SEMINARS IN THROMBOSIS AND HEMOSTASIS—VOLUME 18, NO. 4, 1992

Use of Platelets and Other Transfusion Products in Patients with Malignancy

Transfusion support is vital for the management and care of patients with neoplastic disease. Not only is the quality of their life improved, but, because of chemotherapy and radiation, these patients are often totally dependent on transfusion of red blood cells and platelets for survival. This essential service, which is dependent on the willingness of healthy people to voluntarily donate their blood, is both taken for granted by physicians and feared by patients. In truth, these patients are at far greater risk from their disease and the therapies used to treat them, than the perceived risk of infection caused by transfusion. Patients hope for a cure from their treatment or, at the least, palliation from their discomfort. They expect little risk and hope that expenses will not overburden them. Transfusion therapy illustrates a microcosm of this personal and societal phenomenon. The more we treat, the greater the chance of remission, the greater the personal risk, the more the cost. Appropriate transfusion therapy requires that we use only what is necessary to meet an individual patient's needs. To understand appropriateness, we must understand risk and weigh it against benefit. What are the risks associated with transfusion? How great are they and how can they be reduced? Blood, while a living tissue, is considered a drug and so comes under the control of the Food and Drug Administration's (FDA) Code of Federal Regulations. These Codes include the 200 series, which emphasizes good manufacturing practices, and the 600 series, which emphasizes biologies. All collecting agen-

From the Community Blood Center of Greater Kansas City, and the Departments of Medicine and Pathology, University of Kansas Medical Center, Kansas City, Kansas. Reprint requests: Dr. Bayer, Director, Community Blood Center of Greater Kansas City, 4040 Main Street, Kansas City, MO 64111. 380

cies involved in interstate shipping must be licensed and inspected by the FDA annually for the release of any products used in transfusion. Blood-collecting facilities involved only in intrastate collection are registered with the FDA and are periodically inspected. Most collecting facilities are also accredited by the American Association of Blood Banks (AABB) and follow their Standards and are inspected by the AABB biannually.

SAFETY OF A VOLUNTEER BLOOD SUPPLY The tests for infectious disease markers listed in Table 1 are either required or recommended by the FDA (recommendation has the same legal effect). In addition, the AABB Standards require the use of alanine aminotransferase (ALT), a surrogate marker for non-A, non-B (NANB) hepatitis. All of these tests raise costs and improve the safety of the blood supply. However, they also eliminate many donors who are not likely to be contagious. In Kansas City during 1990 and 1991, the present testing system for infectious disease markers resulted in the loss of donors who had, in their lifetime, given 15,661 donations at our Blood Center. These donors are lost primarily because they failed ALT, hepatitis B core antibody (anti-HBc) and hepatitis C virus antibody (anti-HCV) tests. As new tests are added to assure safety, more donors will be lost and it becomes exceedingly important to encourage additional donors to volunteer as well as to balance the perceived benefit of safety against the risk of inadequate supply. A test that is nonspecific, such as ALT, should be reevaluated in terms of benefit. Almost 40% of the total units discarded were due to elevated ALTs in donors who had previously given (Table 2); the overwhelming majority of these donors did

Copyright © 1992 by Thieme Medical Publishers, Inc., 381 Park Avenue South, New York, NY 10016. All rights reserved.

Downloaded by: Universite de Sherbrooke. Copyrighted material.

WILLIAM L. BAYER, M.D., DAVID C. BODENSTEINER, M.D., LOWELL L TILZER, M.D., Ph.D., and MARY E. ADAMS, M.D.

USE OF PLATELETS IN CANCER PATIENTS—BAYER ET AL TABLE 1. Required or Recommended Testing of Donated Blood and Blood Components Food and Drug Administration

381 TABLE 3. Total 1991 Deferred Units with Unconfirmed Repeat Reactive Tests for Anti-HCV or Anti-HBc*

Test*

American Association of Blood Banks

Required

HBsAg

Yes

Yes

Anti-HCV

414

288 (69.6)

15(3.6)

Anti-HIV-1-2

Yes

Yes

Anti-HBc

618

241 (39.0)

25 (4.0)

Anti-HTLV-I-II

Yes

Yes

Total

1032

529(51.3)

40(3.9)

Anti-HBc

Yes

Yes

ALT

Yes

Anti-HCV

Yes

* Total donations for 1991 were 86,754 U. Total percent not used was 1.2%; percent of total donations not used with unconfirmed repeat reactive hepatitis B core antibody (anti-HBc) and hepatitis C virus antibody (anti-HCV) was 0.61%.

RPR

Yes

Recommended

Yes

Repeat Reactive Marker

Total Units Repeat Reactive

Units with Units with Negative Confirmed Indeterminate Confirmed Test (%) Testing (%)

not have any other markers for infectious disease. Donors who are negative for all other hepatitis markers should be considered for possible reentry programs. The loss of donations in 1991 due to ALT, anti-HBc and the new test, anti-HCV, was 2109, or 2.4% of all donations. The anti-HBc test needs specificity improvement. Prevalence data in our donors show a reactivity rate of 3% when they were tested for the first time. However, 0.37% of the donors who have previously passed this test were rejected last year. Only 0.01% of these same donors, however, had corroboration of this finding by radioimmunoassay (RIA) for HBc antibody (HBcAb) or positive results by enzyme immunoassay (EIA) for antibodies to hepatitis B surface antigen (anti-HBs). Table 3 shows that half of the total donations were lost due to failed anti-HCV and HBc Ab testing and that half of these could not be corroborated. This continued loss of committed volunteers because of nonspecificity of tests requires an increased recruiting effort, since these donors average 1.6 donations yearly. Table 4 shows that two thirds of units lost by failed tests of HBcAb and anti-HCV donated by repeat donors were not confirmable.

TABLE 2. Units Deferred in 1991 Due to Elevated Alanine Aminotransferase (ALT) (Donated by Repeat Random Blood Donors)

No. Units

Percent of All Donations by Repeat Donors*

ALT >100 IU/liter

151

0.22

ALT 60-100 IU/liter

648

0.92

Total

799

1.14

* Total donations by repeat donors were 70,006 U. Total number of repeat donors was 43, 122; average, 1.62 U/repeat donor.

