outcome related to detection method—you’re really comparing the outcome of two different diseases, indolent versus aggressive relapse.”

According to Dr. El-Galaly, the design of the current study—which focused on follow-up methods rather than detection methods and used cohorts from two similar Scandinavian countries with different standards of care—allowed it to potentially avoid lead- and length-time biases that affected previous retrospective studies. In fact, he said, “The study, while retrospective and non-controlled with all its potential limitations, is the closest thing to a randomized controlled trial” in that from the outset, the Danish and Swedish cohorts followed two distinct paths, similar to the random design of a clinical trial.

“This study shows us that it’s safe to have a follow-up protocol that doesn’t use imaging,” said Dr. El-Galaly. “This finding should be considered in terms of the cost-effectiveness of different treatments in an environment of limited healthcare resources and should also be considered in terms of avoiding exposing patient to the very high doses of radiation that is delivered by imaging.”


Tarec Christoffer El-Galaly, MD, is a Consultant and a Clinical Associate Professor of Hematology in the Department of Hematology at Aalborg University Hospital, in Aalborg, Denmark.


Abstract of the original JCO article.

PDF of the original JCO article.

El-Galaly TC, Jakobsen LH, Hutchings M, et al. Routine imaging for diffuse large B-cell lymphoma in first complete remission does not improve post-treatment survival: a Danish–Swedish population-based study. J Clin Oncol. Epub 2015 Oct 5. 


The Exclusive Coverage series on ASCO.org highlights selected research from JCO and JOP with additional perspective provided by the lead or corresponding author.

@ 2014 American Society of Clinical Oncology


Back to Top