Episode 63- “I don’t think it means what you think it means”: INR values in liver disease

Hello and welcome to Episode 63 of ER-Rx. In this Mini Grand Rounds series, we’re going to discuss something that I see a lot of people getting wrong. And that is, misunderstanding and misinterpreting coagulation parameters in the setting of acute and chronic liver failure. The parameter that we’re going to focus on today specifically is the INR. What does it mean when a patient with liver disease has an elevated INR? Are they at a greater risk of bleeding? Let’s talk about this.

Acute liver failure (ALF) is defined as the development of hepatic encephalopathy and coagulopathy, defined as an INR of >/= 1.5. Notice that by definition, these patients will have an INR elevation. In general, the best use of the INR in patients with liver disease is to monitor the degree of reduction in synthetic function or to predict mortality using scores such as the MELD. But, many of us use the INR to assess bleeding risk in acute and chronic liver failure patients- despite that fact that overwhelming evidence shows that in this context, the INR isn’t very useful.

In fact, we shouldn’t be using the INR as a measure of coagulopathy in this setting. First, the INR was developed for one reason only: to evaluate the vitamin K- dependent clotting pathway in patients on warfarin. In the setting of liver failure, both vitamin K-dependent and independent factors contribute to coagulopathy. Second, the INR only tells us about procoagulant factors. The INR comes from the prothrombin time (PT) and is a calculated ratio of a specific patient’s PT to a standardized PT. PT and the activated partial thromboplastin time (aPTT) only reflect the reduction in procoagulant factors- they don’t tell us anything about reductions in anticoagulant factors such as proteins C and S and antithrombin- all of which are also reduced in liver failure. Lastly, the INR is unreliable. There can be huge interlaboratory differences between INRs in patients with liver disease, again because this test wasn’t developed to tell us anything about coagulopathy in this setting. Labs can have wide variability in reported INRs due to different sample storage times, sensitivity indexes, instrumentation, and methodologies.

Despite the elevated INRs that we might see in acute and chronic liver failure, bleeding is rare. In one study of over 1,000 patients with ALF who had a mean INR of 3.8 (range 1.5 – >10), bleeding was uncommon, and the INRs of patients who bled weren’t much different from those who didn’t bleed. Bleeding complications from invasive procedures, like placing an ICP monitor, is also comparable to patients not undergoing invasive procedures. Actually, there is evidence that ALF may be a hypercoagulable state, as shown by the high rates of portal vein thromboses and catheter clotting during renal replacement therapy.

Even though the exact mechanism of coagulopathy in liver disease isn’t quite clear, evidence tells us that this process is a complex and fragile balance of a reduced amount of procoagulant factors and anticoagulant factors, weakened fibrinolytic systems, and reduced and defective platelets. Any tipping of the scales in either direction can lead to either clotting or bleeding. So overall, these patients live in a state of “rebalanced hemostasis.”

In the setting of liver failure there is a reduction in coagulation factors, given that hepatocytes make almost all of them, including fibrinogen and factors II, VII, IX, X, and XI. But remember hepatocytes also make anticoagulant proteins such as C, S, Z, and antithrombin. We know that fibrinogen function is reduced in liver failure, but on the flip side, fibrinolysis is also reduced because proteins involved in fibrinolysis are also synthesized in the liver (things like plasminogen, antiplasmin, and factor XIII). Finally, although there is sometimes a reduction in the number of platelets in liver failure, overall platelet activity is normal or even increased given the increase in von Willebrand factor (vWF). This is all to say that the increased INR does almost nothing to fully describe the in vivo complexity of coagulation in these patients.

So, how can we measure hemostasis in liver failure? Although high quality data for their use in this setting is lacking, global assays like viscoelastic tests (VET), including thromboelastography (TEG) and rotational thromboelastometry (ROTEM) have been used. These point-of-care assays show us real-time functional properties of coagulation including clot formation, clot strength, and clot stability. Interesting to note that in ALF, hemostasis values are usually normal. For example, in one study of 50 ALF patients, despite an INR average of 3.4 (range 1.5-9.6), mean and median TEG parameters were normal for the entire population; 60% had normal TEG studies, 8% had hypercoagulable parameters, and the number of thrombotic complications was higher than the number of bleeding complications (6 pts vs 11 pts). Again, this proves that despite elevated INRs, hemostasis is maintained, and patients may even be hypercoagulable.

In conclusion, in the setting of acute or chronic liver failure, don’t rely on the INR to tell you much about a patient’s bleeding risk. Both the European guidelines and the Society of Critical Care Medicine recommend against relying on the INR as the sole measure of coagulopathy in this setting. We also shouldn’t routinely correct coagulation tests such as the PT/INR in an attempt to prevent bleeding. Even The American Association for the Study of Liver Disease Practice Guidelines for liver biopsy say that there is no specific PT/INR cutoff that can predict bleeding complications.

Unfortunately, given the ease of obtaining the results, the INR is often used incorrectly as a general indicator of a patient’s overall bleeding risk. The practice of correcting an INR value with things like vitamin K, Kcentra, and blood products is common, but it is not evidence based, has been show to do little to nothing, and puts our patients at risk for transfusion reactions, thrombosis, and exacerbation of portal hypertension.

As always, thank you so much for your time. I want to say a big “thank you” to Laura for supporting the show by buying me 5 coffees on buymeacoffee.com. The money donated to the show goes to things like paying monthly website and podcast hosting fees, new and improved podcasting tools, and yes, even sometimes coffee to help me bust through an episode. It also ensures that the show will always be free to you. You can donate to the show by clicking on the hyperlink at the bottom of the show-notes wherever you get your podcasts. I also want to thank listener RxLaura- who may or may not be the same person that made the donation- for the great podcast review on Apple Podcasts. If you haven’t done so, please leave a review of the show. Again, this is super helpful and super easy- just go to the show notes. Anyone who leaves a review will be entered into a periodic drawing for giveaways of small, medically related gifts.