The United States has essentially been totally dependent on volunteers for its blood and component transfusion supply since the late 1970s. Plasma collected from paid donors continues to be the main source of supply for derivatives, such as albumin and the Factor VIII complexes. In principle, this policy of volunteerism has been far more important than all the testing in reducing the risk of transfusion-induced infectious disease. In 1970, the Community Blood Center of Greater Kansas City implemented a policy of total volunteerism, which was accomplished by March of 1972. In 19701971, 88% of the reported transfusion-associated icteric hepatitis cases involved paid donors and prison volunteers (high-risk donors). Blood from these high-risk donors comprised 40% of the blood supply. Patient risk for icteric hepatitis is shown in Table 5. Although reporting of all transfusion-associated hepatitis cases may be incomplete, it is likely that most icteric cases were reported. It can be noted that, in terms of reducing jaundiced cases, no great impact was made by the introduction of additional surrogate tests begun in 1987. In 1989-1990, we had eight reported cases of icteric NANB hepatitis prior to implementing the screening test for hepatitis C. All of these were later confirmed to be due to hepatitis C. Since testing for hepatitis C began 19 months ago (May 1990), we have had no reported cases TABLE 4. Deferred Units in 1991 Donated by Repeat Donors with Negative Confirmatory Tests* Repeat Reactive Marker†

No. Units

All Donations by Repeat Donors (%)

Total Repeat Reactive Units (%)

Anti-HCV

222

0.32

79.9

Anti-HBc

181

0.26

51.7

Total

403

0.58

64.2

* Total donations by repeat donors were 70,006 U. Total number of repeat donors was 43,122; Average, 1.62 U/repeat donor. † HCV: hepatitis C virus; HBc: hepatitis B core.

Downloaded by: Universite de Sherbrooke. Copyrighted material.

Yes

*HBsAg: hepatitis B surface antigen; HIV-1-2: human immunodeficiency virus, types 1 and 2; HTLV-I-II: human T-cell lymphotropic virus, types I and II; HBc: hepatitis B core; ALT: alanine aminotransferase; HCV: hepatitis C virus; RPR: rapid plasma reagin.

SEMINARS IN THROMBOSIS AND HEMOSTASIS—VOLUME 18, NO. 4, 1992 TABLE 5. Risk for Icteric Hepatitis in Blood Recipients

Year

Low-Risk Donors

High-Risk Donors

1970

1:5924

1:266

1978

1:5686

1985

1:5880

1988

1:5826

of icteric hepatitis. We have, however, had reports of 30 transfusion recipients who were not known to be jaundiced and were being investigated for liver dysfunction. Although 19 of 30 are either known or reported to be anti-HCV reactive by EIA, confirmatory recombinant immunoblot assay results are not yet complete. These patients were transfused predominately in the period from 1988 through 1990, but many have extended transfusion histories dating into the 1970s and even the late 1940s. Although our investigations are not finished, the impression is that most, if not all, of these HCV EIA reactive patients were infected by transfusions received prior to HCV donor screening or from another source. These patients are being reported now as possible transfusion-associated cases, we believe, because of the availability of possible diagnostic tests. There has been no increase in the rate of acute NANB hepatitis, whether icteric or not, in the years 1989-1990. We have no substantiated cases of HCV transmission since screening of donors for HCV was initiated. Hepatitis may still be transmitted by transfusion, but the risk is extremely unlikely and should be improved even further with the second generation of hepatitis C tests that became available in March 1992. Acquired immune deficiency syndrome (AIDS) is probably the most feared of the infectious diseases caused by transfusion. As of December 1991, 4347 adults and 289 children, less than 13 years old, developed AIDS from transfusion according to the Centers for Disease Control Surveillance Report. Twenty of these developed AIDS after receiving blood that had been screened and was negative, since testing began in 1985. The total of all AIDS cases through December 1991 is 206,392. Over 3 million people are transfused annually in the United States. Platelets are prepared from 40 to 50% of

all donated units and approximately half of all the platelet transfusions are given to patients with hematologic malignancies and cancer (Table 6). In our experience, where 22,000 to 24,000 patients are transfused annually, we are aware of only one patient who, in 1987, received a transfusion that passed the HIV screening performed at that time but that nevertheless transmitted the HIV virus. This false-negative unit would have been detected by the improved HIV detection test utilized today and would not have been detected by using the p24 antigen test. We must recognize that there will be occasional failures in detection, but the strength of the overall system lies in the fact that blood donors are volunteers. Those people whose behavior patterns make them prone to HIV infection are precluded from donating. Other virus infections, such as cytomegalovirus (CMV) and parvovirus may also be transmitted by transfusion. CMV can cause problems in leukemic and bone marrow transplant patients. In our institution, we attempt to give CMV seronegative blood to recipients who are seronegative and also administer the transfusion through leukocyte filters, which should also decrease the risk. Bacterial contamination does occur; the greatest risk, in terms of relative frequency for patients with malignancy, is from platelet transfusion, as they are stored at 20° to 24°C. Graft-verus-host disease is another uncommon risk. All patients receiving units from first-degree relatives should have that blood irradiated and, by present recommendations, all patients with Hodgkin's disease should have every transfusion production irradiated before administration. Hemolytic transfusion reactions do occur and, unfortunately, are still most often due to patients receiving ABO incompatible blood. Present laboratory testing safeguards and careful clerical checks by the transfusionist make this another unlikely problem. Occasionally, patients develop a rapid amnestic response to an antigen they previously have made antibody to, causing a delayed transfusion reaction. This problem cannot be avoided by the crossmatch nor by the use of blood screened for irregular antibodies, but it can be diminished by those procedures, as well as a check of past records to determine if the patient ever had an antibody at an earlier transfusion.

TABLE 6. Random Platelet Use by All Types of Patients (Kansas City Region, July-August, 1991) No. Patients

% Total

Patient Type

Units

% Total

Nonmalignant diagnosis

4412

51.7

388

74.9

11.37

Malignant diagnosis

4128

48.3

130

25.1

31.75

Total

8540

100.0

518

100.0

16.49

Units/Patient

Downloaded by: Universite de Sherbrooke. Copyrighted material.

382

USE OF PLATELETS IN CANCER PATIENTS—BAYER ET AL

383

TABLE 7. Blood Usage* by Platelet Recipients (Kansas City Region, July-August, 1991) % Total†

No. Patients

%Total†

Units/Patient

Nonmalignant diagnosis

8,533

32.1

388

6.4

21.99

Malignant diagnosis

5,188

19.5

130

2.2

39.91

13,721

51.6

518

8.6

26.49

Patient Type

Total

Units

* Includes whole blood, packed red blood cells, fresh frozen plasma, cryoprecipitate, random donor platelets, and single donor platelets, †Total usage by all patients in the Kansas City region was 26,542 U, 6011 patients, average 4.41 U/patient.

