Gold standards come and go

Uneasy is the head that wears the crown, they say.  Same goes for “gold standard” CLL regimens, there is always a new kid on the block that would like to replace the present day top gun and take its place.  FCR (fludarabine, cyclophosphamide and Rituxan) has been the gold standard for several years now.  Bendamustine (brand name “Treanda”) has been getting a lot of press and marketing hype recently.  Combination of bendamustine plus Rituxan is a new chemoimmunotherapy combination that is getting attention as a possible rival to better understood combinations such as FCR and FR.

As a potential salvage therapy for people who have relapsed soon after FCR, bendamustine combinations are a very welcome development.  But how about B+R as a frontline therapy?  I can think of at least a couple of CLL experts here in the USA who are pushing this angle, many more are doing so in Germany (bendamustine was discovered in East Germany, around the time of the second world war.  Nothing like home town pride, it seems).  Is B+R a  better choice than FCR, for frontline therapy?  Unfortunately, we do not (yet) have a head-to-head double arm trial results  to make truly credible comparison.  However, this latest article in Journal of Clinical Oncology detailing B+R as frontline therapy gives us a pretty good idea of what to expect.  Abstract is given below, do send me a private email if you would like help locating the full text article.

J Clin Oncol. 2012 Aug 6. [Epub ahead of print]

Bendamustine in Combination With Rituximab for Previously Untreated Patients With Chronic Lymphocytic Leukemia: A Multicenter Phase II Trial of the German Chronic Lymphocytic Leukemia Study Group.

Fischer K, Cramer P, Busch R, Böttcher S, Bahlo J, Schubert J, Pflüger KH, Schott S, Goede V, Isfort S, von Tresckow J, Fink AM, Bühler A, Winkler D, Kreuzer KA, Staib P, Ritgen M, Kneba M, Döhner H, Eichhorst BF, Hallek M, Stilgenbauer S, Wendtner CM.

PURPOSE: We investigated the safety and efficacy of bendamustine and rituximab (BR) in previously untreated patients with chronic lymphocytic leukemia (CLL). PATIENTS AND METHODSIn all, 117 patients, age 34 to 78 years, 46.2% of patients at Binet stage C, and 25.6% of patients age 70 years or older received BR chemoimmunotherapy for first-line treatment of CLL. Bendamustine was administered at a dose of 90 mg/m(2) on days 1 and 2 combined with 375 mg/m(2) rituximab on day 0 of the first course and 500 mg/m(2) on day 1 during subsequent courses for up to six courses.

RESULTS: Overall response rate was 88.0% (95% CI, 80.7% to 100.0%) with a complete response rate of 23.1% and a partial response rate of 64.9%. Ninety percent of patients with del(121q), 94.7% with trisomy 12, 37.5% with del(17p), and 89.4% with unmutated IGHV status responded to treatment. After a median observation time of 27.0 months, median event-free survival was 33.9 months, and 90.5% of patients were alive. Grade 3 or 4 severe infections occurred in 7.7% of patients. Grade 3 or 4 adverse events for neutropenia, thrombocytopenia, and anemia were documented in 19.7%, 22.2%, and 19.7% of patients, respectively.

CONCLUSION: Chemoimmunotherapy with BR is effective and safe in patients with previously untreated CLL.

PMID:22869884

Devil is in the details..

The conclusion stated in the abstract is pretty blunt:  “Chemoimmunotherapy with BR is effective and safe in patients with previously untreated CLL.”  

I suppose one can say that.   That is surely one way of looking at these results.  But just as sure as death and taxes, there is also a different perspective – one that you should be aware of before you make that all important frontline therapy decision.  So, let us dig just a bit deeper and see what is really going on here.

As the abstract points out,  starting in March 2007, 117 previously untreated patients were recruited for this trial.  In keeping with modern criteria for start of therapy, slightly less than half of these guys were Binet stage C (same as Rai stage 3-4).  Median age was 65 years, but about a quarter of the patients were older than 70 years.  This is an important point, since bendamustine is being touted as the kinder and gentler chemotherapy agent, likely to be better tolerated by older patients for whom FCR may be contra-indicated.  That is the claim.  Further down in this analysis we will see how well that claim holds up.

