Crowning the new king
Which of the chemo-immunotherapy combinations out there today is the undisputed “gold standard” for chemo-naïve CLL patients?
Several large scale studies at M. D. Anderson and elsewhere have established FCR (fludarabine, cyclophosphamide and Rituxan)as a very potent chemoimmunotherapy combination. Especially in chemo-naïve patients the response rates have been very impressive. Complete Response (“CR”) and better still, “PCR-negative remission” (really squeaky clean CR) are the holy grail that all patients hope to achieve as they go into therapy. FCR combination gives higher percentage of CRs and “PCR-negative remissions” than just about any other combination we have seen thus far. Of course, if you are not quite a ‘chemo virgin’ (even exposure to single agent Rituxan disqualifies you as being truly chemo-naïve), then FCR response statistics are significantly reduced.
But FCR therapy is not without its risks. Patients have to be aware of a non-trivial list of adverse effects, including neutropenia that can last for a significant length of time. During this period patients can be at risk of opportunistic infections – hence the need for carefully managed prophylactic therapy with broad spectrum antibiotics, anti-viral drugs and in some case anti-fungal drugs as well. The search goes on for combinations that give terrific response statistics but far fewer adverse effects, better bang for the buck.
Is PCR “kinder & gentler” than FCR?
Pentostatin is similar to fludarabine in the sense that it too is a purine analog and its mechanism of action in killing cancer cells is quite similar. Some researchers think pentostatin is a kindlier version of fludarabine and that it is therefore easier to tolerate and likely to have fewer adverse effects. Which leads us to the obvious question, would substituting pentostatin for fludarabine in the FCR combo give us the best of both worlds? Would PCR combo yield as good response statistics As FCR but with reduced risk of infections?
Most often it is very hard to answer questions like this with any degree of accuracy. That is because often we do not have results from strictly valid comparison trials between two competing therapy regimens. Apples to oranges comparisons are by definition imprecise and often mis-leading. For a change, we now have a well conducted and honest apples-to-apples comparison between FCR and PCR. The results were published at the recent ASH 2008 conference. Anyone who is considering therapy and trying to decide between PCR and FCR needs to read this article. Here is the link to the ASH abstract, followed by my cheat-sheet version that I hope is an easier to follow.
FCR versus PCR: straight up comparison
Until this latest study our understanding of FCR combo comes mostly from studies done at M. D. Anderson. Mayo Clinic has taken the lead in evaluating PCR. (Fludarabine was invented at M. D. Anderson, while pentostatin was invented at Mayo. I am sure you would agree, a certain amount of home court advantage and local pride is to be expected.) In these earlier, single arm trial PCR response statistics seen in the Mayo trial were similar to FCR statistics from M. D. Anderson. The hopeful news was that there were indications PCR may give rise to fewer infection complications and for that reason PCR may be easier to tolerate in elderly patients.
The present study was undertaken to see if indeed there is less risk of infections with PCR combination compared to FCR, whether the early indications can be confirmed in a head-to-head comparison. Thus, infection risk was the primary research objective of this latest study, followed by response statistics as a secondary objective.
The study was conducted at a number of community based oncology practices. Just that by itself is a good thing when considering the value of this latest clinical trial. Results obtained at expert centers like Mayo and MD Anderson may not be representative of what may be possible at local oncology practices. But majority of patients are treated by local practitioners and the results of this study are therefore more likely to be realistic indicators of what we can expect locally. In addition, this trial design is the most valid and head-to-head comparison we can hope for, with enough patients recruited to give the results good statistical credibility.
Well matched groups
Patients were randomly assigned to receive either FCR or PCR combinations. There were 92 patients in each group, a pretty hefty number. Random assignment takes care of biases. Here is how the two groups stacked up:
Pretty well matched groups I would say, and therefore the results give PCR a fair hearing. If anything, a few more of the patients in the PCR group were in earlier Rai Stages.
The table below gives the drug dosage information. The “FCR” combo is pretty much what we have come to expect from earlier work. In the PCR arm of the trial the “C” and “R” are exactly comparable to the doses in FCR. Don’t get excited about the lower milligram number for pentostatin in the PCR arm. This is very much a standard dose for pentostatin and compares well to the fludarabine dose in the FCR arm. In both the FCR and PCR arms of the trial Rituxan was given in a split dose in cycle 1, to reduce infusion related side effects.
Did PCR have the same ‘oomph’ as FCR?
Based on the response statistics shown in the chart below, it does not appear pentostatin did as well as fludarabine in combination with cyclophosphamide and Rituxan. The difference between FCR and PCR in terms of “CR” (complete remission) was statistically significant while in overall response rate (this includes all the folks that got any kind of a response: complete, partial and “nodal” partial responses) the difference was not statistically significant. A bunch more people had stable disease.