TABLE 8. Use of Red Blood Cells in Platelet Recipients (Kansas City Region, July-August, 1991) Patient Type

Units

% Total*

No. Patients

% Total*

Nonmalignant diagnosis

2105

14.2

241

4.8

8.73

836

5.6

121

2.4

6.91

2941

19.8

362

7.2

8.12

Malignant diagnosis Total

Units/Patient

TRANSFUSION USAGE IN THE KANSAS CITY REGION In general, patients whose conditions warrant infusions of platelets tend to also require multiple infusions of blood and other components. Cancer patients are obviously included in this group. In a comparison of platelet recipients with malignant diagnoses versus those with nonmalignant diagnoses in the Kansas City region (which includes 40 counties in western Missouri and eastern Kansas, population 1.875 million) for the months of July and August 1991, it was found that, of a total of 518 platelet recipients, 130 (25.1%) had a diagnosis of cancer, with the majority of them having hematologic malignancy. The total random donor platelet usage was 8540 U, yielding an average of 16.49 U/patient. The cancer patients accounted for 48.3% of random donor platelets and 98.8% of single-donor platelets used during the study period. The breakdown of usage of all blood products between patients with and those without malignancies is

shown in Tables 7 through 11. Patients with malignancies, transfused with random platelets, averaged almost 32 U/person compared with an average of 11 U for all other patients. Of the patients with malignancy who were transfused with platelets, 93% also received red blood cells. However, only 8.5% of the patients received fresh frozen plasma (FFP) and only 1 of 130 was transfused with cryoprecipitate. Although the cancer patients represented only 25.1% of the patients in this group, they received 37.8% of all units. Tables 7 through 11 also show the usage of the various blood products by the study group in comparison to the total usage of these products by all patients in the Kansas City region.

A UNIVERSITY HOSPITAL EXPERIENCE WITH HEMATOLOGIC MALIGNANCIES Bleeding in patients with acute leukemia is a major cause of morbidity and mortality. The most common

TABLE 9. Use of Fresh Frozen Plasma in Platelet Recipients (Kansas City Region, July-August, 1991) Patient Type

Units

% Total*

No. Patients

% Total*

Units/Patient

Nonmalignant diagnosis

1484

67.8

155

39.1

9.57

Malignant diagnosis Total

34

1.6

11

2.8

3.09

1518

69.4

166

41.9

9.14

* Total usage by all patients in the Kansas City region was 2189 U, 396 patients.

TABLE 10. Use of Single Donor Platelets by All Patients (Kansas City Region, July-August, 1991) Units

% Total

2

1.2

2

7.4

1.00

Malignant diagnosis

170

98.8

25

92.6

6.80

Total

172

100.0

27

100.0

6.37

Patient Type Nonmalignant diagnosis

No. Patients

% Total

Units/Patient

Downloaded by: Universite de Sherbrooke. Copyrighted material.

* Total usage by all patients in the Kansas City region was 14,857 U, 5041 patients.

384

SEMINARS IN THROMBOSIS AND HEMOSTASIS—VOLUME 18, NO. 4, 1992 TABLE 11. Use of Cryoprecipitate in Platelet Recipients (Kansas City Region, July-August, 1991)

Patient Type

Units

% Total*

No. Patients

% Total*

Units/Patient

Nonmalignant diagnosis

530

67.6

34

60.7

15.59

20

2.6

1

1.8

20.00

550

70.2

35

62.5

15.71

Malignant diagnosis Total

cause of bleeding is severe thrombocytopenia, but other disorders, such as disseminated intravascular coagulation (DIC), leukostasis, and protein synthetic defects, can also contribute. Thrombocytopenia can be due to inadequate production either because of marrow replacement by leukemic cells, or marrow ablation by chemotherapeutic agents, or it can be due to enhanced destruction because of hypersplenism, DIC, fever, sepsis, or bleeding.1 DIC occurs primarily in patients with sepsis and a DIC-like syndrome commonly occurs in acute promyelocytic leukemia. Leukostasis occurs in patients with hyperleukocytosis and protein synthetic defects occur with underlying liver disease or with certain drugs such as L-asparaginase. The following discussion includes a review of our experience during the past year at the University of Kansas Medical Center.

Blood Usage in the Treatment of Acute Leukemia During calendar year 1991, 31 adult patients with acute leukemia were treated at our hospital. Of these, 27 were newly diagnosed and four patients were in their first relapse. Ablative chemotherapy was given to 25 patients and six patients received supportive care alone, including four patients who had a preexisting myelodysplastic syndrome, one patient with chronic myelogenous leukemia (CML) in blast crisis, and one patient with overwhelming sepsis and early death. A total of 33 courses of induction chemotherapy, to induce remission of active disease, was given to the 25 leukemic patients. Of the 15 patients who entered a complete remission, 12 received consolidation therapy to eradicate residual leukemic cells. Of the three patients who did not receive consolidation therapy, one proceeded to bone marrow transplant, one refused further chemotherapy, and one was transferred to another institution. The blood usage for these patients is shown in Table 12. Blood usage for the six patients receiving supportive care is shown in Table 13. A total of four patients received FFP, three in the treatment group and one in the supportive group. The criteria used for transfusion were as follows: packed red blood cells (PRBC) transfusions were given for a hemoglobin of less than 8 g/dl unless the patient was bleeding or symptomatic, and platelet trans-

fusions, either pooled random donor platelet concentrates (RDPC) or single donor platelet concentrates 15,000 (SDPC) were given for a platelet count of less than 15,000/µl unless the patient was bleeding. All red blood cell and platelet transfusions were made leukocyte poor by treatment with a commercially available leukocyte depletion filter. As can be seen, the induction chemotherapy group and the supportive care group used a similar amount of blood products, whereas the consolidation group used considerably less. The total blood products used by these 31 patients were 597 PRBC, 146 SDPC, and 3085 RDPC. A total of 45 cycles of myeloablative therapy was given and for each cycle patients received a mean of approximately 12.0 PRBC, 3.2 SDPC, and 8 RDPC (based on a typical pool size of 8 U). A total of 58 U of FFP were given to four patients, one of whom had DIC, one of whom had liver failure, and two of whom had had major bleeding. Serious bleeding occurred in seven patients, or 22.6%, and lethal bleeding in 2, or 6.5%. Bleeding episodes were associated with sepsis in five patients, all of

TABLE 12. Usage of Blood Products in Adult Acute Leukemia Patients Receiving Chemotherapy Induction*

Consolidation†

Mean

Total

Mean

Total

RDPC ‡

67.8

2238

17.9

304

SDPC ‡

4.3

143

0.1

2

PRBC ‡

13.8

457

3.7

63

* Twenty-five patients given a total of 33 courses. †Twelve patients given a total of 17 courses. ‡RDPC: random donor platelet concentrate; SDPC: single donor platelet concentrate; PRBC: packed red blood cells.

TABLE 13. Use of Blood Products in Six Adult Acute Leukemia Patients on Supportive Therapy Only* Mean

Total

RDPC

90.5

543

SDPC

0.2

1

PRBC

12.2

77

* For abbreviations, see Table 12.

Downloaded by: Universite de Sherbrooke. Copyrighted material.

