Saturday, December 31, 2011

No mortality benefit with enoxaparin in acutely ill medical patients... but do these patients resemble yours?

Cardiology blogs and news websites were buzzing earlier this week with the publication of the LIFENOX trial, where prophylaxis for venous thromboembolism (VTE) using enoxaparin (Lovenox®) failed to improve 30-day survival among acutely ill medical patients [1].  Enoxaparin is a low molecular weight heparin used for the prevention and management of several types of thrombotic disorders, including deep venous thrombosis (DVT) and pulmonary embolism (PE).

In the investigation conducted by Kakkar, et al, a once daily dose of enoxaparin 40 mg failed to improve the rate of death among hospitalized, acutely ill medical patients when compared to graduated compression stockings (GCS). While the incidence of all bleeding events was higher among patients randomized to enoxaparin, this was primarily driven by minor bleeding; no differences in major bleeding were observed.

While the results of LIFENOX seem pretty straightforward, I think it is important to highlight a few caveats when making applications to clinical practice. Here are a few limitations, some of which have already been noted by the authors of the trial:
  • The LIFENOX trial assessed mortality and not the prevention of DVTs, a benefit of enoxaparin already demonstrated in previous analyses; to say that enoxaparin is of no benefit in hospitalized patients is not supported by the results of this investigation.  
  • The trial was actually underpowered to detect a difference in mortality as a result of a lower-than-expected mortality rate among patients randomized to GCS alone.  In previous investigations (e.g., MEDENOX [2]) -- where a potential mortality benefit with enoxaparin was observed -- patients randomized to the placebo group received non-pharmacologic DVT prophylaxis according to site-specific standards of care. In LIFENOX, it was specified that patients not receiving enoxaparin wear GCS devices, which may have impacted the ability of the trial to detect a mortality difference between the two groups.
  • The pathophysiology of VTE among surgical and medical patients has traditionally been thought to result from similar processes (i.e., immobility, systemic inflammatory states induced by trauma or other illnesses); however, this may not be the case.  Even in previous analyses of acutely ill medical patients, the rates of confirmed PE as the cause of death is very low.  The authors of LIFENOX hypothesize that enoxaparin may actually reduce the risk of other events not captured by their investigation.
I think the key to interpreting LIFENOX is a comparison of its patient population with those typically admitted to inpatient cardiology services (or other hospitalized patient populations) here in the US.  LIFENOX enrolled patients at sites in Asia, Mexico, the Philippines, and Tunisia.  While there are probably some differences in the baseline risk of VTE between various ethnic groups, I think it is more beneficial to evaluate the non-genetic differences in VTE risk.

After looking at the baseline characteristics of patients enrolled in LIFENOX, I was interested in seeing how these patients compared to those currently admitted to the inpatient cardiology services at our institution. In LIFENOX, the mean body mass index (BMI) was about 23 kg/m2, which would place these patients in the normal BMI range. However, the mean BMI of the patients currently admitted to the cardiology services at our institution (n = 31) is 30.2 kg/m2, which would place our patient population in the obese category.  Given the known risk for VTE associated with obesity, this difference alone would place our patient population at a significantly higher risk than those enrolled in LIFENOX.

I also thought it would be interesting to look at the past medical histories of the patients admitted to our service in order to evaluate what co-morbid conditions might also contribute to their risks for VTE.  Since prophylaxis with enoxaparin may reduce the risk of a variety of thrombotic events among medically ill patients (as the authors note in their investigation), I thought it would be helpful to evaluate the incidence of risk factors for myocardial infarction and stroke, the two most common types of thrombosis among cardiology patients.  The following table highlights these results (n = 31):

Disease StatePrevalence
Coronary Artery Disease48%
Heart Failure60%
Hypertension56%
Hyperlipidemia32%
Diabetes Mellitus44%
Atrial Fibrillation16%
Previous Cancer32%


While the investigators of the LIFENOX trial only provide baseline characteristics for heart failure and active cancer (as well as history of VTE or coagulation disorders), it is pretty clear from these numbers that the population admitted to our service are at a much higher risk for thrombotic complications when compared to those enrolled in LIFENOX.  Using heart failure as one example, the incidence on our service is nearly double that of patients enrolled in LIFENOX (60% vs. 31.4%).

One of the risks of popular trials such as LIVENOX is that the results often become generalized to different patient populations.  The failure of enoxaparin to reduce mortality was quickly tweeted, re-tweeted, and tweeted again.  Unfortunately, the 140 character limit prevents tweeters from detailing the caveats and limitations of a trial alongside its headline results. As one example, a tweet from theheart.org said the following:
"Enoxaparin prophylaxis doesn't cut mortality in acute-care setting."
While I hope readers would click the tweet's included link and read more about the trial, I suspect many did not -- even then, few probably made it past the abstract.

So, while LIFENOX failed to demonstrate an improvement in all-cause mortality among acutely ill medical patients, I am not sure the results are very applicable to our practice (nor many cardiology practices here in the US), where patients are at increased risk for any number of thrombotic events, including DVT, PE, myocardial infarction, and stroke.  While I have no evidence to support it, I suspect -- based on the percentages listed above -- that pharmacologic prophylaxis for VTE is probably reducing the risk of other thrombotic complications in a significant number of our patients.

(As a point of clarification, we use subcutaneous unfractionated heparin rather than enoxaparin for VTE prophylaxis at our institution.)

References
  1. Kakkar AK, Cimminiello C, Bergmann JF, et al; LIFENOX Investigators. Low-molecular-weight heparin and mortality in acutely ill medical patients. N Engl J Med. 2011 Dec 29;365(26):2463-72.
  2. Samama MM, Cohen AT, Weisslinger N, et al. A comparison of enoxaparin with placebo for the prevention of venous thromboembolism in acutely ill medical patients. Prophylaxis in Medical Patients with Enoxaparin Study Group. N Engl J Med. 1999 Sep 9;341(11):793-800.

No comments: