On September 9th 2013 the article we have all been eagerly awaiting (or perhaps just myself!), was published in Circulation by Sarode et al (1). What was formerly known as Trial 3002 was officially named “Efficacy and Safety of a 4-factor Prothrombin Complex Concentrate in Patients on Vitamin K Antagonists Presenting with Major Bleeding”. Surprisingly this important trial was published with very little fanfare and may have gone unnoticed by myself, if it wasn't for the alarm set on PubMed which notifies me of any recent articles published on Prothrombin Complex Concentrates.
Trial 3002 was the randomized control trial the FDA had initially used to justify 4-factor PCC, Kcentra’s approval in the United States (See original post on Kcentra). It was an open label RCT comparing 4 factor PCC to FFP in patients with an elevated INR (>2) who were having “acute major bleeding”, defined as 1 of the following: life-threatening or potentially life-threatening event, acute bleeding associated with a fall in hemoglobin ?2 g/dL; and bleeding requiring blood product transfusion. The patients were treated with either PCCs or FFP according to a sliding scale based on their initial INR value. This trial was somewhat of a straw man comparison as those in the PCC arm received 25, 35, or 45 IU/kg of PCC depending on INR levels. The FFP groups received a relatively lower dose of FFP at 10, 12, or 15 ml/kg according to their respective INR levels. An update of consensus guidelines for warfarin reversal published in The Medical Journal of Australia in March 2013, recommended at least 15ml/kg in patients with INR >2 with severe acute bleeding(9). In a trial published by Chowdhury et al in the British Journal of Haematology in 2004, patients who received a high dose of FFP (30ml/kg) corrected their INR much faster and more frequently when compared to patients who received a low dose of FFP (7).
Did the underdosing of FFP influence the rate in normalization of INR?
Sarode et al developed a unique scale, which they used to assess the hemostatic efficacy of each intervention. As per the authors own admission this is an entirely novel scale, which was never formally derived or validated before its use in this trial. The scores were determined by a blinded Endpoint Adjudication Board (EAB), which was composed of 9 adjudicators (4 Hematologists, 2 Emergency Medicine Physicians, 2 Trauma Surgeons, and 1 Gastroenterologist all seemingly free of conflicts of interest). The scale was divided into 3 general endpoints: excellent, good, and poor, which the authors then compressed into effective (excellent and good) or non-effective (poor).
The trial’s co-primary endpoints were hemostatic efficacy of the intervention (4F-PCC or plasma) assessed over a 24-hour period from the start of infusion and rapid INR reduction (?1.3) at 0.5 hour after the end of infusion. A change in a lab value and a scale of unknown significance are not the ideal endpoints for evaluating patient oriented outcomes. Thankfully the authors also included the rates of thromboembolic events and death.
As was previously discussed in “The Sign of Four“, 4-factor PCC performed significantly better in every disease oriented outcome as compared to FFP. INR reduction (?1.3) at 0.5 hour after the end of infusion was 61% vs 9.6%. Mean time to correction of INR was significantly lower in the PCC group, while mean levels of all the vitamin K dependent factors, protein C and protein S were all significantly higher in the PCC group.
On the other hand PCCs did not outperform FFP when the patient oriented outcomes were examined. In relation to hemostatic efficacy, PCC and FFP groups were statistically equivalent. More importantly the 45 day mortality was 9.7% in the PCC group vs 4.6% in the FFP group.
Sarode et al conclude that 4 factor PCCs are an effective alternative to plasma given their hemostatic efficacy and ability to achieve a swift reduction in INR. Despite hemostatic equivalence and double the mortality associated with PCC as compared to FFP, the authors claim non-inferiority based on PCC’s superior rate of reduction of INR. This is yet another example in the medical literature of a surrogate endpoint used to prognosticate morbidity when there is clear evidence that the events it is meant to predict have not occurred. It is the methodologic equivalent of standing in the rain and asking for the weather report. INR is not the ideal outcome variable for assessing interventions for anticoagulation reversal. We have seen many examples now of fancy drugs that are very effective at reversing a patient’s INR, which have little to no effect on actual clinically relevant outcomes.
INR was originally developed as a more stable measurement of the Vitamin K dependent factors of the coagulation system to adjust for the lab-to-lab variability in the prothrombin time. As the assays evolved they have become more standardized. Whereas in the past tissue thromboplastin from goats, rabbits, and even human brains was used, today we utilize recombinant tissue thromboplastin (rTT). rTT’s major advantage over its predecessors is its greater reliability between assays. One of rTT’s major disadvantages is that it is far more sensitive to changes in factor VII. (2). This may be the main reason why activated factor VII is very effective at reducing a patient’s INR but does little to control bleeding (4).
INR was derived and validated as a prognostic marker for risk of hemorrhage of patients on warfarin over the long term. There is excellent data on the increased risks of bleeding with elevated INR chronically (3) but very little data on how INR affects the acutely bleeding patient. INR even has the capability of providing limited prognostic guidance during an acute bleed (5,6)but it has never been shown to have any accurate capability of quantifying the effectiveness of our attempts at coagulation reversal.