* Total usage by all patients in the Kansas City region was 784 U, 56 patients.

USE OF PLATELETS IN CANCER PATIENTS—BAYER ET AL

385

TABLE 14. Mean Units of Blood Products Given to Bone Marrow Transplantation Patients* Good Risk

Poor Risk

All Patients

Total Transfusion

Autologous Transplantation No.

10

7

17

RDPC

89.2

67.3

71.0

1207

SDPC

25.2

11.3

19.3

328

PRBC

14.7

17.3

15.8

268

Allogeneic Transplantation 9

4

13

RDPC

26.6

188.0

68.5

891

SDPC

6.1

18.5

9.9

129

PRBC

6.9

20.3

11.0

143

No.

whom had gastrointestinal bleeding, and bleeding without sepsis occurred in two patients. Both of these were intracerebral hemorrhages, one patient died and the other has significant cognitive deficits. One of these two patients was refractory to platelet transfusions and bled with a platelet count of 8000/µl, but the other patient bled with a platelet count of 23,000/µl.

Blood Usage in Bone Marrow Transplantation During calendar year 1991, 29 adults and one child received bone marrow transplantations at our hospital. There were 13 allogeneic transplants, of which nine were classified as good risks, including four patients with chronic granulocytic leukemia in stable phase, three with acute nonlymphocytic leukemia (ANLL) in first remission, one patient with ANLL in second remission, and one patient with high-grade lymphoma in remission. Four were classified as poor risk, including three patients with acute leukemia in relapse and one patient with a myelodysplastic syndrome. The blood product usage is shown in Table 14. In addition, two patients received a total of 18 U of FFP. The criteria used for transfusion were the same as for the acute leukemia group. All red blood cell and platelet concentrates were irradiated and leukocyte depleted. As can be seen, the poor-risk patients used three times as many PRBC and SDPC and seven times as many RDPC per patient compared with the good-risk patients. There were 17 autologous bone marrow transplants, of which ten were classified as good risks (no active disease) and seven were classified as poor risks (active disease). The blood product usage is also shown in Table 14. In addition, three patients received a total of 16 U of FFP. In this group, the good-risk patients used more

SDPC and RDPC than the poor-risk patients. This is directly attributable to two patients in the good-risk group who used a total of 80 PRBC, 173 SDPC, and 606 RDPC. One of these patients had good erythroid and myeloid engraftment but required 8 months for megakaryocytic engraftment to occur. During this time, she required 54 PRBC, 100 SDPC, and 365 RDPC for the treatment of recurrent epistaxis and gastrointestinal mucosal bleeding. The second patient failed to engraft and required 26 PRBC, 73 SDPC, and 241 RDPC before having a relapse of her acute nonlymphocytic leukemia 4 months after bone marrow transplantation. The mean results for the good-risk patients deleting the two outliers just described would be PRBC 9.13, SDPC 9.1, and RDPC 36.4. The total blood product transfused for the 30 bone marrow transplant patients was 411 PRBC, 457 SDPC, and 2098 RDPC. Based on our typical pool of 8 U of RDPC, each patient would have received approximately 14 PRBC transfusions, 15 SDPC transfusions, and 8.7 RDPC transfusions. There were five episodes (16.7%) of severe bleeding among the 30 transplant patients. Severe bleeding was associated with sepsis in four patients and without sepsis in one. This patient had protracted thrombocytopenia refractory to platelet transfusion and had massive gastrointestinal bleeding. FFP was given to two patients with liver failure, one with DIC, and two after massive bleeding. The overall incidence of severe bleeding in both the transplant group and the leukemic group was 12 of 55 patients, or 22.8%. Bleeding associated with sepsis occurred in nine patients (16.4%) and without sepsis in three patients (5.5%). Bleeding was a primary cause of death in two patients and a contributory cause in an additional four patients. The remaining six patients survived because of aggressive blood product and antibiotic

Downloaded by: Universite de Sherbrooke. Copyrighted material.

* For abbreviations, see Table 12.

SEMINARS IN THROMBOSIS AND HEMOSTASIS—VOLUME 18, NO. 4, 1992

support. The total blood product usage for the leukemic and bone marrow transplant patients combined was PRBC, 1008; SDPC, 604; and RDPC, 5726. The total blood product usage for 11 (20%) of the largest users was PRBC, 420 (41.6%); SDPC, 233 (38%); and RDPC, 520 (44%).

Thrombocytopenia In 1964, a report from the National Cancer Institute reviewed the experience with 414 patients with leukemia evaluated between 1954 and 1963 and demonstrated the value of platelet transfusions in their treatment. Before platelet transfusions became available in 1959, 67% of these patients had a fatal hemorrhage, but from 1960 to 1963 only 37% of such patients died of hemorrhage.2 Other studies have confirmed this observation.3 With aggressive platelet transfusion support, more recent reports have indicated a much lower incidence of fatal hemorrhage. Walters and coworkers4 reported 3 of 41 patients dying of hemorrhage alone, whereas 2 of 55 patients died of hemorrhage alone in our study. However, bleeding in combination with sepsis continues to be a major cause of morbidity and mortality. Although it is obvious that patients with severe thrombocytopenia and bleeding should be transfused, there is some controversy in regard to the role of prophylactic transfusions. In particular, at which platelet level should they be given, and are they more beneficial than therapeutic transfusions. A platelet level of 20,000/µl has been generally accepted as the level at which prophylactic transfusions should be given based on the study by Gaydos and coworkers.5 However, the authors did not note a threshold platelet value at which there was no bleeding. Gross bleeding was most frequent at platelet counts less than 5000/µl. In patients who had platelet counts of 10,000 and 20,000/µl, the percentage of days with bleeding was 8% and 4%, respectively. Intracranial bleeding in eight patients who did not have CML in blast crisis occurred at platelet levels of less than 10,000/µl. A recent prospective study by Gmur and coworkers6 describes a more stringent prophylactic transfusion policy. If the morning platelet count was 5000/µl or less, platelets were always given. If the platelet count was 6000 to 10,000/(xl, platelets were given for fresh minor bleeding manifestations or for body temperature of 38°C or greater. When the platelet count was 11,000 to 20,000/ µl, platelets were only given if the patient had an underlying coagulation disorder, was receiving heparin, or was undergoing a bone marrow biopsy or lumbar puncture. Platelets were given to control major bleeding or for minor surgical procedures when the platelet count was above 20,000|xl. Total platelet usage was 972 transfu-

sions for 254 cycles of marrow ablative therapy in 102 patients, or approximately four platelet transfusions per cycle. There were 28 nonlethal bleeding episodes, of which ten required PRBC transfusions, and three lethal bleeding episodes, or 3.0%. This is in comparison to our data using 15,000 platelets/|xl as the level at which prophylactic transfusions are given and the average number of transfusions per cycle was 11.2 and lethal bleeding occurred in one of 31 patients, or 3.2%. Another approach is the use of therapeutic transfusions that are given only to control bleeding. There are two controlled studies that prospectively evaluated prophylactic versus therapeutic platelet transfusions. In both studies, the prophylactic group used twice as many platelets, but there was no difference in the survival, number of remissions, or bleeding deaths between the two groups.7'8 There was, however, a reduction in the number of days with active bleeding in the prophylactic group in the first study during the first 4 months of transfusion, but with time bleeding in some of these prophylactically treated patients could no longer be controlled, presumably due to platelet alloimmunization (PA). These studies indicate that more stringent transfusion polices, or therapeutic transfusion policies, can decrease platelet usage without changing the incidence of lethal bleeding.

