Defining high-risk disease has been a bit of a moving target. We have traditional classifiers such as the International Prognostic Index (IPI) and cell of origin, and now we have molecular and genetic classifiers that indicate double-hit biology (ie, translocations of MYC and either BCL-2 or BCL-6). When MYC is translocated with an immunoglobulin gene partner, a worse prognosis is conferred. Some patients have a double-hit gene signature without the classically identified translocations by fluorescence in situ hybridization (FISH), and the use of double-hit genetic signatures may offer more precision than using translocation data from FISH alone. There are various classification schemata that are in development that move away from cell of origin, and these may also help inform the development of precision-targeted therapy in the treatment of either newly diagnosed or relapsed/refractory high-risk disease.

The management of double-hit lymphoma with MYC translocated to an immunoglobulin partner is an area of ongoing debate. To date, prospective randomized data to help guide decision making have been very limited. Some retrospective data have suggested that intensified induction strategies such as dose-adjusted etoposide, prednisone, vincristine sulfate, cyclophosphamide, doxorubicin hydrochloride, and rituximab (EPOCH-R regimen) may improve outcomes compared to rituximab plus cyclophosphamide, doxorubicin hydrochloride, vincristine sulfate, and prednisone (R-CHOP regimen), but a recent abstract using data from the Flatiron Health electronic health record database found no difference in survival between these regimens in patients with MYC-rearranged double- or triple-hit DLBCL. 

There are several clinical trials that have been completed, are currently accruing, or are in development, each of which focuses on a different way to improve outcomes in patients with higher-risk DLBCL. These often use the IPI to create a high-risk disease category and may incorporate biologically agnostic treatments. Early nonresponse to treatment is another indicator of high-risk disease, but identifying those early nonresponders is a challenge. However, we may be able to use the rate of disappearance of cell-free DNA or circulating tumor DNA after treatment initiation to predict prognosis and to help guide response-adaptive approaches in the future.

Risk can be assessed by clinical features or by biologic markers. We know that patients with higher IPI scores are less likely to be cured with standard chemoimmunotherapy, whereas those with an IPI of 0 or 1 have very high cure rates with R-CHOP. Thus, some of the ongoing trials are looking to make improvements in the treatment of high-risk populations. 

For example, the ongoing POLARIX trial (NCT03274492) is comparing polatuzumab vedotin with rituximab, cyclophosphamide, doxorubicin hydrochloride, and prednisone (R-CHP) vs standard R-CHOP in patients with an IPI score of 2 or higher. Polatuzumab vedotin plus R-CHP is a modified version of R-CHOP, where polatuzumab vedotin is added and vincristine is removed from the regimen. Early data are expected soon. 

Another ongoing global study is the ESCALADE trial (NCT04529772), which uses cell of origin and clinical features (IPI ≥2) to create a high-risk category. The trial is limited to patients with nongerminal center B-cell–like (NGCB) disease, who generally have higher risk than patients with GCB disease.  Additionally, it appears that Bruton tyrosine kinase inhibitors have activity in NGCB disease but very little activity in GCB disease. So, the ESCALADE trial is evaluating the use of R-CHOP plus the Bruton tyrosine kinase inhibitor acalabrutinib in these patients. If positive, POLARIX and ESCALADE would both be practice-changing clinical trials. 

For future trials, we will likely be more sophisticated in our approach. For instance, we now know that classifications such as cell of origin may not be sufficiently precise, as there are distinct molecular subtypes within these categories that may respond differently to therapy. Dr Abramson is involved in another interesting trial that is using molecular markers to assess risk, looking at either the double-hit or double-expressor population. These patients generally overexpress the BCL-2 protein and they are known to have inferior outcomes, and the trial is evaluating the addition of venetoclax to standard chemoimmunotherapy in an effort to improve outcomes (NCT03984448). 

Biological factors beyond cell of origin include the double-hit and double-expression populations. In patients with double-hit high-grade BCL, the prognosis has traditionally been seen as very poor with conventional therapy, whereas double expressors seemed to have intermediate outcomes, falling between double-hit and other DLBCL patients. However, the prognosis in both groups may not be as grave as initially described, and I think that this is an important point to raise when counseling our patients.

R-CHOP remains the standard of care for double-expressor lymphomas, and there is no evidence that a more intensified induction regimen is better; however, that does beg the question of whether a more biologically driven strategy might be more successful. Double-hit DLBCL is a heterogeneous group, with some patients having highly kinetic disease, presenting with leukemic phase disease and lactate dehydrogenase levels in the thousands, whereas others present with relatively typical-appearing DLBCL. Thus, there are a lot of biological differences beyond just those cytogenetic changes. However, a biologically driven strategy might improve outcomes in double-hit patients and in double-expressor patients, in whom BCL-2 is uniformly expressed. This is currently being investigated in the clinical trial that was mentioned by Dr Kahl (NCT03984448), which is looking at venetoclax plus chemoimmunotherapy for MYC/BCL-2 double-hit and double-expressing lymphomas.

We are eagerly awaiting the results of the ESCALADE trial incorporating acalabrutinib, and we are also interested in potentially seeing more data with lenalidomide to know whether we should be incorporating that into our treatment of selected patients. POLARIX is an appealing concept across the board, since CD79b is a ubiquitously expressed target, and it appears poised to change our upfront standard of care. Another intriguing strategy is being employed in the ZUMA-12 trial (NCT03761056), which is a nonrandomized study evaluating patients with high-risk LBCL (double or triple hit, or IPI ≥3) as a component of first-line treatment. Patients receive 2 initial cycles of an anti-CD20 monoclonal antibody and an anthracycline-containing regimen, and those with a persistently positive positron emission tomography scan after 2 cycles then receive chimeric antigen receptor T-cell therapy with axicabtagene ciloleucel. And we will continue to see bispecific antibodies getting into the mix as well, to examine whether dialing in additional immunotherapy for both relapsed and upfront therapy might cure more patients with high-risk disease.