Technically INR is a surrogate for the amount of functional coagulation factor a patient has. The idea is that if the patient’s INR is elevated then they are deficient in one or more VKD factors. The assumption is that in correcting the INR we have corrected the coagulation deficit. The problem is that INR is not an exact correlation to clotting factors. In a study from Thrombosis Research, INR had a sensitivity of 84% and a specificity of 88% of identifying acute trauma patients with deficits in one or more of their coagulation factors (10). An elevated INR has reasonable albeit not perfect test characteristics but it does not inform us which factors specifically are missing. In addition, an INR can be normalized with the addition of a single factor (as is the case with activated factor VII) without correcting the actual deficit(4). There is no evidence that correcting a patient’s elevated INR actually affects true patient oriented outcomes. In a study in Blood, Brett et al demonstrate this concept. The authors examined healthy subjects who were anticoagulated with bleeding times measured before and after being given activated factor VII. For patients who were given activated factor VII, INR normalized but there was no difference in bleeding times between those given the activated factor VII and those given placebo(4).
This is not an attempt to demonize 4-factor PCCs. It may in fact be a clinically relevant medication with true benefits over FFP. What is clear is that at this moment the claims of its superiority are based on physiologic theory and fanaticism. More studies are required to determine PCC’s true efficacy. More importantly it is apparent that our method of evaluating the efficacy of anticoagulation reversal is short-sighted and ineffective. From now on we must be more critical of studies addressing these questions and demand more from an agent then the fact that it quickly and effectively normalizes a patient’s INR. Both TEG and TEM are point of care assays currently in vogue in the management of coagulopathy in trauma that may more accurately and reliably quantify a patient’s ability to form a clot, but they are still a work in progress (11). Replacing one surrogate endpoint with a slightly more accurate one will not provide us with the answer we require. Instead it is necessary to measure true patient oriented outcomes such as hemostasis efficacy, neurological outcomes (in the cases of ICH), and of course mortality. These concepts are doubly important in the era of the direct thrombin and 10a inhibitors where the role of INR is even murkier. It is critical that we maintain high standards of scientific scrutiny and demand high quality patient oriented evidence before accepting prothrombin complex concentrates or any other agent as the preferred treatment for reversal of oral anticoagulants.
- Sarode R, Milling TJ Jr, Refaai MA, Mangione A, Schneider A, Durn BL, Goldstein JN. Efficacy and safety of a 4-factor prothrombin complex concentrate in patients on vitamin K antagonists presenting with major bleeding: a randomized, plasma-controlled, phase IIIb study. Circulation. 2013; 128: 1234-1243.
- Sarode R, Rawal A, Lee R, Shen, Y, Frenkel E. Poor correlation of supratherapeutic international normalised ratio and vitamin K-dependent procoagulant factor levels during warfarin therapy. British Journal of Haematology. 2005; 132, 604–607.
- Palareti G, Leali N, Coccheri S, Poggi M, Manotti C, D’Angelo A, Pengo V, Erba N, Moia M, Ciavarella N, Devoto G, Berrettini M, Musolesi S, on behalf of the Italian Study on Complications of Oral Anticoagulant Therapy, Bleeding complications of oral anticoagulant treatment: an inception-cohort, prospective collaborative study (ISCOAT), The Lancet. 1996 Aug;348(9025): 423-428.
- Brett E. Skolnick, David R. Mathews, Naum M. Khutoryansky, Anthony E. Pusateri, and Marcus E. Carr1 Exploratory study on the reversal of warfarin with rFVIIa in healthy subjects. Blood. 2010 Aug;116(5): 693-701.
- Fang MC, Go AS, Chang Y, Hylek EM, Henault LE, Jensvold NG, Singer DE. Death and disability from warfarin-associated intracranial and extracranial hemorrhages. Am J Med. 2007 Aug;120(8):700-5.
- Rosand J, Eckman MH, Knudsen KA, Singer DE, Greenberg SM. The effect of warfarin and intensity of anticoagulation on outcome of intracerebral hemorrhage. Arch Intern Med. 2004 Apr 26;164(8):880-4.
- Chowdhury P, Saayman AG, Paulus U, et al. Efficacy of standard dose and 30ml/kg fresh frozen plasma in correcting laboratory parameters of haemostasis in critically ill patients. Br J Haematol. 2004;125:69–73.
- The Sign of Four
- Tran HA, et al. An update of consensus guidelines for warfarin reversal. Med J Aust. 2013 Mar 4;198(4):198-9.
- Yuan S et al. Comparing the prothrombin time INR versus the APTT to evaluate the coagulopathy of acute trauma. Thromb Res. 2007 August 2;120(1):29-37.
- Afshari A et al. Thrombelastography (TEG) or thromboelastometry (ROTEM) to monitor haemotherapy versus usual care in patients with massive transfusion. Cochrane Database Syst Rev. 2011 Mar 16;(3).