A registered clinical study investigating the effect of platelet activation status on transfusion outcomes was conducted at Vancouver General Hospital (VGH) from 2011 to 2014 under the guidance of Dr. Kate Chipperfield. This study enrolled 200 hematology-oncology patients with informed consent.
Recently, a retrospective analysis was conducted investigating the patients in that study who became refractory to platelets. An abstract describing the analysis, written by Dr. Elisabeth Maurer and reviewed by Dr. Chipperfield, has been accepted by the International Society of Hematology (ISH) for oral presentation. The post below gives an overview of the content in the abstract and provides supporting commentary. The abstract will be presented at the 2018 ISH meeting to be held in Vancouver, British Columbia, September 13th-16th.
Platelet Refractoriness
Platelet refractoriness is defined as two consecutive platelet transfusions with 24-hr corrected count increments below 5x109/L. Up to 35% of hematology-oncology patients who depend on platelet transfusion support become refractory to platelets during their treatment. One possible cause of platelet refractoriness is alloimmunization, however, literature suggests this only accounts for 3-5% of patients. The term “non-immune” refractory is commonly used to describe cases where alloimmunization is not the cause of refractoriness. However, this term may be misleading as complement and other immunomodulatory factors likely play a major role in the removal of platelets. As this map depicts, there are many immunomodulatory factors associated with activated platelets.
The figure below is adapted from this guide on platelet refractoriness by the University of Michigan and shows platelet transfusion responses measured as Corrected Count Increments (CCIs) in the recipient over time. The intended response of a platelet transfusion is a substantial increase in CCI within 1 hour of transfusion and good survival of platelets for about two days, but this is only observed in 65-85% of patients (green curve). If the transfused platelets lead to a sufficient 1-hour CCI but do not survive until the next day patients are said to be “non-immune refractory”. This is observed in 10-30% of transfusion recipients (yellow curve). When there is insufficient recovery of transfused platelets this leads to inadequate 1-hr CCIs in 3-5% of transfusion recipients and they are said to be immune refractory due to alloimmunization (red curve). There are serious patient care and cost consequences to platelet refractoriness making it important to understand the causes of the largely untreated “non-immune” refractoriness segment.
Refractoriness in the Vancouver General Hospital Study
Of the 200 patients enrolled in the VGH study, 52 became refractory. Of the 52 refractory patients, only 3 received HLA matched platelets during the study implying a low alloimmunization rate. This means the remaining 49 were “non-immune” refractory patients. Although refractory patients represented only 26% of the study population, they were the driving force in platelet consumption receiving over 58% of the platelet transfusions. Refractory patients also received over 3 times the number of activated platelets compared to non-refractory patients.
The distribution of underlying diagnoses was consistent between the refractory and non-refractory patient populations. The sex distribution of refractory patients was nearly even, with 56% male patients and 46% female patients.
Incidents of WHO bleeding scores of 2 or higher were rare in this study, however, bleeding was more common for refractory patients with 5.3% of refractory patients recording a WHO bleeding score above 2 compared to only 2.2% of non-refractory patients. Furthermore, refractory patients’ average body temperature was 0.5°C higher, their average length of stay was 42% longer and the end of study survival rate was 10% lower than non-refractory patients.
Conclusions from the Retrospective Analysis
Based on this retrospective analysis alloimmunization, bleeding, gender and underlying diagnosis were not major factors associated with platelet refractoriness. However, elevated body temperature and the activation status of the platelets received were associated with platelet refractoriness. It has long been thought that fever can cause poor platelet response, perhaps it is possible that activated platelets contribute to a febrile response and subsequent transfusion failures due to the immunomodulatory factors associated with activated platelets. Further investigation is needed to determine the precise relationship between these factors.
How was the conclusion that ‘activation status of the platelets received were associated with platelet refractoriness’ ?
Also, body temperature seems to have significant difference.
Dear Dr. Andavolu,
Thank you for your question. The conclusion that transfusions of activated platelets may be associated with platelet refractoriness was drawn from the significantly higher rate of activated platelet transfusions received by refractory patients. The difference in overall body temperature between refractory and non-refractory patients was indeed significant.
Considering the immunomodulatory potential of activated platelets it is reasonable to ask whether a connection between activated platelets and elevated body temperature exists. This possibility has not been considered in the past because the extent of platelet activation in transfusions was not known.
