Reassessing Imaging Surveillance in Diffuse Large B-Cell Lymphoma

Dr. Brian Link, Professor of Internal Medicine at the University of Iowa’s Holden Comprehensive Cancer Center, spoke with CancerNetwork at the 2013 ASCO meeting about topics in diffuse large B-cell lymphoma (DLBCL). Here he discusses ASCO abstract 8504, a large Mayo Clinic/ University of Iowa study presented by Dr. Carrie Thompson of the Mayo Clinic, which has implications for the role of surveillance imaging in detecting relapses in patients with DLBCL .

Cancer Network: You were a coinvestigator on the study that will be reported later today by Dr. Carrie Thompson from Mayo Clinic, and that’s abstract 8504. Would you please give a brief summary of the study and its findings?

Dr. Link: This study emerges from a database that is a joint database maintained by the University of Iowa and Mayo Clinic on all newly diagnosed lymphoma patients, and it dates back to 2002. What Dr. Thompson did, with substantial help from our statistician Matt Maurer [Matthew J. Maurer, MS], is look at all patients who were diagnosed with de novo diffuse large B-cell lymphoma who were treated with an anthracycline-based immunochemotherapy combination.
And then they wanted to explore patterns of relapse. So they identified 644 patients with diffuse large B-cell lymphoma treated with anthracycline-based immunochemotherapy. Of those, 537 eventually made it through therapy that was deemed successful and had reached a point where they and their physician agreed, “no more therapy, let’s now observe for progression or relapse.” Of those 537, as would be expected, about 20%, or 109 patients, ultimately progressed or relapsed. And Dr. Thompson and Matt very carefully looked at patterns, with a focus on CT imaging, to determine, “How were those relapses detected?”

I think the most interesting findings that they observed were that 68% of patients who relapsed did so and it was detected prior to a scheduled routine surveillance visit, presumably mostly on the basis of a patient identified a concern, notified the care team, and further exploration documented relapse. [A total of] 32% had relapse detected at the time of a routine scheduled surveillance visit. But among those, a very high portion of patients had documented at that visit either an abnormal physical exam finding that would have been suspicious, or gave a history of a symptom that was concerning, or had an elevated level of lactate dehydrogenase level

in the blood. Among patients who had none of those three features, there were only eight patients in whom diffuse large B-cell lymphoma relapse or progression was detected on the basis purely of the CT scan. So, ultimately what I think that suggests is that the symptom process was more dominant in identifying relapses than routine imaging in the absence of clinical findings.

Cancer Network: Patients with DLBCL typically get two or three CT scans per year, but Dr. Thompson said the current study findings are important to help physicians decide how frequently to order scans following treatment of DLBCL, and that these decisions should be tailored to an individual patient. She said she’d like a randomized study to see if clinically directed scans, prompted by patient symptoms, would work better for patients than scheduled scans.
So, what’s your perspective on the impact of the current study results in terms of tailoring surveillance scans for patients with DLBCL?

Dr. Link: As excited as we are to present our abstract here at the ASCO meeting, we have to keep in mind that this is the first step in a process. This data ideally would be validated in another setting, and as Dr. Thompson described, there’s a French cohort of patients that have a similar finding-and those details will be [discussed] further at the international lymphoma meetings at Lugano this summer [12th International Conference on Malignant Lymphoma, June 19–22].

The next step is we have several opportunities to put the statistician to work, to look at further patterns within this set-can we identify subsets of patients for whom CT scans were more useful; perhaps those with high IPIs [International Prognostic Indices], or those who had a unique biomarker, maybe a cell of origin, that identified a pattern of relapse that was more or less susceptible to routine surveillance. Ultimately, once we complete these analyses, we’ll use the peer review process to really focus and clarify our final report. Then, that can go out, and I think we can start making summary conclusions about altering care.

But my initial impression is that, if 68% of the patients are presenting prior to a planned visit, that it really gives us an opportunity to identify early relapses through enhanced communication with patients about symptoms, and you’d like to think in 2013 that if there’s anything we can do to make progress, it ought to be finding ways to communicate easily with our patients, whether that means bringing them into the office to take a history and evaluate them, or reach out to them through traditional or novel methods, to check on them from a symptom standpoint on a regular basis.