Sunday, May 08, 2011

Richter's Syndrome: important new information

Richter's syndrome (RS) is the development of a diffuse large cell lymphoma from a background of conventional CLL. It has clearly been recognized that this can occur in two separate ways: either from the progression of the CLL clone to a different, more aggressive, histology or by the development of a clonally unrelated large cell lymphoma.

Much of the literature about Richter's syndrome is inadequate, relying on small or old series, often not confirmed histologically, so the paper in Blood of March 24th is welcome. This Italian study looks at a series of 86 patients.

Median time from the diagnosis of CLL to the diagnosis of RS was 3.7 years (range 0.2-6.7 years). 83/86 were evaluable for survival.The median survival time was 19 months. The follow up was 57 months during which 62.6% had died. Only 40 of the cases of RS were CD5 positive (48.2%); 92.7% displayed a non-germinal center type cell of origin and EBV infection was demonstrated in only 5.9%. The IGHV genes were unmutated in 64.7% and stereotyped VH CDR3 occurred in 36.4%

47.1% carried disruption of TP53 by either deletion or mutation or both. c-MYC abnormalities were present in 26.3%, but BCL2 and BCL6 translocations did not occur. In 50% of cases the c-MYC abnormality was paired with a TP53 abnormality.

Disruption of TP53 was a major prognostic factor. Median survival for those with a disruption was 9.4 months versus 47.1 months for those without a problem at TP53. By univariate analysis other factors associated with poor survival were age over 60, Performance score of >1, tumor size >5cm, platelet count <100, LDH >1.5 x upper limit of normal. Associated with good survival were achievement of CR after remission induction therapy and allogeneic transplant after remission.

By multivariate analysis TP53 disruption was an independent predictor of poor survival along with failure of achievement of CR after remission induction therapy and poor performance score. Use of these three factors stratifies patients into risk groups. Patients with poor performance status had a median survival of 7.8 months. Patients with good performance status but harboring TP53 disruption or not achieving a CR had a median survival of 24.6 months. Patients with none of the adverse risk factors had a 5-year survival of 70%.

It was not possible to determine the clonal relationship between the CLL and RS in 23 cases because paired material was not available, but in the remainder there was a clonal relationship in 79.3% of cases. Those that were clonally unrelated had significantly less TP53 disruption (3 of 13 compared with 30 of 50) and had a lower prevalence of stereotyped VH CDR3. Clonally unrelated cases had a longer median survival (62.5 months) compared to clonally related cases (14.2 months).

I think this is an extremely important paper because we are at a loss as where to go with RS. Generally, patients are treated on clinical suspicion of RS without histological proof and without molecular dissection of the tumor, usually with CHOP-R. What this paper tells us is that while resembling diffuse large B-cell lymphoma (DLBCL), RS is mostly very different. Approximately 4 in 5 cases derive from the CLL clone and in these cases particularly the pattern is quite distinct. DLBCL carries mutated IGHV genes and is usually positive for BCL2 and BCL6. RS carries unmutated IGHV genes in 60% of cases and is negative for BCL2 and BCL6. RS tends to be like refractory CLL in having disruption of TP53 and is therefore unlikely to respond to conventional drugs.

The RS cases that are unrelated to the CLL presumabl occur because of the genetic damage inflicted on normal tissue by the cytotoxic drugs and the release of T-cell suppression caused by drugs like fludarabine. Thankfully, CHOP-R is a good choice in this context, but it just isn't going to work in TP53 disrupted cases.

How then should we proceed? It seems to me that there should be a high index of suspicion. The old literature suggested that the incidence of RS in CLL was 3.5%, but with the advent of more aggressive therapy it might be as high as 10% in multiply treated patients. It is important to treat early when performance status is still good and the tumor small. It is also important to get histological proof so the physician knows what he is dealing with. One thing that should be avoided is blind treatment with CHOP-R, which in many cases will be futile and only serve to impair performance status of the patient.

Which drugs are useful in TP53 resistant cases of intermediate grade lymphoma? Only one that we are sure of, high dose steroids. Other drugs like Revlimid, Campath and flavapiridol have no track record in DLBCL. Perhaps the Btk or PI3Kdelta inhibitors might have a place?


Brian Koffman said...

Important article, but the practical clinical decisions are very tough.
If the enlarged suspicious node is not easily accessible for biopsy, which they are most often are not, one is left using the PET data to make the diagnosis, but with no info on its cytogenetics or P53 status.
Should CLL doctors be more aggressive going as far as open biopsies to get the important prognostic details only possible from a tissue sample?
Does knowing the cytogenetics of the CLL increase our ability to predict the details of the RT and what therapies might worK?
Should HDMP replace R-CHOP as 1st line therapy?
All I can say is: HELP!

Terry Hamblin said...

Those are important points, Brian but we don't know the answers. I would favor getting a biopsy whenever remotely possible. Previous cytogenetics are no guide. A trial of HDMP would not be a bad thing.