How about prognostics?  Only 7% of patients had the dreaded 17p deletion, and the other high risk FISH defect of 11q deletion was seen in 19% of patients.  16% were positive for ZAP70 expression.  Far higher percentage (88%) had the poor risk CD38 expression.  62% were unmutated IgVH (higher risk group).  Median B2M was 3.4.   All in all, I would classify these guys as sort of middle of the pack in terms of prognostics.  They were by no means toughest patients to treat.   For starters, each and every one of them is a chemo naive patient, no relapsed or refractory folks in this group.  They are kind of patients that could reasonably expect to get a pretty good remission from FCR, if that had been the the therapy they opted for.

The protocol called for  the present day standard of 6 courses, administered over 6 months.   Drug dosages are pretty standard for bendamustine and Rituxan.  Notice, unlike FCR that uses a purine analog (fludarabine) and alkylating agent (cyclophosphamide) along with Rituxan, in this combination bendamustine is the only chemo agent.  That is because it is often suggested bendamustine acts in ways that are similar to both alkylating agents and purine analogs.  That remains to be proven rigorously.  From where I am sitting, bendamustine looks and acts a lot more like alkylating agents (chlorambucil, cyclophosphamide etc) – with a very similar toxicity profile. More on that later.

The abstract is a little skimpy on the safety and adverse effects, you have to read the full text article to gt the juicy details.  Over a followup period of 27 months, the following were noted:

  • 11 patients died during this time period.
  • There were 3 patients who suffered Richter’s transformation.
  • Three patients died within the first five months, due to treatment related infections.
  • All in, roughly two thirds of the patients (64%) experienced grade 3 – 4 level adverse effects.  In case you are not familiar with the grading of adverse effects, grade 1 is nothing to write home about, grade 2 is a bit tougher but considered tolerable (given these guys have cancer!), but grade 3-4 is serious business.  There is another level, grade -5.  That is just a euphemism for the patient died.  So, I suppose you can say 11 patients had grade -5 adverse effect during the 27 months.

What did our guys get for their troubles?  We are all grown-ups here, we understand a certain amount of adverse effects are baked into the chemo cake and we are willing to tolerate them, provided – and this is an important condition – there is good news to be had at the end of it all.  And the good news we are looking for boils down to just three things:  length and quality of remission and overall survival.  Here is how that panned out.

  • The overall response rate was 88%, but only a measly 23% got a “CR” response.  Bulk of the remissions were partial (64.9%) and stable disease in 9.4%.  Bummer!  Since you and I know very well that chances of a partial response or merely stable disease holding the line very long are not all that good, the low rate of CR responses is indeed a disappointment.  Sure enough, as you can see in the graph below, the rate of relapse was pretty steep right from the get go.  Half the patients needed to start new therapy, had progressive disease or died by about 34 months.  (Follow the red lines I drew in, to see how long it took for exactly half the patients to have relapsed). If anything, the relapse rate was even steeper after that point.

  • How about quality of life during and after this therapy regimen?  I wish researchers would make that part of their data collection, ask folks how they rate their quality of life.  But that does not happen too often.  So, you and I have to do our best to guess at how life might have been for these guys.  Hematological toxicities such as anemia and neutropenia take their toll.  The former can cause deep fatigue, the later can precipitate infections and mandate party pooping “social distancing”.  Thrombocytopenia (reduced platelet counts)  increases risk of bleeding and bruising.  In this context, 64% of the patients had one or more high grade (3-4) hematological toxicity.  Don’t know about you, but  this does not sound like a “kinder and gentler” therapy regimen to me.
  • As discussed in the adverse effect section above, 11 patients died during the observation period of 27 months.  I found it frankly difficult to sort out how many of these deaths were attributable to the therapy under discussion, and how many were due to unrelated causes.  Write to me if you wish to read the full text article and try to figure this out for yourself.

