Anas Younes, MD: There are multiple prognostic factors used to stratify patients, mainly for clinical trials and to provide some sense of prognosis for patients with follicular lymphomas. The initial one in the pre-rituximab era was FLIPI, and this was followed by what we call FLIPI-2 in the rituximab era. I should point out that these prognostic factor models do not guide therapy outside of clinical trials. So, whether your score is good, intermediate, or bad doesn’t tell you what should you use for treatment, but they are good for stratification in the context of clinical trials.

Bruce Cheson, MD: Follicular lymphoma is a very heterogeneous disease, and what we’ve learned over the last few years is that there are several time points at which we can predict patients who are likely to do well versus those who are not likely to do so well.

For example, there’s CR30—patients who remain in remission at 30 months—but that’s waiting 2-and-a-half years. Then we found from several studies that if a patient is in remission at 2 years and perhaps even 1 year, their survival is comparable to that of an age-match population without lymphoma. That accounts for about 80% of patients, but it’s the other 20% of patients who have an awful outcome. Their survival is just a few years, and clinical trials are now focusing on that specific group of patients because of their poor survival. But that’s still waiting 2 years or 1 year. If we’re going to affect improvement in patients with follicular lymphoma who are not likely to do so well, we need to bring this closer to treatment. There are a couple of tests we can do following treatment, immediately after therapy, looking at minimal residual disease and at PET scans, and these predict outcome. But then again, at that point, the patient has already been treated with a potentially ineffective therapy. So, where we need to go is looking at predictive markers before treatment. Several of these methods have been evaluated. Dr. Younes talked about m7-FLIPI, for example. There’s also the metabolic tumor volume using PET scans, which also seem to predict outcomes, particularly in combination with the FLIPI-2 score.

Incorporation of patient comorbidities is important in treatment planning. There are some drugs—for example, anthracyclines—which cannot be given to patients with reduced heart function. There are other drugs, such as vinca alkaloids, that you can’t give to patients who have peripheral neuropathy.

Now in the era of targeted therapies, comorbidities become somewhat less relevant, although not totally. For example, idelalisib, the PI3 kinase inhibitor, was approved, at least in CLL, for patients who are not suitable candidates for chemotherapy because of comorbidities—in patients who have abnormal renal function, abnormal bone marrow function, or reduced performance status. So, we still need to recognize the specific toxicities of a drug and tailor our therapies so that we don’t risk a patient getting worse because we gave them a drug with a specific toxicity. But the targeted therapies are much safer than some of the more standard toxic chemotherapies.

Transcript Edited for Clarity