When the crystal ball is crystal clear (for a change!)
Most of the time, I find myself having to waffle and hedge my bets when I describe results of clinical trials. The problem is that it costs a huge amount of money, time and effort to do credible clinical trials that give clear and unambiguous answers. Why are trials such as this one so important to us? Because there is often a lot of hype about “quickie” clinical trials and it is important to know what you can trust and what you have to take with a big pinch of salt.
- The large scale of this trial makes the results statistically significant. Results based just a handful of patients are hard to trust, they are no better than gushy anecdotal stories one hears on the internet – dime a dozen and expensive at that. This trial had more than 800 patients participating in it.
- Majority of trials done these days are single arm trials. Results from such trials are compared against “historical norms”. First, it is not always easy to define historical norms, without bias. Second, as medical knowledge improves, historical norms are no longer the right comparison. Did the new trial look better than historical norms because we now know how to take better care of patients, how to protect them against infections for example, because we have better anti-viral drugs available today? This trial does not use historical norms since it has its own built in control group.
- Every one arm trial and comparisons of its results with previous data suffers from being an apples and oranges comparisons. The two sets of patients are never quite the same. Besides unavoidable differences between trial subjects and historical norms, there is quite a lot of room for “cherry picking”, researcher bias and downright human nature. This trial avoids all that. The 800+ patients who participated in this trial were randomly assigned (by a computer program, no human bias or error) to either the FC arm or the FCR arm. Net result, we got two groups of patients that are very well matched: similar on age, disease status, general health etc.
- This study was spread over many top level institutions in Europe, eliminating risk of bias possible in single institution studies. Some of our most credible and well respected researchers were involved in the trial. The oversight committee was top-notch.
- As of now, most of the patients have gone beyond the 3 year mark in this study. As they continue to monitor, we will get more detailed picture of how the two arms (FC versus FCR) differ over the long haul. I have no doubt the researchers will continue to publish data from this trial as the results mature over many more years. And I will review them as soon as they are published.
- With the irrefutable fire power of the results of this clinical trial, most governmental health care agencies and insurance companies will have a hard time insisting on far cheaper FC, in an attempt to skip the expensive “R” part of FCR. I will repeat something I said a while back: with availability of this latest information if your doctor still says you should have FC rather than FCR, get yourself a new doctor.
The abstract of this trial report is given below. I will review most of the relevant details, but I strongly urge you to read the full text of the paper. Just send me a personal email and I will be happy to help you locate it.
Lancet. 2010 Oct 2;376(9747):1164-74.
Addition of rituximab to fludarabine and cyclophosphamide in patients with chronic lymphocytic leukaemia: a randomised, open-label, phase 3 trial.
Hallek M, Fischer K, Fingerle-Rowson G, Fink AM, Busch R, Mayer J, Hensel M, Hopfinger G, Hess G, von Grünhagen U, Bergmann M, Catalano J, Zinzani PL, Caligaris-Cappio F, Seymour JF, Berrebi A, Jäger U, Cazin B, Trneny M, Westermann A, Wendtner CM, Eichhorst BF, Staib P, Bühler A, Winkler D, Zenz T, Böttcher S, Ritgen M, Mendila M, Kneba M, Döhner H, Stilgenbauer S; International Group of Investigators; German Chronic Lymphocytic Leukaemia Study Group.
Collaborators (5)Hillmen P, Dighiero G, Bosch F, Brugiatelli M, Pigeot I.
Department I of Internal Medicine and Centre for Integrated Oncology, University of Cologne, Cologne, Germany. firstname.lastname@example.org
BACKGROUND: On the basis of promising results that were reported in several phase 2 trials, we investigated whether the addition of the monoclonal antibody rituximab to first-line chemotherapy with fludarabine and cyclophosphamide would improve the outcome of patients with chronic lymphocytic leukaemia.
METHODS: Treatment-naive, physically fit patients (aged 30-81 years) with CD20-positive chronic lymphocytic leukaemia were randomly assigned in a one-to-one ratio to receive six courses of intravenous fludarabine (25 mg/m(2) per day) and cyclophosphamide (250 mg/m(2) per day) for the first 3 days of each 28-day treatment course with or without rituximab (375 mg/m(2) on day 0 of first course, and 500 mg/m(2) on day 1 of second to sixth courses) in 190 centres in 11 countries. Investigators and patients were not masked to the computer-generated treatment assignment. The primary endpoint was progression-free survival (PFS). Analysis was by intention to treat. This study is registered with ClinicalTrials.gov, number NCT00281918.
FINDINGS: 408 patients were assigned to fludarabine, cyclophosphamide, and rituximab (chemoimmunotherapy group) and 409 to fludarabine and cyclophosphamide (chemotherapy group); all patients were analysed. At 3 years after randomisation, 65% of patients in the chemoimmunotherapy group were free of progression compared with 45% in the chemotherapy group (hazard ratio 0·56 [95% CI 0·46-0·69], p<0·0001); 87% were alive versus 83%, respectively (0·67 [0·48-0·92]; p=0·01). Chemoimmunotherapy was more frequently associated with grade 3 and 4 neutropenia (136 [34%] of 404 vs 83 [21%] of 396; p<0·0001) and leucocytopenia (97 [24%] vs 48 [12%]; p<0·0001). Other side-effects, including severe infections, were not increased. There were eight (2%) treatment-related deaths in the chemoimmunotherapy group compared with ten (3%) in the chemotherapy group.