An expert comparison

Usually, when discussing these single arm trials I have to scramble around to find a good comparison without too much researcher bias built into it. In this case, my life is made a lot easier since the JCO article above is accompanied by an editorial that does a detailed comparison.  The author of the editorial is no less than Dr. Bill Wierda of M. D. Anderson, one of the lead researchers of the FCR regimen.  He has a wonderful table summarizing the results of this study and comparing it to a bunch of other studies using FCR, PCR, FR and a lot of single agents as well.  I have picked through his table and made the comparison much simpler for you.  Here it is.

Let us put this information into proper context.  The Fischer study is the present study of frontline B+R that we are discussing.  The FCR study from M. D. Anderson is the pivotal study that got chemoimmunotherapy into the limelight several years ago.  Dr. Wierda was closely associated with that whole effort and we can cut him some slack for perhaps feeling a bit more love towards that particular study.  But the third column is the clincher of the argument.  It is the “CLL 8” report of the justly famous German CLL Study Group, a very large scale study conducted at many different institutions.  Heck, the present B+R study was also conducted by the same prestigious group! No one can complain about researcher bias when we compare results from the same group of researchers.   It is about as credible as it gets and I think the results are pretty solid.

Looking over the chart, I am truly taken aback by the differences, especially in the percentage of complete responses and the length of remission (also called PFS, progression free survival).  B+R suffers in the comparison.   Choosing FCR instead of B+R seems to  double the chances of getting a CR (23% versus 44%) and increasing the remission length by a year and a half, even if I take the GCLLSG results as the representative ones. 

Sure, this is not an apples to apples comparison since we are not talking of a double arm trial where every other variable is held constant.  But as Dr. Wierda points out, “Although the patient populations included in the trial may be different, it is difficult to conclude that the BR efficacy outcomes are consistent with FCR in the (GCLLSG) CLL-8 trial.”

It must be pointed out that incidence of grade 3-4 neutropenia was lower with B+R, compared to FCR in the GCLLSG study.  But making up for this, incidence of anemia and thrombocytopenia (reduced platelets) was higher in the B+R study.  Not a whole heck of a lot to point to, if you are trying to sell B+R as being kinder and gentler, compared to FCR.

One of the adverse effects that I worry about with “super” alkylating agents such as bendamustine is the long term risk of myelodysplastic syndrome and secondary acute leukemias down the road.  Researchers tend to be rather coy in discussing this risk, I find.  Longer term monitoring has shown even FCR has a certain percentage of patients who go on to develop myeloid cancers.  Richter’s transformations and  myeloid cancers are the proverbial second shoes yet to drop when considering bendamustine combinations – in the opinion of this layperson patient advocate.

Do send me a personal email if you want to read the rest of Dr. Wierda’s well reasoned and pithy editorial.  I enjoyed reading it.  It clarified the picture a great deal.

Editorial

Not much to add to what Dr. Wierda had to say, just tying a few loose ends to finish the job.  As you can see from the author list, this is a pretty heavy hitting team of German CLL experts.  Also as you would expect, majority of them had financial support from the pharmaceutical companies involved (Mundipharma, Roche), in terms of paid consultant or advisory role, or honoraria – not unexpected, that is how most clinical trials get funded these days.

So, does this important paper succeed in replacing FCR with BR as the new gold standard for chemo naive patients?  We  have to wait a bit more to see one-on-one comparison in terms of two arm trials, but what we have seen thus far suggests we may want to hang on to FCR for a while longer.  Is BR the kinder and gentler version of FCR style chemoimmunotherapy we were hoping for?  I am not sure that has been shown to be the case either, not in view of the high percentage of grade 3-4 adverse effects.  Is it good to have  this drug combination thoroughly tested in clinical trials?  You bet.  It is always good to have choices, and bendamustine combinations are important choices to have.  At this stage of the game, if there is no reason to rule out FCR, would I choose BR as frontline therapy?  I do not think so, the case has not been made to my satisfaction.  I would stick with the “better known devil” of FCR, until and unless more positive results of BR are demonstrated in future studies.

For a change, there was a clear acknowledgement “We thank all patients and physicians for their participation in the study“.  Wow.  I appreciate that.  We must be doing better getting our message out.