All in all, the results for both groups were less than thrilling and I will have more to say about that in the editorial section at the bottom of this article.
Surprise, surprise! The comparison of adverse effects between the two arms was not what I expected. PCR did not come across as kinder and gentler. Both arms had higher infection and hospitalization rates than I expected. Both regimens had significant toxicity, only about 50% of patients completed all the scheduled cycles. More than a third of the patients were hospitalized at some point while undergoing therapy, and the percentage of hospitalizations was higher in the PCR arm! Are you confused yet? You should be.
I have to admit I am surprised by the comparison in the adverse effects profiles of the two arms. Here are some direct quotations from the abstract, highlighted for your convenience. My admittedly ‘snarky’ comments are in parenthesis.
- “Both regimens possess significant toxicity“. (Yeah, tell me about it. More than a third of our guys were hospitalized! I would call that significant toxicity.)
- “Response rates in this multi-institution, community-based randomized trial were lower than previous phase II trials of previously untreated patients“. (I wonder why? Are expert centers so much better than local oncology practices in treating patients?)
- “This trial did not demonstrate a lower infection rate with PCR using pentostatin at the 4 mg/m2 dose level“. (Hunh?? What gives? I thought PCR’s claim to fame is that it is going to be kinder and gentler than FCR?)
- “In early follow-up, no statistically significant differences with respect to overall response rate or survival were observed between FCR and PCR“.
- “CR rate was significantly higher with FCR“. (You telling me PCR has less bang and cost more buck? That does not sound like a winning proposition to me.)
- “We conclude that both PCR and FCR have significant activity in CLL and can be given safely in the community setting”. (Sure. Only half of the patients completed all six cycles of therapy in either arm, a whopping 27% dropped out due to adverse effects, and more than a third of patients were hospitalized. All in a days work, no problem mate, except if you happen to be the one facing the sharp end of the infusion needle.)
I am disappointed and baffled by these results. First and foremost, in this rigorous comparison PCR does not come off well. There was no advantage on the response side (in fact the CR rate was statistically worse), and there was no improvement on the adverse effect profile. This was both surprising and disappointing.
But what baffles me is why the response statistics for both FCR and PCR were so much worse in this study compared to earlier results reported by expert centers (M. D. Anderson and Mayo Clinic). Two possible explanations come to mind for this important discrepancy:
First, it may be that expert centers are trulyexpert at monitoring, protecting and treating patients, hence the better responses and less scary adverse effect profiles in the earlier studies. If this is true, then all of us have to think twice about getting treated at our friendly neighborhood oncology practice. However, not everyone has the choice of high-tailing it to Mayo or M. D. Anderson when therapy time rolls around. If this is the reason for the disparity in results, it further highlights the need for CME (continuing medical education) to bring local guys up to speed, we need to do everything we can to get the darn trickle-down of expert information to speed up. And, from my perspective as a patient advocate, it is all the more reason why you have to take an active role in your own health-care, get your ducks in a row before you make therapy decisions. What you and your local guy do not know may kill you!
The second explanation why the results of this study differ from the earlier study is more troubling. A randomized late stage study such as this with large patient cohorts and multiple centers is the gold standard for clinical trials. Randomization and use of multiple centers removes some of the bias built into single institution and single arm studies. Were the patients in the earlier studies recruited at earlier stages? Did researcher and/or instituional bias contribute to the more benign adverse effect profile and better response rates?
Randomized studies with sizable patient cohorts such as this clinical trial are expensive and time consuming. Neither drug companies nor research institutions are anxious to go the extra mile and more often than not we have to make-do with results published based on single arm and single institution studies. If there can be so much difference between single arm / single institution studies and well conducted “gold standard” randomized studies such as this one, how the heck are we to judge results? Is it all a case of beauty is in the eyes of the beholder?
Oy vey. If there is so much wiggle room in reported results from large scale but single arm and single institution studies such as the FCR and PCR trials at M. D. Anderson and Mayo, how much can we trust other study results obtained in early stage and often Phase -1 clinical trials with less than a couple of dozen or so recruits? And yet company press releases (with an eye on potential investors and stock price) present such early results as the next great thing and the lay press (with no clue as to the technical or statistical credibility of the results) dutifully echoes the glowing quotes from conflicted researchers. Sensational journalism sells.
There is always the third possibility that something is seriously wrong with the results reported in this trial. I am going to be doing some digging in the next little while to see if I can get a better reading of these tea leaves. If I hear anything that sheds light on this situation, on or off the record, I will let you know. For now, I suggest you stay skeptical about stuff you read in the lay press. And if you are trying to decide between FCR and PCR, based on these latest results you may want to stick with FCR. It has longer track record, seems to yield better response statistics (at least as far as percentage of CR responses are concerned and it sure as heck does not seem to be any worse than PCR when it comes to adverse effects. Who would have thunk it?