INTERPRETATION: Chemoimmunotherapy with fludarabine, cyclophosphamide, and rituximab improves progression-free survival and overall survival in patients with chronic lymphocytic leukaemia. Moreover, the results suggest that the choice of a specific first-line treatment changes the natural course of chronic lymphocytic leukaemia.
FUNDING: F Hoffmann-La Roche.
This is a Phase-III ( = late stage), double arm ( = built in comparison group, FC versus FCR), randomized ( = computer divided incoming patients into the two groups on a random basis), open label ( = patients knew which arm they were in), multi center ( = 11 countries, 190 centers) study. Little over 800 patients were recruited and the two arms got 400 patients each. All were previously untreated but only those needing therapy at the time of recruitment were included. All had to be reasonably fit, no second cancers, no autoimmune disease and good kidney function.
Each cycle was 28 days long. Fludarabine (25 mg/m2) and cyclophosphamide (250 mg/m2) were given intravenously on days 1,2 and 3 of each cycle. In addition, patients in the FCR arm got Rituxan (375 mg/m2) on the first day of the first cycle. This dosage of Rituxan was increased to 500 mg/m2 for all subsequent cycles.
The aim was 6 cycles for each patient, roughly 6 months of therapy. As it worked out, 26% of the patients in the FCR did not complete all 6 cycles. A larger percentage (34%) of the FC patients did not complete the allotted 6 cycles. Patients who were clearly not responding or their disease progressed while on the trial were allowed to drop out and pursue other options. In other words, fewer people on the FCR had to drop out because of lack of response, compared to the FC arm.
How close was the match between the two groups of patients? As you can see below, it was close to perfect. Small differences seen were not worth the mention, not statistically significant.
As you can see, this is not an easy to treat cherry picked lot. Quite a few “Bucket C” cases with poor prognostics. While the median age was 61, the range was from 30 to 80 years old. So, quite a few older patients. But it is important to remember these were physically fit older patients. Anyone with significant co-morbidities were excluded. In other words, your mileage may vary from these results if you are elderly, and you have a bunch of other – often age related – medical issues.
Results – Responses
The major value of this study is the detailed analysis of the results obtained in the two groups. Since this study took place after better understanding of the importance of prognostic indicators such as FISH, IgVH gene mutation status etc, the researchers were able to parse the results according to these characteristics. The table below gives the percentage of complete responses (“CR”) and the overall responses (ORR) in each case. Remember, CR is defined as normalized blood counts, no swollen lymph nodes measured by physical exam, no swollen spleen, and less than 30% CLL cells in the bone marrow. This is the standard definition of CR, defined by the NCI guidelines. It does not look to see if there are swollen lymph nodes buried deep inside that can only be seen by a CAT scan. In other words, a CR response in this study is not the same as a MRD (minimum residual disease) negative response.
Below I have tabulated the CR and ORR (overall response, which is the sum of all patients who responded) for the two groups, further splitting out the responses by age, disease status, prognostics etc.
As you can see, FCR did better than FC on all counts. On most of them, the difference was large enough to be statistically significant. (p = 0.01 means there is only a 1% chance that the difference is simply due to happenstance. p =0.01 means a very high degree of confidence that the difference is very real and not imaginary or wishful thinking).
A few points are worth highlighting. Chances of getting a CR are very slim for 17p deleted patients, with or without Rituxan thrown into the mix. But the other “Bucket C” FISH abnormality of 11q deletion did far better if they got FCR rather than FC. In fact, 11q deleted patients did just as well as the more benign 13q deleted folks, if they got FCR as their frontline therapy. This lesson is very clear and should be something that patients with 11q deletion should take to heart. Would prior exposure to multiple rounds of single agent Rituxan take away this important edge for 11q deleted folks? Would FCR look more like FC for them, when they finally get around to doing it, having exhausted Rituxan single agent therapy? That would be a huge price to pay.
Percentage of CRs were very high for Trisomy 12 candidates with FCR. IgVH gene mutation continues to have prognostic significance, and FCR works better than FC even for unmutated IgVH folks.
Progression Free Survival
This is where rubber meets the road. How long did the remissions last? How long were the patients alive and well, and continued to be in remission? That is the meaning of progression free survival. Below is the chart of PFS for the two arms, all patients included. Please refer to the full text paper if you want to see how PFS worked out for individual groups within the two arms.
We discussed how to read these charts in our second workshop. What you have here is the percentage of patients who continue to be in remission in the two arms, starting the clock at the start of this clinical trial. I drew the red vertical line upwards at the 36 month point, and where this line intersects the two curves for FC (blue) and FCR (green), I drew the red horizontal lines to meet the vertical axis. As you can see, roughly 65% of the FCR folks are still in remission but only 45% of the FC people are still in remission. The clear take home message: patients have longer remissions after FCR compared to FC.