Platelet Alloimmunization PA is a common problem occuring in 30 to 70% of patients requiring long-term platelet support. PA can be defined on the basis of both laboratory and clinical criteria. Laboratory criteria can be described as the presence of antibodies in the recipient serum directed against antigens on the donor platelets. These antibodies are usually directed against HLA determinants on the donor platelets, although 25% of the patients may also have plateletspecific antibodies. The clinical criterion for PA is an inadequate platelet increment following platelet transfusion. A corrected count increment (CCI) of less than 10 × 103 platelets/µl at 1 hour post-transfusion is defined as platelet refractoriness in patients who are not febrile, septic, actively bleeding, in DIC, or have palpable splenomegaly.9 The formula is as follows:

PA can occur as early as following the initial transfusion of platelets, but some patients have received hundreds of platelet concentrates without any evidence of alloimmunization.10 Once PA develops, therapeutic options depend on the expected natural history of the patient's underlying

Downloaded by: Universite de Sherbrooke. Copyrighted material.

386

disease. Patients in whom the duration of thrombocytopenia is expected to be short, for example, postchemotherapy, can be managed with either HLA-matched platelet transfusions, if available, or SDPC selected with a platelet crossmatch procedure. HLA-matched platelets can be obtained from family members if tissue typing has been done, or, in some centers, a pool of HLA typed nonrelated donors is available as needed. Platelets matched for three or four of the four HLA-A and B loci will yield good increments in the majority of patients. However, many times HLAmatched platelets cannot be found for a patient. In this instance, crossmatched selected SDPC can be used. In platelet crossmatching, recipient sera are incubated with donor platelets to determine if the recipient antibodies will bind to the donor platelets.11 At our institution, this technique is used in all alloimmunized patients who do not have potential HLA identical donors. Compatible donors can be identified for most patients, but, as time goes on, fewer and fewer donors will remain compatible. Patients in whom thrombocytopenia is expected to be a long-standing problem can be managed with HLAmatched or crossmatched selected platelets. In addition, transfusing only when active bleeding occurs is a reasonable approach, since continued transfusion may result in antibody production of wider specificity. Other approaches that have been tried include highdose immunoglobulin infusions given prior to platelet transfusion, which has been of limited success, and epsilon-aminocaproic acid (12 to 16 gm/day in divided doses) which has been shown to decrease mucosal bleeding.12 There are strategies available for the potential prevention of PA. The use of HLA-A,B identical platelets will prevent or delay the development of alloimmunization in most patients. However, HLA identical platelets are not available at many centers. Random single donor transfusions have also been shown to delay PA.13 Several recent studies have indicated that PA can be delayed or prevented by using leukocyte-depleted blood products. The incidence of PA dropped from 50% in the control group to 15% in the group receiving leukocyte-poor blood products in the study by Sniecinski and coworkers14 and from 48 to 16% in the study by Murphy et al. 15 A third study using leukocyte-depleted blood products exclusively had a 21% incidence of detectable platelet antibodies and a 9% incidence of platelet refractoriness.16 Commercially available leukocyte removal filters, supplied by several vendors, consistently remove 99% of residual leukocytes in PRBC or platelet concentrates.17 A new generation of filters recently available can result in a 3-log removal.18 Preliminary studies indicate that ultraviolet irradiated blood products also delay PA, presumably

387 by inactivating antigen-presenting cells, but this approach is not FDA approved.19 Potential new approaches to maintain platelet counts after chemotherapy or in the cases of marrow failure include the use of the interleukins. Several of these hemopoietic growth factors, in particular interleukin-3, have a stimulatory effect on megakaryocytes. Clinical trials in this area are only beginning.

Disseminated Intravascular Coagulation DIC is a syndrome that results in the widespread deposition of fibrin in small blood vessels due to the indiscriminant formation of fibrin monomers and soluble polymers. DIC may present as a thrombotic disorder due to fibrin deposition or may paradoxically present as a bleeding disorder due to consumption of fibrinogen, other coagulation factors, and platelets. DIC can be triggered by any mechanism that results in the discharge of ungovernable amounts of thrombin into the circulation. Sepsis is the most common cause of DIC, accounting for 20 to 29% of cases. Unfortunately, gram-negative sepsis and shock as well as gram-positive infections frequently occur in neutropenic leukemic patients and can precipitate DIC. 20 Another common cause of DIC is malignancy, particularly acute promyelocytic leukemia (APL).21 The DIC associated with APL is attributed to spontaneous or chemotherapy-induced release of a tissue factor with procoagulant activity that is present in the granules of promyelocytic leukemia cells.22 In addition, primary fibrinolysis mediated by plasmin and digestion of coagulation factors by elastase-like proteases may also play a role.23 Two large studies have evaluated the incidence of DIC in APL. In the series reported by Kantarjian et al, 24 58 of 60 patients had a DIC-like syndrome and 16 early deaths due to fatal hemorrhage occurred. Heparin therapy was associated with the trend toward decreased hemorrhagic death. Hoyle and coworkers25 reported on 115 patients and found that the use of heparin significantly decreased hemorrhagic death from 30 of 80 (37.5%) in patients not receiving heparin to 3 of 35 (8.6%) in patients receiving heparin. Although there is no consensus as to dose or route, a reasonable recommendation for treating APL should include heparin in a dose of 400 to 1000 U/hour given intravenously as well as aggressive blood product support with FFP and platelet concentrates to control bleeding. Patients not in DIC at the time of presentation should be monitored closely because chemotherapy may precipitate the consumption coagulopathy. The management of these patients is controversial. Prompt institution of broad-spectrum antibiotics is the cornerstone of therapy. Controversy exists in regard to

Downloaded by: Universite de Sherbrooke. Copyrighted material.

