PACER: Prophylactic platelets before central lines

There is an odd paradox in medicine: our sickest patients are in the most need of intervention, but they are also the most likely to have complications from those interventions. (This is why I have always opposed the “kitchen sink” philosophy, in which we throw everything at an apparently dying patient. That philosophy assumes that what you are throwing provides more benefit than harm, but if you wouldn’t use the treatment in a healthier population because of lack of proven benefit, you shouldn’t use it in the sickest patients, where harms are likely to be even higher.) When it comes to procedures like central lines, the patients who need them are more likely than the average patient to have coagulopathies which increase their risk. The PACER trial looks at the role of prophylactic platelet transfusion prior to central venous catheter (CVC) placement. 

The paper

The PACER trial: van Baarle FLF, van de Weerdt EK, van der Velden WJFM, et al. Platelet Transfusion before CVC Placement in Patients with Thrombocytopenia. N Engl J Med. 2023 May 25;388(21):1956-1965. doi: 10.1056/NEJMoa2214322. PMID: 37224197

The Methods

This is a multicenter, randomized, controlled, noninferiority trial at 10 hospitals in the Netherlands. 

Patients

Patients with severe thrombocytopenia (platelet count, 10,000 to 50,000 per cubic millimeter) who were being treated on the hematology ward or in the intensive care unit and required a CVC. Exclusions: Therapeutic anticoagulant, coagulation factor deficiency, INR 1.5 or higher. 

Intervention

Prophylactic platelet transfusion (one unit of platelet concentrate) prior to ultrasound guided CVC placement.

Comparison

No platelet transfusion prior to ultrasound guided CVC placement.

Outcome

The primary outcome was catheter-related bleeding (grade 2-4). A noninferiority margin was set at 2.5%.

Physicians in both groups had to have placed at least 50 central lines previously.

The Results

They included 373 CVC placements among 338 patients. The median platelet count was 30,000. About 50% of lines were placed in the IJ, and 40% subclavian.

They were unable to demonstrate noninferiority, with 4.8% of the transfusion group and 11.9% of the no-transfusion group having grade 2-4 bleeding after CVC placement (ARR 7.1%, 90% CI 1.3-17.8%). In fact, although this is how they phrased their results, they actually demonstrated that prophylaxis is superior. (More on these distinctions in an upcoming post on the ins and outs of noninferiority trials.)

There were no grade 4 bleeds in the trial. Grade 3 bleeds followed the same trend as the overall bleed rate, with 2.1% in the transfusion group and 4.9% in the no-transfusion group. 

There were 2 allergic transfusion reactions in the transfusion group as compared to 1 in the no transfusion group. There was one case of acute lung injury in the transfusion group. 

My thoughts

I really like this study, because it is the kind of clinical question we often try to address using observational data. Having an RCT is incredibly valuable, and the fact that this is a well done multicenter RCT, with a pre-published protocol, excellent follow-up, and very few protocol violations is fantastic.

Unfortunately, this trial was open-label, and there was really no reason for that. It feels like a very easy intervention to blind. Worse, the physicians were aware of the purpose of the trial, and the outcome is pretty subjective. Furthermore, when you know your bleeding rates are being monitored, you might be far more careful in your procedural technique than you usually would be (the Hawthorne effect). Thus, instead of a near perfect study, which would have settled this issue nicely, we are left with a lot of uncertainty.

Furthermore, although the study was probably powered appropriately for the bleeding events, it is much to small to assess the harms of platelet transfusions. It only takes a few cases of acute lung injury to completely overwhelm the benefits of reduced minor bleeding, because minor bleeding is easily controlled with local pressure (even if it takes a lot of clinician time). We will need a much larger study to compare the real harms of transfusion again the true complications (bleeding significant enough to have patient impact). 

It is also worth noting that this study has a very high bleed rate, even compared to other thrombocytopenia literature. That might be explained by a more rigorous documentation of bleeding, but it makes extrapolation somewhat difficult.

Finally, different guidelines have suggested prophylaxis at different levels, anywhere between 20 and 50,000. This trial doesn’t clarify those levels, and the data is not granular enough to assess the risk at the higher end (for example, the patients with platelet levels between 40 and 50,000, who some people might not be transfusing based on current guidelines). 

Bottom line

This is a well done (although far from perfect) RCT that demonstrates better outcomes with platelet transfusion prior to CVC placement in patients with platelet counts between 10 and 50,000. Based on this study, I will aim to provide a prophylactic platelet transfusion for patients with thrombocytopenia and platelet counts less than 50,000 per mm³. That being said, neither hospital I work in has platelets on hand, so there I will need to continue to assess the urgency of the CVC against the risk of performing the procedure prior to the platelets arriving (which usually takes a few hours).

Other FOAMed

Platelet Transfusion before CVC Placement in Patients with Thrombocytopenia on The Bottom line

References

van Baarle FLF, van de Weerdt EK, van der Velden WJFM, et al. Platelet Transfusion before CVC Placement in Patients with Thrombocytopenia. N Engl J Med. 2023 May 25;388(21):1956-1965. doi: 10.1056/NEJMoa2214322. PMID: 37224197

Photo by Diana Polekhina on Unsplash