Issues in the Management of Cancer-Related Thrombocytopenia
Issues in the Management of Cancer-Related Thrombocytopenia
The evolving use of
intensive chemotherapy regimens in oncology and bone marrow/stem cell
transplantation programs has increased the demand for platelet products,
particularly for patients with severe thrombocytopenia or bleeding
complications. The use of apheresis platelet transfusions has also increased
substantially, from 352,000 units in 1989 to 1,003,000 units in 1999 (Table 1)[1-3]; this increase is being driven partly by the need for
alternative platelet inventories to support cardiac surgery and peripheral blood
stem cell (PBSC) transplantation programs, and partly by the use of leukoreduced
Emerging issues are renewing interest in platelet transfusion practices.
This review discusses the current risks associated with platelet transfusion and
the results of recent studies of alternative strategies in platelet transfusion
therapy, including: (1) reevaluation of the platelet threshold for
prophylactic transfusion, (2) modification of the platelet transfusion
dose, (3) the potential role of thrombopoietin therapy, and (4) current
investigational and pharmacologic options for the treatment of cancer-related
The risk of acquiring transfusion-transmitted diseases is estimated to be
lower than ever (Table 2).[3,6] Nucleic acid testing has decreased the risk of
viral infection by shortening the window of infectivity, and thus reducing the
risk of posttransfusion infection with the hepatitis C virus and human
immunodeficiency virus (HIV).
Transfusion-associated HIV infection was first reported in late 1982 and
early 1983. HIV antibody testing was implemented in March 1985, and over the
next 5 years, only about five cases of transfusion-associated HIV infection were
reported annually. In the year before testing began, 714 cases had been
reported. In late 1995, blood banks began testing donors for the p24 antigen
to further decrease the risk of transfusion-transmitted HIV disease. In 1999,
nucleic acid testing was introduced to close the window of infectivity (from
infection to detection) by at least 50%, lowering the estimated risk of HIV
transmission by transfusion to approximately 1 in 2 million units.
Labeling of blood from paid donors (a practice initiated in 1972) and the
implementation of third-generation screening tests for the hepatitis B
surface antigen markedly reduced transfusion-transmitted hepatitis B. By 1995,
however, these measures were found to eliminate only about 10% of all
posttransfusion hepatitis cases.
The incidence of non-A, non-B posttransfusion hepatitis was further reduced
when potential HIV-positive donors were excluded and was reduced again when
donors were tested for the surrogate markers alanine aminotransferase (a marker
for acute liver inflammation) and antibody to hepatitis B core antigen (evidence
of previous hepatitis B infection). Even greater reductions in the risk of
transmission of non-A, non-B hepatitis were described after implementation of a
test for antibody to the hepatitis C virus (Table
implementation of nucleic acid testing reduced the current estimated risk of
hepatitis C transmission to approximately 1 in 2,000,000 units.
Platelet Product Contamination
The estimated risk of sepsis related to apheresis platelets is 1:2,000. This
risk is greater with transfusions of pooled platelet concentrates from multiple
donors. Because the risk of bacterial overgrowth increases with time, the shelf
life of platelets stored at 20oC to 24oC is limited to 5 days. The organisms
most commonly implicated in fatalities, in descending order, are Staphylococcus
aureus, Klebsiella pneumoniae, Serratia marcescens, and Staphylococcus
Clinical presentation of infection with bacterially contaminated platelets
can range from mild fever (potentially indistinguishable from febrile,
nonhemolytic transfusion reactions) to acute sepsis, hypotension, and death.
Sepsis caused by transfusion of contaminated platelets is unrecognized, in part,
because the organisms found in platelet contamination are often the same as
those implicated in catheter-related sepsis. The overall mortality rate of
identified platelet-associated sepsis is 26%.
No widely accepted test is available to detect bacterially contaminated blood
products. Currently, the most promising approach is the use of psoralen and
ultraviolet (UV) light to sterilize blood products. In the clinical setting,
any patient who develops fever within 6 hours of receiving platelets should be
evaluated, and empiric antibiotic therapy should be considered.
Cytomegalovirus (CMV) infection caused by platelet transfusions has been
associated with substantial morbidity and mortality in immunocompromised
oncology patients. Patients who undergo allogeneic bone marrow/stem cell
transplantation are at risk of contracting the virus present in blood products
due to their use of cytotoxic preparative regimens or immunosuppressive therapy
(cyclosporine and corticosteroid), or graft-vs-host disease. Up to 60% of
this patient population will become infected with CMV, and 50% will develop CMV
disease if no preemptive therapy is administered.
The risk of developing CMV infection ranges from 28% to 57% for seronegative
bone marrow transplant patients who receive standard blood products. Even
with the use of CMV-negative blood products, CMV seroconversion has been
reported in 1% to 4% of CMV-negative donor-recipient transplant patients.