USE OF PLATELETS IN CANCER PATIENTS—BAYER ET AL

SEMINARS IN THROMBOSIS AND HEMOSTASIS—VOLUME 18, NO. 4, 1992

the use of heparin and blood product support. Heparin is effective for "thrombotic DIC," but its role in "bleeding DIC" is less clear. Some institutions give platelet concentrates and FFP, whereas others avoid blood product support for fear that replenishment with coagulation products will perpetuate the DIC. In any case, to treat DIC, the underlying disorder must be brought under control.

Leukostasis Leukostasis refers to the sludging of blood flow in the microcirculation due to a high concentration of intravascular leukemic cells, with the subsequent lodging and growth of these cells in the vascular system. These foci of leukemic cells proliferate, and destroy the vessel wall, frequently producing severe hemorrhage and macroscopic extravascular tumors referred to as leukostatic tumors. In an autopsy study, McKee and Collins26 reported a high incidence of leukostasis and leukostatic tumors in 15 of 30 patients with CML in blast crisis and 33 of 82 patients with acute myelogenous leukemia (AML).26 They were rare in CML in chronic phase (2 of 17), in chronic lymphocytic leukemia (3 of 33), and in acute lymphocytic leukemia (0 of 39). There was a direct relationship between the degree of elevation of leukemic cells in AML and blastic CML and the frequency of leukostasis. The organs most frequently involved were the lungs and the brain. Severe respiratory compromise can occur with leukostasis. Fatal intracranial hemorrhage has also been reported. Gaydos et al5 described eight patients with blastic crisis who died of intracerebral bleeding, and on autopsy they found leukostasis and leukemic nodules. The treatment of leukostasis requires the prompt institution of antileukemic therapy to lower the leukemic cell count rapidly. In patients with marked hyperleukocytosis, leukapheresis may also be beneficial. Patients who present with hyperleukocytosis, who do not have signs of leukostasis, should also be treated promptly to prevent its occurrence.

Coagulopathy Extensive liver disease can cause protein synthetic defects that may result in the inadequate production of coagulation factors. This can predispose patients to bleeding problems. In addition, vitamin K deficiency can occur in malnourished patients with resultant functional deficiences of coagulation Factors II, VII, IX, and X. This is a common problem in hospitalized patients who have either no or poor dietary intake, especially in combination with antibiotic therapy. These patients should be replaced with parenteral or oral vitamin K two or three times a week.

In addition, certain drugs such as L-asparaginase and mithramycin can induce a coagulopathy by inhibiting the production of coagulation factors. L-asparaginase is an enzyme that deaminates asparagine, making it unavailable for protein synthesis. It is effective in certain acute leukemias, which are unable to synthesize asparagine. Unfortunately, interruption of protein synthesis may result in undesirable side effects within normal host tissues, including hypoalbuminemia and a decrease in plasma clotting proteins. Hypofibrinogenemia and low levels of multiple coagulation factors (II, V, VII, VIII, IX, X) may be observed.27 Concomitant liver dysfunction will further diminish synthesis of these proteins. Mithramycin has also been reported to decrease the levels of coagulation factors and also has been shown to induce abnormal platelet function through the depletion of platelet adenosine diphosphate (ADP).28 Fortunately these patients rarely bleed and these abnormalities are reversible with discontinuation of the drugs. In fact, many times L-asparaginase does not need to be discontinued even though the levels of fibrinogen and other clotting factors are low.

Bleeding in Patients with Solid Tumors The most common malignancies, carcinomas, lymphomas, and sarcomas (solid tumors), have little written about them in terms of use of blood components for bleeding. The need for components in our university hospital for oncology patients is less than 10% of that needed for the hematology patient (Table 15). Thus, the major epithelial tumors, including lung, breast, colon, prostate, kidney, and skin, use small amounts of platelets and FFP compared with hematologic malignancies and bone marrow transplants. The same is true for sarcomas, non-Hodgkin's lymphomas and Hodgkin's disease. Most component therapy for the bleeding cancer patient goes to leukemias, myelomas, myelodysplastics, and bone marrow transplants, either allogeneic or autologous. Of TABLE 15. Fresh Frozen Plasma and Platelet Use at a University Hospital in 1991 Service

Platelets Used*

FFP† Used

All med. services ‡

8,207

1820

All surg. services

1,426

616

Hematology

6,224

838§

Oncology Total

192

24

9,633

2436

* Includes random and single donor platelets. †FFP: fresh frozen plasma. ‡Includes hematology and oncology. § Includes nonleukemic patients also (e.g., patients with thrombotic thrombocytopenic purpura).

Downloaded by: Universite de Sherbrooke. Copyrighted material.

388

course, many of the autologous bone marrow transplant patients have a primary solid tumor. With this in mind, we will discuss the patient with solid tumors requiring component therapy for bleeding or for prophylaxis to prevent bleeding.

Thrombocytopenia The usual cause for bleeding in the oncology patient with solid tumors is lack of adequate platelets. Qualitative platelet defects can occur also, but the vast majority of bleeding episodes are due to low platelet numbers. The cause of the thrombocytopenia (and neutropenia) in oncology patients is chemotherapeutic agents aimed at destroying primary and metastatic disease. These agents' mechanisms of actions usually have in common interference with cell cycling. Since hemopoietic elements are continuously being produced at a rapid rate, they are susceptible to the chemotherapeutic agents. Cytokines given to cancer patients to increase thrombopoietic activity to counteract the toxic chemotherapies are still in their infancy.29 In the future these therapeutic agents may help stem the tide of hemorrhage due to thrombocytopenia. However, as noted by Belt et al, 30 less than 11% of 718 patients with solid tumors had hemorrhagic complications due to thrombocytopenia induced by 3942 rounds of chemotherapy. Almost half of these hemorrhages were confined to skin or mucosal bleeding. Fatal hemorrhage was almost nil when the platelet count was more than 10,000/µl. In a retrospective study (1972-1980) done by Heyman and Schiffer,31 11% of 1274 patients experienced platelet counts below 20,000/|xl. Death due to hemorrhage occurred in less than 1% of the reviewed population. Severe hemorrhage usually occurred only when infection (usually sepsis), tumor invasion, or coagulation abnormality were simultaneously present. In recent years chemotherapy, especially for nonsmall cell carcinoma of the lung, has become much more intense, leading to prolonged myelosuppression and thrombocytopenia.32-38 In fact, one drug, hexamethylene bisacetamide, had as its major toxic side effect thrombocytopenia. Young et al39 showed that of 33 patients receiving 10 day intravenous courses of the drug, four had bleeding. In most cases the bleeding was within the tumors (leiomyomsarcoma and yolk sac tumor) and bleeding from a gastric carcinoma into the gastrointestinal tract. As the drug regimens become more toxic, one can expect longer periods of thrombocytopenia and subsequent hemorrhagic complication. In addition, Panella et al40 have shown that high-dose chemotherapy causes qualitative platelet defects as measured by bleeding time and platelet aggregation. Although bleeding times were only slightly prolonged, secondary aggregation was sig-