One feature of this and similar clinical trials is this: not everyone enters the clinical trial on the same day. In fact, it takes many months (or even years!) to recruit this many patients. Only some of the early recruited patients have been in the trial for much longer than four years. Some of the “wiggle” you see in the curves past the three year mark is due to the fact that not many patients have been in the trial much longer than that, and the data at the 5 year mark is based on a small handful of people who were there right from the start. The good news is that the patients continue to be monitored. As the study matures over time, we will have more robust statistics about remission rates and survival at the four and five year points. You will see this wiggle in the curves disappear in reports from these researchers in the years to come. But I think you will agree, the trend is crystal clear. Longer remissions with FCR compared to FC. As I said earlier, the break out of how each group fared (by age, FISH and IgVH mutation status etc) is given in the full text paper.
The other big thing we are interested in, besides how long the remissions lasted, is how long our gives stayed alive. Getting a CR is of no value if the remission is short lived and the patient dies soon after. So, I was really keeping my fingers crossed for this one as I read the article. Fortunately, the news continues to be good.
What you are looking at here is the percentage of patients in the two arms (green for FCR, blue for FC) that are alive, starting the clock at the time they started in the trial. Once again, I drew the vertical red line at the three year mark (36 months), and then swung left to meet the vertical axis. 83% of the patients in the FC arm are alive, where as 87% of the patients in the FCR arm are alive. Bear in mind that these are early days for the clinical trial and since CLL is a “good cancer”, not too many people have died. As the trail matures and patients have been monitored for more years, the trends will become even more clear. I am willing to bet that another three years out, we will see statistically significant differences in the overall survival statistics between the two arms.
Allow me to rant a bit. Statistics are not easy to understand, and these survival curves are not easy to read. It is all too easy to get confused by the small difference between 83% and 87% (the overall survival at the 3 year mark, at this point in time). A couple of the more excitable members of our chat rooms have gotten all upset by this. The truth of the matter is that this is an artifact of the time that has passed since the trial started. Give it a bit more time, and as I said, there will be clear and statistically significant differences between the two groups. Take home message:patients live longer if they get FCR, compared to FC. I think that much is obvious even with this early stage data.
With the exception of increased neutropenia (too few neutrophils) and leukopenia (too few white blood cells), addition of Rituxan did not increase the toxicity. In fact, there was no increase in grade 3 or 4 sever adverse effects, there was no increase in severe or opportunistic infections. Less than 1% of the patients suffered infusion related adverse effects during first exposure to Rituxan (chills, hives, allergy type of reaction to the mouse juice). I think this last bit has to do with the fact that we are learning how to administer Rituxan for the first time. Ever so slowly. And you will notice they used a lower dose of 375 mg/m2 Rituxan for the first cycle, bumping it to 500 mg/m2 for all cycles after that. All in all, not much of a price to pay for a whole lot of improvement in efficacy. I like that.
For those of you with attention deficit issues, here are the bullet points to remember:
- This is a very well done and credible clinical trial. The results are therefore trustworthy.
- You really should read the full text article. Lots of details there that I could not cover.
- Patients in the FCR arm got better responses, more CRs and more overall responses than patients in the FC arm. This was true across the board. In a few subsets, this did not reach statistical significance – very few. Most often the difference was statistically significant and obvious even to laypersons.
- Older patients in this trial did very well. But remember these guys were probably more fit and healthy than most patients in that age group. If you are going to go for FCR, you may want to consider doing it while you are still relatively fit and healthy, so you can tolerate all six cycles without dose reduction – that should give you the best possible response.
- 17p deleted patients did not get much joy in either arm. Frankly, patients with deleted 17p (or abnormal 17p function) should be looking at other options – such as Revlimid, Campath, flavopiridol, some of the newer experimental drugs.
- 11q deleted patients did extremely well with FCR – a huge relief to patients with this poor prognostic FISH result. By comparison, FC did not a very good job for them.
- Overall, patients stayed in remission longer after FCR than after FC.
- There is clear indication patients will live longer after FCR than after FC and this difference will get even more pronounced and statistically significant as the study matures.
- Don’t get confused by all the chatter. If you want to read a real professional review of this clinical trial, please visit Dr. Terry Hamblin’s blog. He has published a detailed and multi-part analysis of this study.
- If I get a chance, I will review an excellent review of this study in recent issue of Hematologica. (See caveat below).
- If your insurance company will not cover FCR because it is still not ‘proven’, here is valuable ammunition to change its mind.
- If your national health-care plan will not cover FCR, here is your chance to back up strong patient advocacy and change a few things.
I was planning to stay in India for several months more, getting back to the USA in late February. As it turned out, I will be flying back home in a few days. It seems I have a bit of a medical emergency, a detached retina. Not having good vision scares the heck out of me, especially since I read and write so much. Depending on how things work out, I may have to sign off for a while if I am told to rest my eye after whatever surgery they have to do. Wish me luck.