A recent analysis of our program at Washington University identified CMV
viremia in only 1 (2.5%) of 39 CMV-negative donor-recipient pairs undergoing
allogeneic PBSC transplantation. Our analysis included 59 patients who had
undergone allogeneic PBSC transplantation in an investigational study of
prophylactic granulocyte infusions from stem cell donors. Notably, results
showed that CMV-positive granulocytes did not alter the risk of viremia compared
with CMV-negative granulocytes; the incidence of CMV viremia was 34.5% vs 26.6%,
respectively (95% confidence interval [CI] = 0.47-4.41).
CMV infection and CMV disease occur much less commonly than other virally
transmitted diseases in patients receiving conventional chemotherapy or
autologous bone marrow/stem cell transplantation and are not a significant
clinical problem except in CD34-selected or T-cell-depleted stem cell
A randomized, controlled clinical trial in allogeneic bone marrow
transplantation patients compared the value of CMV-seronegative blood products
vs unscreened blood products subjected to bedside leukofiltration. Of 252
patients in the CMV-seronegative cohort, 4 (1.3%) developed CMV infection, with
no CMV disease or fatalities; 6 (2.4%) of 250 patients in the leukoreduced
cohort developed CMV disease, and 5 of these patients died. The leukoreduced
cohort had an increased probability of developing CMV disease by day 100 (2.4%
vs 0%, P = .03). Even when investigators eliminated the CMV infections that
occurred within 21 days of transplantation, two patients in the leukoreduced arm
and none in the seronegative arm died of CMV disease. The investigators’
conclusions that leukoreduced blood products are "CMV safe" remain
In a consensus conference held by the Canadian Blood Service, 7 of 10
panelists concluded that patients considered at risk for CMV disease should
receive CMV-negative products, even when blood components are leukoreduced.
Debate over the merits of "universal" leukoreduction (cellular
components with < 5 ´ 106 leukocytes) has focused on several potentially
important clinical effects, including transfusion-related alloimmunization to
platelets, febrile-associated transfusion reactions, and transfusion-related
immunomodulation. Use of both leukoreduced and nonleukoreduced blood components
is currently approved by the US Food and Drug Administration (FDA).
Table 3 summarizes the number of leukoreduced blood units collected in the
United States from 1994 to 1999.[2,3] While the percentages of red cell
transfusions remained static, the generation of leukoreduced apheresis
platelets, especially by blood centers, increased substantially. For the
first 9 months of 1999, 13% of all red cells transfused at our hospital
were leukoreduced. Our indications for leukoreduction reflect those published
previously (Table 4).
Febrile ReactionsFebrile-associated transfusion reactions occur
in only 0.5% of patients transfused with red cells, and of these, 18% and 8%
experience a second and third event, respectively. Approximately 18% of all
platelet transfusions are associated with febrile-associated transfusion
reactions, although the prevalence of these responses can be as high as 30%
in frequently transfused populations, such as oncology patients. Reactions
characterized as severe occur in only 2% of platelet transfusions, and
bedside leukofiltration has not reduced the overall prevalence of such
effects.[22,23] Moreover, bedside leukoreduction filters can cause significant
hypotensive events by activating the bradykinin/kininogen systems, particularly
in patients who are receiving angiotensin-converting enzyme inhibitors.
Alloimmunization to PlateletsTransfusion-related
alloimmunization to platelets was studied in a multicenter trial in newly
diagnosed leukemia patients. The study found that clinical platelet
refractoriness associated with human leukocyte antigen seropositivity was
reduced from 13% in patients transfused with unprocessed platelet concentrates
to from 3% to 5% in patients receiving leukoreduced apheresis platelets,
leukoreduced platelet concentrates, or psoralen/UV-B-treated platelets.
Although this difference is statistically significant, no clinically important
differences were found between patient cohorts in prevalence of transfusion
reactions, hemorrhagic events, length of hospital stay, number of platelet
transfusions, number of red cell transfusions, or mortality.
ImmunomodulationTransfusion-related immunomodulation has been
cited as clinically important in patients undergoing renal transplantation and
in women who have had multiple miscarriages. However, a multicenter,
controlled study found no evidence of such an effect, and the authors
recommended against the use of allogeneic mononuclear infusions as treatment for
unexplained recurrent miscarriages. Similarly, patients who received
transfusions before renal transplantation had a superior 1-year renal allograft
survival rate compared with untransfused patients. Nevertheless, among
patients who did not receive a transfusion prior to surgery, blood transfused at
the time of transplantation had no effect on 1-year renal allograft survival.
Only a few prospective studies have attempted to clarify the potential
immunomodulatory effects of allogeneic transfusion in other settings.
Table 4 lists the indications for leukoreduction published in a review in
1992. These guidelines continue to be applicable, pending future controlled,
prospective clinical trials.