At the 64^th Annual ASH Meeting, Kathryn Tringale, MD, MAS, Memorial Sloan Kettering Cancer Center, New York, NY, presented her study on patients with primary CNS lymphoma, which sought to understand survival outcomes after consolidation regimens and analyze patterns of disease failure among over 500 patients treated for primary CNS lymphoma at her institution.

Transcript:

My name is Kathryn Tringale. I'm a chief resident in radiation oncology at Memorial Sloan Kettering Cancer Center [MSK]. At ASH this year, I presented our work on the outcomes after treatment for primary CNS lymphoma in over 500 patients treated at our institution.

So really, the background of our study was to understand the survival outcomes after consolidation regimens and also look at the patterns of disease failure in patients with primary CNS lymphoma.

Primary CNS lymphoma is typically treated with a high-dose methotrexate-containing regimen as induction therapy, and then after that, there's a consolidation regimen. Currently, the optimal consolidation regimen is not clear for patients, especially when we think about individual patient characteristics. There are different options that we can use, like stem cell transplant, or whole-brain radiation therapy, or even chemotherapy alone.

In more recent years, we've seen a shift away from whole-brain radiation therapy of doses around 45 gray, given there are concerns about neurotoxicity that were seen in several prospective trials. So, we've shifted away from that dose to lower doses of radiation, more like 23.4 gray in 13 fractions, or even away from radiation altogether, just with chemotherapy or transplant.

So, the purpose of our study was really to evaluate the outcomes with these different regimens and compare them. Moreover, this disease is really thought to be multifocal in nature, or sometimes even disseminated throughout the entire central nervous system. So, we wanted to see, in a modern cohort of patients in a large series, whether there actually are patients who have local disease relapse, meaning involving the original site of disease or adjacent to that original site of T1 post–contrast-enhancing disease.

To do this, we looked retrospectively at patients at Memorial Sloan Kettering treated from 1983 until 2020. We looked at 559 patients, ultimately, when we limited our cohort to those with diffuse large B-cell histology, and those with adequate treatment and follow-up information.

When we looked at our cohort of 559 patients, we looked at their patient demographics and characteristics, including the RPA, which stands for “recursive partitioning analysis.” It's an MSK-derived prognostic model incorporating both age and performance status. An RPA of 1, those patients have the best prognosis. They're patients age less than 50. Then there's patients who are RPA 3. Those are patients who have a high age but a low performance status.

We looked at that, and we also looked at the induction regimens that were used. And 95% of patients actually received methotrexate-containing reduction regimens. So they were homogeneously treated. Fortunately, we also had a very long follow-up, a median follow-up of over 7 years.

So when we looked at our original, the baseline presentation of these patients, on their MRI, they actually were more likely to have multifocal disease, to have supratentorial disease. Fewer patients had deep structure involvement.

Our first question, as I had mentioned before, was really to look at the outcomes by consolidation regimen. So, we looked at the 351 patients who had a complete response to consolidation, and we looked at whether they failed down the line, and if they did fail, we looked at those sites of failures.

When we looked at the sites of failures, we see that there's no clear relationship between those baseline MRI characteristics and where they failed. That's to say we really can't predict where these patients are going to fail later on down the line, in patients who achieve a complete response to consolidation.

Then when we look at their outcomes by consolidation regimens, we see that there's no statistically significant difference in the different consolidation regimens when we account for patient characteristics like the RPA and like induction regimen in response to induction.

But interestingly, when we look at RPA 1 patients, those are the best prognostic group. Those patients actually had very good outcomes to both reduced-dose, whole-brain radiation and stem cell transplant, with many of these patients being cured, which is very exciting for this disease.

Then, interestingly, when we look at RPA 3 patients, those are the patients who are older with poor performance status, they actually do relatively okay with reduced-dose, whole-brain radiation. We did have patients who had overall survival beyond 10 years even after their treatment. That's very encouraging.

Then, interestingly, we also wanted to look at patients who had early disease failure, so not the patients who had a complete response to consolidation and failed later on, but we looked at patients who actually had an early failure, as in refractory on induction or failed within 3 months of induction.

We saw that those patients were actually significantly more likely to fail locally, at the original site of disease, than outside that original site. This is really paradigm shifting for this disease that has historically been considered to be multifocal and disseminated in nature.

We are very excited to see this result of these early failures being local because it does introduce the opportunity to consider a local therapy, like involved-site radiation therapy, as opposed to targeting the whole brain with radiation. If we're able to use involved-site radiation, possibly as a bridge to some of these new exciting therapies like cellular therapies, it'll be really exciting to be able to focus the radiation on this site, to give a good response so that they can go along to more curative treatment options.

That's one of our most exciting findings. And then the other is really to highlight that even though we've shifted away from the use of whole-brain radiation, the reduced-dose, whole-brain radiation results are very encouraging, and they support the preliminary results published on RTOG 1114, which looked at the use of reduced-dose, whole-brain radiation with chemotherapy, and they showed no statistically significant differences in terms of neurotoxicity but did show improved progression-free survival with the use of reduced-dose, whole-brain radiation.

We're excited to really think about using more local therapies as a bridging strategy for these early failures, and also for us to reevaluate the role of reduced-dose, whole-brain radiation, especially when we're starting to consider cost to the health care system. Although our study did not have this data, it's interesting to consider doing a cost-effectiveness type analysis, comparing the cost of reduced-dose, whole-brain radiation with stem cell transplant or some of these other more involved cellular therapies.