While diffuse large B-cell lymphoma (DLBCL) occurs in the vast majority of chronic lymphocytic leukemia (CLL) patients who develop Richter’s transformation, the treatment approach depends on a number of factors, but in most cases there exists no standard of care and outcomes are poor, even in the era of novel agents.
“The evidence to support use of one therapy vs the other is not very strong at all, especially for the clonal-related Richter’s, and I believe the majority are clonal related,” said Wei Ding, MD, PhD, of the Mayo Clinic in Rochester, Minnesota, during a presentation at the American Society of Hematology (ASH) annual meeting.
Richter’s transformation — when CLL transforms into a more aggressive lymphoma — occurs in roughly 5% to 10% of patients with CLL. According to past data, nearly all of these (95%) become DLBCL, but are highly heterogeneous, complicating treatment decisions.
According to data from the chemoimmunotherapy era, about 50% of Richter’s patients have CDKN2A deletion, TP53 disruption, and c-Myc activation; 30% have trisomy 12 and NOTCH1 mutations; and 20% have heterogeneous genomic aberrations. “Clearly this is a very high-risk cohort,” said Ding.
In the era of novel therapies, the biology is less well established. More recent studies have reported that 75% of patients have complex karyotype and 70% have disrupted TP53.
Also unclear is whether certain treatments are more likely to result in higher rates of Richter’s in CLL patients. In one study of fludarabine plus cyclophosphamide with or without rituximab (Rituxan), patients receiving the CD-20 antibody had a transformation rate of 3% compared with 6% in the chemotherapy-alone arm, raising the question of whether rituximab may provide some protective benefit.
In patients treated with the Bruton tyrosine inhibitor (BTK) ibrutinib (Imbruvica), studies have reported rates of Richter’s ranging from 3% in first-line to 5% to 9% in the relapsed or refractory setting, with transformation mostly occurring within the first 18 months.
Data for patients on venetoclax (Venclexta) are more limited, but range from 16% to 23% in a small study that included a heavily pretreated population. In those who were 17p deletion negative, or had fewer previous lines of therapy, the rate was around 10%, but dropped to 5% when PET imaging was used to exclude already-existing Richter’s.
“Clearly the prevalence of Richter’s is not decreased in the era of novel agents,” said Ding, noting that long-term follow-up for multiple phase III trials will hopefully inform the true rate associated with different therapies.
Treatment approaches are also poorly researched. Ding said that protocols for these patients are based mostly on small trials and retrospective data. “We don’t have very concrete evidence,” she said.
In the chemoimmunotherapy era, a number of regimens were studied in Richter’s:
- R-CHOP: rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone
- R-EPOCH: rituximab plus etoposide phosphate, prednisone, vincristine sulfate (Oncovin), cyclophosphamide, and doxorubicin hydrochloride (Hydroxydaunorubicin)
- O-CHOP: ofatumumab (Arzerra) plus CHOP
In general, outcomes in these studies were poor, with high treatment-related mortality and rates of grade 3/4 hematology toxicity. Median overall survival (OS) ranged from 6 to 22 months across the different studies. Given the heterogeneous nature of DLBCL after Richter’s transformation, Ding said that a universal approach of R-CHOP or R-EPOCH for every one of these patients would lead to poor outcomes in the majority of cases.
But before any Richter’s-specific treatment begins, PET-guided tissue biopsy is “critical,” as aggressive CLL can also show activity on PET, said Ding, and clinical suspicion needs to be high early on for patients on targeted therapy.
Presentation of Richter’s transformation often includes worsening of discordant lymphadenopathy, rapid clinical deterioration with symptoms such as fevers, chills, night sweats, weight loss, rising lactate dehydrogenase (LDH), hypercalcemia, and cytopenia.
“If the Richter’s diagnosis is confirmed, the first thing that we recommend is to do a clonal relationship test,” she said. “If it is determined to be clonal unrelated, those patients would be best treated as de novo large B-cell lymphoma.”
Data from the chemoimmunotherapy era found that 80% of Richter’s cases are clonally related to the original CLL, with 20% being unrelated or unknown. Some patients have a hybrid disease of both CLL and DLBCL concurrently.
For patients that have clonal-related Richter’s, or where this is unknown, Ding said a clinical trial is clearly the preferred option, especially one testing CAR T-cell therapy or novel agents. Among others, agents under investigation include the XPO1 inhibitor selinexor, blinatumomab (Blincyto), and acalabrutinib (Calquence) or venetoclax, which showed overall response rates of 38% and 40%, respectively.
“But what about if we don’t have a clinical trial?” said Ding. “This is very hard to answer.”
There’s a lack of evidence to support any consensus, she said. “I actually surveyed seven different academic centers — every center had a different approach,” Ding said.
At the Mayo Clinic, the approach to treating Richter’s patients depends on a number of key factors, whether they’ve received prior therapy and what type (chemoimmunotherapy, chemo alone, BCL2 or BTK inhibitors), and whether they have high-risk disease features such as complex karyotype or TP53 mutations — multiple studies have shown these are tied to much worse outcomes.
If a patient is treatment naive, they are treated as though they have de novo DLBCL and these patients can do well without necessarily requiring cellular therapy. Another group that can be approached this way is CLL patients treated with chemoimmunotherapy or chemotherapy alone that don’t have TP53 mutations or complex karyotype, and in this case, those that achieve a partial or complete response should be evaluated for consolidative allogeneic stem cell transplant.
For chemoimmunotherapy- or chemotherapy-treated patients that have high-risk features such as TP53 mutations or complex karyotype, then the typical de novo DLBCL approach is unlikely to yield favorable outcomes. Ding recommended novel combinations, followed by CAR T-cell therapy or transplant evaluation for those with good responses.
The same goes for those with prior ibrutinib exposure, which will likely become more and more common in this setting, as they are hard to treat typically, can have a very aggressive disease course, and often have hybrid disease, with a median OS of <4 months in one study.
For patients after progression on either chemoimmunotherapy or venetoclax who develop Richter’s with poor disease features, Ding suggested a checkpoint plus BTK inhibitor combo if available. A small clinical trial tested pembrolizumab (Keytruda) plus ibrutinib in 25 CLL patients, nine of whom had Richter’s transformation, and selected efficacy was seen in 40% of the Richter’s patients, demonstrating an OS of 11 months.
But none of the CLL patients responded.
“Interestingly, in the patients who have hybrid disease — both Richter’s and CLL — we only see the Richter’s phase responded; the CLL phase continued to progress,” said Ding.
Positive results have also been seen in Richter’s patients on nivolumab (Opdivo) with BTK inhibition.
“A subset of patients treated with nivolumab/ibrutinib can enjoy relative long-term benefit,” Ding noted.
If a PD-1 inhibitor is unavailable, the Mayo clinic would use ibrutinib plus steroid, or ibrutinib in combination with venetoclax and an anti-CD-20 antibody (for the chemoimmunotherapy group). For Richter’s patients who relapse or are refractory, the only option is CAR T-cell therapy — one study in CD-19-directed-CAR T-cell therapy reported responses in three of five Richter’s treated patients.
During the question and answer portion of the session, Kanti Rai, MD, of the Feinstein Institute for Medical Research in Manhasset, New York, said that while from a theoretical point of view it makes sense to find out whether Richter’s is clonally related to the original CLL, “we really do not have a damn treatment whether it is clonal, connected, or unconnected.”
“We still are in a very sorry state of therapy of Richter’s no matter what the origin is,” Rai added.
“I agree with you,” said Ding. “That’s why we need a lot of study.”