Who Should-or Should Not-Receive RT for DLBCL?

December 15, 2014

As we examine the question of which patients with DLBCL do not need RT, the first step must be to confine our review to patients who have received optimal chemotherapy.

In their nicely assembled article, “Improving Outcomes in Advanced DLBCL: Systemic Approaches and Radiotherapy,” Dr. Boyle and coauthors discuss a recurring topic in lymphoma management: What is the role of radiation therapy (RT) in the management of diffuse large B-cell lymphoma (DLBCL)?[1] They conclude that RT has much to offer and should be routinely included in standard management but immediately qualify that recommendation by indicating that RT is most appropriately used only for “select” patients. Thus, the real task is to define who, among the entire set of patients with DLBCL, are the better candidates for RT. An examination of the data reviewed by the authors provides quite helpful guidance, but when I consider this evidence I find it more helpful in addressing the question of who should not receive RT. Let us consider the relevant information with that goal in mind.

As we examine the question of which patients with DLBCL do not need RT, the first step must be to confine our review to patients who have received optimal chemotherapy, which is R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone), as noted by Boyle and coauthors. DLBCL is a systemic disease, even when the presentation appears localized, so the core ingredient in successful treatment must be optimized chemotherapy, currently R-CHOP. The next step in our deliberations is to recognize the value of functional imaging, which has transformed lymphoma management as profoundly as any other single advance over the past 2 to 3 decades. After much searching, we finally have a genuinely useful way to assess the completeness of response: A negative fluorodeoxyglucose positron emission tomography (PET) scan equals a complete response. This valuable tool allows us to determine whether chemotherapy with R-CHOP has eliminated detectable disease when treating patients with either limited-stage or advanced-stage DLBCL. Fortunately, a negative PET scan, now best defined as a score of 1 or 2 using the so-called Deauville scoring system,[2] strongly and reproducibly predicts a very high likelihood of cure, best defined as failure to relapse within the 2 to 3 years following treatment. Most studies have found this negative predictive value to be about 85% to 90% for DLBCL.

The final aspect of our analysis is a harsh truth: Patients with DLBCL who are not cured by their primary treatment are very unlikely to be cured at all. This painful reality is brought home by the closeness with which progression-free survival approximates overall survival in clinical trials and institutional experiences. Let us examine the information provided by Boyle et al in light of these three observations-use of R-CHOP, post-chemotherapy assessment with PET scan, and lack of broadly curative secondary treatment. We have two scenarios to examine and will begin with the more common one: a PET-negative complete response (CR). The authors cite several studies relevant to this clinical setting in which addition of RT appeared to improve outcome. However, closer inspection reveals that almost all of this apparent improvement is seen in progression-free or event-free survival and does not carry through to overall survival. This tells us that RT is effective in providing local control and therefore preventing DLBCL from relapsing in its presenting sites; however, given that DLBCL is fundamentally a systemic disease, local control, which prevents local and earlier relapse, does not translate into cure because unirradiated disease elsewhere in the body eventually emerges. The most reasonable conclusion is that, in terms of cure, not just local control, RT has little to add when chemotherapy has achieved a CR.

It is more challenging to interpret the other patient scenario: PET-positive residual disease after primary therapy with R-CHOP. Some of these patients have a positive PET scan because of continued metabolically active healing or persistent inflammation and will gain nothing from RT; however, some genuinely harbor residual lymphoma solely in a site that can be fully included in a reasonably configured and dosed involved field of RT. Now we have a situation in which RT may play a decisive role: eradication of isolated local residual lymphoma. To accomplish this, we must accept that some patients with a falsely positive scan will be unnecessarily irradiated, but the potential gain for those with truly localized residual lymphoma justifies this modest over-treatment.

Ultimately, the most important aspect of the analysis by Boyle et al rests on their assertion that RT can be quite useful for selected patients with DLBCL. Fortunately, functional imaging with PET scanning provides strong guidance for identifying patients who do not need RT-those with a PET-negative CR; and helps to isolate the minority who may profit from RT-those with a positive PET scan with fluorodeoxyglucose uptake confined to a reasonably configured involved field. This is a major step forward, allowing us to spare many patients from unnecessary irradiation and to confine its use to the special minority for whom RT provides a good balance of efficacy and toxicity.

Financial Disclosure:The author has no significant financial interest or other relationship with the manufacturers of any products or providers of any service mentioned in this article.


1. Boyle J, Beaven AW, Diehl LF, et al. Improving outcomes in advanced DLBCL: systemic approaches and radiotherapy. Oncology (Williston Park). 2014;28:1074-84.

2. Meignan M, Gallamini A, Haioun C, Polliack A. Report on the Second International Workshop on interim positron emission tomography in lymphoma held in Menton, France, 8-9 April 2010: Deauville criteria. Leuk Lymphoma. 2010;51:2171-80.