389 nificantly affected with ADP and arachidonic acid reagents. These results were found in patients receiving high doses of cisplatin and cyclophosphamide for a variety of solid tumors. The qualitative defect may not in itself cause much bleeding, but in combination with lowered platelet numbers, clinically significant bleeding did occur in one patient. Technologic changes and the perception of when to give platelets has changed. Ten to 15 years ago hematology instrumentation was linear to about 15-20 × 109 platelets/liter. Modern instruments are capable of producing accurate counts below 5 × 109 platelets/liter. In the past, when the clinician saw a platelet count of 17,000, they did not know if that meant 6000 or 20,000, due to the inaccurate instruments. With modern-day instruments, the values are so accurate and precise, that the physician should have complete confidence in the information and should be able to make better clinical decisions. Modern instruments are linear down to 1 × 109 platelets/liter (Coulter Electronics, Hialeah, FL). In this vein, it has recently been argued by Gmur et al6 that prophylactic platelet transfusions can safely be held even when the platelet count reaches 5000 platelets/µl. The authors point out that this is not true when the patient simultaneously has fever, coagulation abnormalities, bleeding, or when a procedure is to be performed. Although the study was only done on leukemia patients, the results should be extrapolated to solid tumor patients. Their protocol for platelet transfusion is as follows: 0-5 (× 109 platelets/liter), give platelet transfusions; 5 to 10 (× 109 platelets/liter) bleeding and fever; 10 to 20 (× 109 platelets/liter) DIC or heparin; more than 20 (× 109) platelets/liter) with major hemorrhage or surgical procedure.

Coagulation Disorders Coagulation problems are frequently found in oncology patients. DIC can be caused by the tumor itself whereas other times it may be secondary, such as a gramnegative sepsis associated with leukopenia. Tumors undergoing necrosis can initiate DIC. As cells from pancreatic, lung, and prostate carcinoma as well as Hodgkin's disease degenerate, release of thromboplastins initiates the clotting process and consumes coagulation factors and platelets. Circulating inhibitors in cancer patients have also been described. Some tumors, such as cavernous hemangiomas, on their own cause DIC. Prostate carcinoma can cause primary fibrin(ogen)olysis with subsequent bleeding. In the case of DIC, treatment is directed at the underlying disease. Supportive measures using platelet concentrates may be used if bleeding occurs.

Downloaded by: Universite de Sherbrooke. Copyrighted material.

USE OF PLATELETS IN CANCER PATIENTS—BAYER ET AL

SEMINARS IN THROMBOSIS AND HEMOSTASIS—VOLUME 18, NO. 4, 1992

Structural Lesions At times, some solid tumors present with bleeding diathesis due to their highly vascular nature. Tumors of the gastrointestinal tract such as gastric carcinoma and colon carcinoma can bleed directly into the lumen, producing a life-threatening problem. Solid tumors of the central nervous system (CNS) present special problems. They may bleed by themselves or create problems for the neurosurgeon performing a craniotomy when the platelet count is low. One standard neurosurgery text suggests the platelet count must be above 100,000/µl. 4 1 Minette and Kimmel42 have noted that subdural hematomas are common in patients with solid tumors metastatic to the CNS. Mild trauma and anticoagulants (warfarin) in these patients were the cause of most of the bleeding. In a second group of patients with solid tumors of the CNS, significant bleeding developed due to the metastasis directly invading and destroying the dura. Only a minority of this second group had either thrombocytopenia or coagulation abnormalities. The conclusion may be drawn that dangerous bleeding episodes can occur in patients with brain cancer, primary or secondary.

SUMMARY The need for blood components for oncology patients is small compared with the need for patients with hematologic malignancies. Appropriate use of blood components is necessary, not only medically, but also because of limited supply and availability. Agreement on when to use components is extremely important. In fact, at the time of this writing, the Transfusion Practices Committee of the AABB is conducting an extensive survey on the use of platelets in the oncology and hematology cancer patients (Questionnaire on Institutional Policy on Platelet Transfusion Practice for Hematology/ Oncology Patients). The results will, it is hoped, provide a consensus on the proper times and counts that require prophylactic use of components for these patients. Since these patients use the vast majority of components (see Table 15), their proper use is imperative to maintaining an adequate platelet and frozen plasma supply. Transfusion support in cancer patients is vital for their survival. Platelets, in particular, are necessary to prevent serious bleeding. However, refractoriness to platelet transfusions can develop. It must be appreciated that refractoriness is not a general problem and need not require the expensiveness of a universal decision for handling all platelet transfusions in the same manner. Total refractoriness probably occurs in 15 to 20% of patients frequently transfused. In patients in whom frequent plate-

let transfusion is anticipated, that is, bone marrow transplantation, the development of platelet refractoriness may be reduced by using SDPC and administering them through leukocyte filters. Patients who become refractory to either random or SDPC can either be crossmatched for single-donor platelets that are compatible or can be given HLA-A,B matched platelets. Certainly, the success of platelet transfusion in leukemic patients cannot be denied, since only a small number of these patients now die because of bleeding due to platelet refractoriness. Most of the serious bleeding still seen is associated with sepsis. The risks from transfusion must always be considered. Fortunately, with increased monitoring of the blood supply, they have been reduced. As with any therapeutic regimen, these risks must be weighed against the benefit the patient may gain. Transfusion should always be used prudently.

REFERENCES 1. Slichter SJ: Controversies in platelet transfusion therapy. Annu Rev Med 31:509, 1980. 2. Levin RH: Handbook for platelet transfusion therapy. Special Publication, National Cancer Institute, Bethesda, MD, 1964. 3. Han T, L Stutzman, E Cohen, U Kim: Effect of platelet transfusion on hemorrhage in patients with acute leukemia. Cancer 19:1937, 1966. 4. Walters RS, HM Kantarjian, MJ Keating: Intensive treatment of acute leukemia in adults 70 years of age and older. Cancer 60:149, 1987. 5. Gaydos LA, EJ Freireich, N Mantel: The quantitative relation between platelet counts and hemorrhage in patients with acute leukemia. N Engl J Med 266:905, 1962. 6. Gmur J, J Burger, U Schanz, et al: Safety of stringent prophylactic platelet transfusion policy for patients with acute leukemia. Lancet 338:1223-1226, 1991. 7. Murphy S, S Litwin, P Koch, et al: The indications for platelet transfusion in children with acute leukemia. Clin Res 24:379A, 1976. 8. Solomon J, T Bokefkamp, JL Fahey, et al: Platelet prophylaxis in acute non-lymphocytic leukemia. Lancet 1:267, 1978. 9. Daly PA, CA Schiffer, J Aisner, PH Wiernik: Platelet transfusion therapy. One-hour post-transfusion increments are valuable in predicting the need for HLA-matched preparations. JAMA 243:435, 1980. 10. Dutcher JP, CA Schiffer, J Aisner, PH Weirnik: Long-term follow-up of patients with leukemia receiving platelet transfusions: Identification of a large group of patients who do not become alloimmunized. Blood 58:1007, 1981. 11. Rachel JM, LT Sinor, OW Tawfik, et al: A solid phase red cell adherence test for platelet crossmatching. Med Lab Sci 42:194— 195, 1985. 12. Zeigler ZR, RK Shadduck, CS Rosenfeld, et al: High-dose intravenous gamma globulin improves responses to single-donor platelets in patients refractory to platelet transfusions. Blood 70:1433, 1987. 13. Gmur J, A vonFelten, B Osterwalder, et al: Delayed alloimmunization using random single donor platelet transfusions: A prospec-

Downloaded by: Universite de Sherbrooke. Copyrighted material.

