DLBCL Risk Assessment and Genetic Signatures

Andre Goy, MD: We know large cell lymphoma is the most common lymphoma, about one-third of the cases. We know that over 50% of patients can be cured with R-CHOP [rituximab, cyclophosphamide, hydroxydaunorubicin hydrochloride, vincristine, prednisone] chemotherapy, about 20% to 30% relapse and 10% to 15% suffer primary failures. The real challenge is that these failures occur in 80% of the cases—80% of these relapses occur in the first 18 months, and they usually are very challenging to administer salvage therapy, even with high-dose therapy and transplant. Therefore, this is very important to try to have markers that help us identify, prognosticate, and stratify patients.

I would say there are 3 types of factors, and they are the lymphoma factors: We know large cell lymphoma is a very heterogeneous entity and primarily large cell lymphoma, for example, is a separate subtype. But under the NOS [not otherwise specified] subtype of large cell lymphoma, there’s a different molecular subtype, the non–GC [germinal center]. But well beyond that, now we know that 7 factors are very important, unique BCL2, double expresser, double-hit lymphoma. Beyond the FISH [fluorescence in situ hybridization]—typically the definition is by FISH—there’s about 50% of these high-grade double-hit lymphoma that have a rare, poor prognosis with standard R-CHOP [rituximab, cyclophosphamide, hydroxydaunorubicin hydrochloride, vincristine, prednisone] that have the signature of a high-grade lymphoma but do not have the FISH by gene expression profiling. TP53 next-generation sequencing is not done yet in routine. But we try to narrow down a molecular subtype.

Then there’s the tumor burden of the lymphoma, and the stage, the LDH [lactate dehydrogenase], bulky, and the baseline metabolic volume if you want to go by PET [positron emission tomography] scan. The second aspect that was developed is a prognostic model trying to put this in context, and we have revised IPI [ipilimumab] in the RQ era. We have the NCCN [National Comprehensive Cancer Center] IPI [ipilimumab] guidelines trying to use the same factors basically, the basic factors of the IPI [ipilimumab], but 25 is better, the age and the LDH and try to identify better smaller subset, particularly on the extreme, the very good-risk and the very high-risk patients.

There is something that is interesting called the PFS [progression-free survival] 24—no failure after 2 years; therefore the PFS 24 name. These patients do very well, similar to the control population that never have lymphoma. To identify these factors, there’s the IPI [ipilimumab] 24. There’s a calculator available online that also uses similar factors where it adds gender, bulky disease, and lymphocyte count at baseline. There’s another factor to take into account the diagnostical delay to start treatment versus immediate treatment, which obviously the patient started within 2 weeks. Those who start later usually have more symptoms.

That’s really important because these factors that are clinical factors are centrally based on the characteristics of each patient that have been very helpful, but not really to predict the long-term outcome or to predict the response to a therapy. There are some host factors that we heard about regarding gender. Comorbidities we’ll talk more in a minute, but there are a number of factors not yet used in practice: the class II MHC [major histocompatibility complex] expression; the beta-2 microglobulin, which is a class I molecule; microenvironment signature inflammation; cytokine; vitamin D. This is obviously growing, and those are more the host factor, and this again predicts for DFS [disease-free survival], PFS, and survival but not for the early failures, which is the real problem that I mentioned before.

Finally, the response to therapy is the PET scan. And the interim PET scan, although it is debated, when it’s a negative PET regardless of the stage, even in advanced stage, this is usually a very good outcome. Sometimes there’s nonspecific inflammation and residual gray zone like PET scan with low SUV [standardized update value], and what matters is the delta, the drop of the SUV. Typically in routine what we really look at is the patient who has a negative PET scan at the end of treatment. That was very predictive.

Here I just want to mention that when someone has a PET scan that is positive, it’s important to biopsy this patient. We need to know if it’s truly active disease because it has some hot spots. Finally I think there is another factor emerging, and that will take more role in the future on prognostic factor, the DNA, or cell-free DNA that is detectable in all patients at baseline, regardless of the stage. It reflects the tumor genetic makeup or signature profile if you want. Studies, not large studies but very impressive, show that patients who are PET negative and cell-free DNA negative after 3 cycles do remarkably well.

This is a number of factors. How do we use this in practice, so in a frontline setting, R-CHOP [rituximab, cyclophosphamide, hydroxydaunorubicin hydrochloride, vincristine, prednisone] has remained the default, but it’s not sufficient for high-risk patient, double-hit, and double expresser, and non-GC lymphoma. But it’s important to understand that we can start to adjust with the treatment. In a non-GC subtype, so far the attempts to do R-CHOP [rituximab, cyclophosphamide, hydroxydaunorubicin hydrochloride, vincristine, prednisone] have not really panned out. Patients, about one-third of these non-GC double expresser, and we know that they do very poorly with R-CHOP [rituximab, cyclophosphamide, hydroxydaunorubicin hydrochloride, vincristine, prednisone], so we try to give them something more intensive. Finally, in the high-grade lymphoma that I mentioned, these guys do very poorly with R-CHOP [rituximab, cyclophosphamide, hydroxydaunorubicin hydrochloride, vincristine, prednisone], and there’s definitely no standard therapy. We tend to use, if you’re a younger patient, the EPOCH [etoposide, prednisone, vincristine, cyclophosphamide, doxorubicin hydrochloride] regimen. That’s the impact, in a nutshell, of these prognostic factors in lymphoma.

Julio Chavez, MD, MS: The question is, what is the prognostic and predictive significance of DLBCL [diffuse large B-cell lymphoma] genetic signatures The management of DLBCL has evolved over the years. While next-generation sequencing has shown valuable information, there is an absence of a specific drug or medication that will target this specific gene signature. Its application in routine clinical practice remains elusive, in my opinion. However, it underscores the fact that there’s a need of developing trials using this genetic testing. One of the challenges I see using these tools, like next-generation sequencing, is the cost, the availability, and especially reproducibility, that it will be difficult for physicians who are not familiar with this testing to maintain their patients.

Transcript Edited for Clarity