…Factors that is.
On April 23rd 2013, the FDA approved CSL Behring’s product Kcentra as the first 4-Factor Prothrombin Complex Concentrate (PCC) available in the United States for the use in treatment of anti-coagulated patients with acute hemorrhage. 4- Factor PCC has been widely available in Europe and Canada for years, but the FDA wanted more research assessing its safety and efficacy before allowing distribution within the US.
On April 23rd the FDA felt those criteria were met when they released the following:
The FDA’s approval of this new product gives physicians a choice when deciding how to treat patients requiring urgent reversal of VKA anticoagulation,” said Karen Midthun, M.D., director, Center for Biologics Evaluation and Research, FDA. “Kcentra is administered in a significantly lower volume than plasma at recommended doses, providing an alternative for those patients who may not tolerate the volume of plasma required to reverse VKA anticoagulation.
The FDA approval is based off a single randomized, open label, non-inferiority study of 212 patients. Now this study is not yet published in any journal I can find, but the Behring’s representative that attended our residency’s in-service was kind enough to provide me with their promotional package insert which surprisingly contained most of the trial’s results
Based upon the information provided by the link and the information that has been publicly posted on clinicaltrials.gov – we have an idea of how little evidence regarding efficacy a drug company has to provide for the FDA to approve its extraordinarily expensive product.
This non-inferiority trial’s goal by definition was to prove that 4-factor PCC was not inferior to the far more affordable Fresh Frozen Plasma (FFP). According to clinicaltrials.gov the trial’s primary end points are:
- Hemostatic efficacy with respect to the adequacy of stopping an ongoing major bleed [Time Frame: 24 h from start of infusion]
- Proportion of subjects who have a rapid decrease of the INR [Time Frame: 30 minutes after end of infusion]
The authors were smart enough to choose end points that they knew would demonstrate PCC to be superior to FFP. The physiological INR of FFP is around 1.6-1.7, so without the help of vitamin K and, more importantly, time it will never reduce a patient’s INR below 1.3. As far as the trial’s primary endpoints are concerned Kcentra, to no ones surprise, outperform its slower competitor. Whether or not these advantages of Kcentra hold true clinical importance we shall examine shortly. Patients given Kcentra attained the primary endpoints more frequently than those given FFP. 72.4% vs. 65.4% achieved adequacy of bleeding cessation at 24 hours, and 62.2% vs. 9.6% of patients reduced their INR under 1.3 at 30 minutes. The FDA deemed these results to be a clinical success and approved Kcentra for use in the US.
It could be argued that the disease-oriented outcomes the authors defined to be their primary endpoints would translate into clinical benefit for the patients. After all, if you stop the bleeding faster patients must do better. Unfortunately this is not the case. The mortality in the Kcentra group was 9.7% vs. 4.6% in the FFP group. Twice the number of patients died in the Kcentra group, a fact the authors gloss over claiming statistical insignificance despite obvious clinical significance.
Below are the exclusion criteria:
- Expected survival of less than 3 days, or expected surgery in less than 1 day
- Acute trauma for which reversal of vitamin K antagonists alone would not be expected to control the acute bleeding event
- Use of unfractionated or low molecular weight heparin use from 24 hours prior to enrollment or expected need within 24 hours after start of infusion
- For patients with ICH: Glasgow coma score (GCS) < 7; intracerebral hematoma volume > 30cc as assessed by ABC/21; for subdural hematomas: maximum thickness ? 10 mm, midline shift ? 5 mm; for subarachnoid hemorrhage: any evidence of hydrocephalus; infratentorial ICH location; epidural hematomas; intraventricular extension of hemorrhage; modified Rankin score (mRS) of >3 prior to ICH
- History of thrombotic event, myocardial infarction, disseminated intravascular coagulation, cerebral vascular accident, transient ischemic attack, unstable angina pectoris, or severe peripheral vascular disease within 3 months of enrollment
- Known history of antiphospholipid antibody syndrome or lupus anticoagulant antibodies
- Suspected or confirmed sepsis at time of enrollment
- Administration of whole blood, plasma, plasma fractions or platelets within 2 weeks prior to inclusion into the study
- Large blood vessel rupture (e.g. in advanced cancer patient)
- Pre-existing progressive fatal disease with a life expectancy of less than 2 months
- Known inhibitors to coagulation factors II, VII, IX, or X; or hereditary protein C or protein S deficiency; or heparin-induced, type II thrombocytopenia
- Presence or history of hypersensitivity to components of the study medication
- Pregnant or breast-feeding women
As you can see this patient population was highly selected as to reduce the risk of thromboembolic events. It is clear that the authors were aware this might become a problem.
Even with this highly selected population, the rate of thromboembolic events was 8.7% vs. 5.5%(Kcentra vs. FFP). Now imagine when Kcentra is applied to a more generalized population, after all people who are put on oral anti-coagulants are inherently prone to thromboembolic events. Hence the reason why the oral anti-coagulant treatment was initiated in the first place.
One of the additional selling points suggested by Behring is that since Kcentra has a far smaller volume for a relative equivalent dose of coagulation factors, it will provide a “safer” alternative to FFP in patients who are vulnerable to large fluid boluses. This appeared to be the case as 5.8% (all not related to the Kcentra as per the investigators) of patients in the Kcentra group vs. 12.8% of patients in the FFP group (6.4% related to the FFP as per the investigators) experienced “fluid overload.” The authors do not clearly define “fluid overload” and what clinical relevance it holds. Though since the mortality rate is almost double in the Kcentra group one can imagine it had little to no clinical effect on the patients. Furthermore, this was an open label trial. By allowing investigators to pronounce a patient in “fluid overload” and determine whether or not this was caused by the drug in question, lends itself to a great deal of bias.
So where does this leave us? We now have a very expensive drug that does wonders for correcting patient’s elevated INR, but is at best non-inferior to FFP when comparing mortality and most likely is killing people.
Based on this study, the FDA has approved Kcentra for use in the United States. What comes next I’m sure will be a barrage of articles highlighting how vastly superior it is to FFP at correcting INR while quietly ignoring the true clinically relevant endpoint, mortality.
Peer reviewed by Brent Thoma of BoringEM.
Special thanks to Rebecca Talmud of DinoPT