390

14.

15.

16.

17. 18. 19.

20.

21. 22.

23. 24. 25.

26.

27.

28.

tive study in thrombocytopenic patients with acute leukemia. Blood 62:473, 1983. Sniecinski I, MR O'Donnell, B Nowicki, LR Hill: Prevention of refractoriness and HLA-alloimmunization using filtered blood products. Blood 71:1402, 1988. Murphy MF, P Metcalfe, H Thomas, et al: Use of leukocyte-poor blood components and HLA-matched platelet donors to prevent HLA alloimmunization. Br J Haematol 62:529, 1986. Brand A, FHJ Claas, PJ Voogt, MNJM Wasser, JG Eernisse: Alloimmunization after leukocyte-depleted multiple random donor platelet transfusions. Vox Sang 54:160, 1988. Bodensteiner DC: Leukocyte depletion filters: A comparison of efficiency. Am J Hematol 35:184, 1990. Bodensteiner DC: Unpublished data. Slichter SJ, J Deeg, MS Kennedy: Prevention of platelet alloimmunization in dogs with systemic cyclosporine and by UVirradiation or cyclosporine-loading of donor platelets. Blood 69: 414, 1987. Bick RL: Disseminated intravascular coagulation and related syndromes: Etiology, pathophysiology, diagnosis and management. Am J Hematol 5:265, 1978. Roseman B: Disseminated intravascular coagulation: A review. Oral Surg Oral Med Oral Pathol 59:551, 1985. Gralnick HR, E Abrell: Studies of the procoagulant and fibrinolytic activity of promyelocytes in acute promyelocytic leukemia. Br J Haematol 24:89, 1973. Groupman J, LE Ellman: Acute promyelocytic leukemia. Am J Hematol 7:395, 1979. Kantarjian HM, MJ Keating, R Walters, et al: Acute promyelocytic leukemia. Am J Med 80:789, 1986. Hoyle CF, DM Swirsky, L Freedman, FGJ Hayhoe: Beneficial effect of heparin in the management of patients with APL. Br J Haematol 68:283, 1988. McKee LC, RD Collins: Intravascular leukocyte thrombi and aggregates as a cause of morbidity and mortality in leukemia. Medicine (Baltimore) 53:463, 1974. Haskell CM, GP Carellos, BG Leventhal, et al: L-asparaginase: Therapeutic and toxic effects in patients with neoplastic disease. N Engl J Med 281:1028, 1969. Ahr DJ, SJ Scialla, DB Kimball: Acquired platelet dysfunction following mithramycin therapy. Cancer 41:448, 1978.

391

29. Kanz L, A Linemann, W Oster, et al: Hemopoietins in clinical oncology. Am J Clin Oncol 14:527-533, 1991. 30. Belt RJ, C Leite, CD Haas, et al: Incidence of hemorrhagic complications in patients with cancer. JAMA 239:2571-2574, 1978. 31. Heyman MR, CA Schiffer: Platelet transfusion to patients receiving chemotherapy. In: Rossi EC, Simon TL, Moss GS, eds: Principles of Transfusion Medicine. Williams & Wilkins, Baltimore, 1991, pp 223-232. 32. Collins C, CS Higano, RB Livingston, et al: Cyclophosphamide, vincristine, cisplatin, VP-16 and radiation therapy in extensive small-cell lung cancer. Cancer Chemother Pharmacol 24:128-132, 1989. 33. Fukuoka M, M Takada, S Negoro, et al: Alternating non-cross resistant chemotherapy for small cell lung cancer. Jpn J Clin Oncol 16:261-270, 1986. 34. Harding M, R Docherty, R Mackie, et al: Phase II studies of mitozolomide in melanoma, lung and ovarian cancer. Eur J Cancer 25:785-788, 1989. 35. Kreisman H, M Goutsou, C Modeas, et al: Cisplatin-carboplatin therapy in extensive non-small cell lung cancer: A Cancer and Leukemia Group B Study. Eur J Cancer 26:1057-1060, 1990. 36. Planting AS, A Ardizzoni, J Estape, et al: Phase II study of ACNU in non-small-cell lung cancer: EORTC study 08872. Cancer Chemother Pharmacol 28:145-146, 1991. 37. Pujol J, J Rossi, T Le Chevalier, et al: Pilot study of neoadjuvant ifosfamide, cisplatin, and etoposide in locally advanced non-small cell lung cancer. Eur J Cancer 26:798-801, 1990. 38. Vokes EE, JD Bitran, PC Hoffman, et al: Neoadjuvant vindesine, etoposide, and cisplatin for locally advance Non-Small Cell Lung Cancer. Chest 96:110-113, 1989. 39. Young CW, MP Fanucchi, TD Walsh, et al: Phase I trial and clinical pharmacological evaluation of hexamethylene bisacetamide administration by ten-day continuous intravenous infusion at twenty-eight-day intervals. Cancer Res 48:7304-7309, 1988. 40. Panella TJ, W Peters, JG White, et al: Platelets acquire a secretion defect after high-dose chemotherapy. Cancer 65. 1990. 41. Plapp FV, WL Bayer: Blood transfusions. In: Wilkins RH, Rengachary SS, eds: Neurosurgery Update. McGraw Hill, New York, 1985, pp 372-383. 42. Minette SE, DW Kimmel: Subdural hematoma in patients with systemic cancer. Mayo Clin Proc 64:637-652, 1989.

Downloaded by: Universite de Sherbrooke. Copyrighted material.

USE OF PLATELETS IN CANCER PATIENTS—BAYER ET AL

Use of platelets and other transfusion products in patients with malignancy.

The need for blood components for oncology patients is small compared with the need for patients with hematologic malignancies. Appropriate use of blo...
2MB Sizes 0 Downloads 0 Views