In addition, a possible association of platelet refractoriness with transfusions of activated platelets is in agreement with our other work that has shown a correlation between platelet activation rates and platelet utilization and a reduction in platelet utilization when hematology-oncology patients are allocated exclusively non-activated platelets.
Best regards,
Elisabeth
This is an interesting relevant study in patients receiving frequent platelet transfusions. They do not comment on the number of transfusion on average that leads to refractoriness.
Dear Dr. Seidler,
Thank you for your kind comment and important question.
From our data, the average number of transfusions patients received before clinical refractoriness was 1.6 with a range from 0 to 11. Again, clinical refractoriness is defined as 2 consecutive platelet transfusions leading to 24-hr CCIs of ≤ 5. The low average and wide range are a consequence of a high incidence of refractoriness at the start of transfusion treatment: 25 of the 52 refractory patients showed insufficient CCIs already with their first transfusion. Interestingly, 16 of those unsuccessful first transfusions (64%) were activated platelets.
While no claim can be made that activated platelets cause refractoriness the data do suggest that transfusion of activated platelets could increase the complexity of these already very complex patient cases.
Best regards,
Elisabeth
Of those patients who were not refractory until >2 platelet transfusions, what percentage received activated platelets? Also, once they were refractory due to non-immune causes, did they remain refractory or did it change and what variables were associated(fever, activated platelets, splenomegaly, etc)?
Of the 52 refractory patients 16 patients became refractory after the second platelet transfusion, on average with the 5th transfusion. Of the 16 patients, 14 (87%) received activated platelets, either before the two consecutive insufficient 24hr CCIs (31%) or on the day of the first failed transfusion (56%). 11 of the 16 patients (69%) had a fever when they became refractory – considering that activated platelets are immunomodulatory it is possible that transfusion of activated platelets could cause fever. Five patients died on study. Splenomegaly was an exclusion criterion. GCSF treatment showed some predictive value for poor clinical outcome overall but only 2 of the 16 patients were treated with GCSF when they became refractory.
This means the remaining 49 were “non-immune” refractory patients. IMO this is a bit of a leap. Did you look to see if HLA and HPA antibody screens were carried out in all 52 patients? Certainly those with a positive screen would receive matched platelets but there may have been some patients who did not have the antibody screen sent ? There may have been a choice not to send this screen in some patients.
Poor increments are frequently seen in patients with sepsis or rarely in consumptive coagulopathy patients. Did you gather any data on blood culture results and correlate same?
Finally , a lot of our patients are also receiving red cell transfusions concurrently which are devoid of platelets so there is also a dilutional effect when they are receiving 2-3 units of red cells on the same day as the platelets which can also confound results.
Dear Dr. Power,
Thank you for your interesting questions. When we started the study, we were aware of these questions and therefore collected information to be in a position to answer them.
Your comment “This means the remaining 49 were “non-immune” refractory patients.” is correct by the textbook definition of non-immune refractoriness (2 consecutive platelet transfusions based on 24-hour CCIs). At the time we conducted our study all patients were screened for HLA antibodies at admission to VGH and – according to our patient data – bi-weekly during their hospital stay. Thus, all 200 study participants were screened. Interestingly, 21 patients (10.5%) tested positive for HLA antibodies, 5 of those had always adequate 1-hour CCIs – so no indication of alloimmunization despite antibodies. Only 8 HLA-positive patients had both inadequate 1-hour and 24-hour CCIs – as expected for a fast, antibody-mediated immune response – and 3 of them received HLA matched platelets. This is in line with previous findings (N Engl J Med 1997;337: 1861-9).
This is not surprising given the low success rate of about 26% for HLA matched platelets – according to a recent study conducted by Canadian Blood Services which is in agreement with the literature.
Consumptive coagulopathy was an exclusion criterion and study patients were monitored for sepsis. We collected the following blood culture results: 8 refractory and 7 non-refractory patients had positive blood cultures.
We also recorded when study patients received red cell transfusions concurrently with platelets. Indeed, on 112 of 580 transfusion days (19%) refractory patients received both red cells and platelets but the mean 24hr CCIs showed no statistically significant difference (2.7 with RBC and 3.2 without RBC, t-test, P=0.23).
I hope this helps to clarify these points.
Best regards,
Elisabeth