Hemapheresis System

Hayden G. Braine, M.D.1

Introduction Historically, the major transfusion requirements for cancer treatment have been for blood and plasma products used in surgery. The relatively long storage time of these products, 35 days for red cells and 1 year for fresh frozen plasma, have made management of inventories of these products amenable to relatively simple data systems. Many hospital blood bank inventory systems have been developed to handle these products. Likewise, systems for the management of the large-scale donor center manufacture and testing of blood components are also available. However systems to assist in the management of platelet concentrates with 1- to 5-day storage times have not been so well developed.

Platelets are small cellular fragments that circulate in the bloodstream. They are primarily responsible for controlling bleeding by initiating thrombosis. Deficiencies in platelet numbers usually result in an increased risk of spontaneous or posttraumatic bleeding. Prior to the development of platelet transfusion therapy, 50% of the patients treated for acute lymphocytic leukemia died of hemorrhage; with effective platelet transfusion technology this was reduced to less than 5%.

Effective platelet transfusion therapy has also allowed the development of intensive curative treatment regimens in many other hematopoietic malignancies. The development of these treatment regimens has increased the use of platelets dramatically in the last decade. In general, most platelet transfusion support has been concentrated in tertiary care hospitals; in 1980 1000 such tertiary centers accounted for two-thirds of the platelet transfusions in the United States.

‘Hayden G. Braine, M.D., joined the Oncology Center in 1976 and established the Hemapheresis and Transfusion Service. During its inception, Dr. Braine initiated work on the automation of this service and its integration with the OCIS. Dr. Braine is an Associate Professor in Oncology with a special interest in the transfusion of cancer patients. He holds joint appointments in both Medicine and Laboratory Medicine at Johns Hopkins, and is presently the Director of the Hemapheresis and Transfusion Services at the Center.

Figure 1. Platelet transfusion planning flow.

The JHOC administers 7900 platelet transfusions annually. This chapter describes the systems used to manage the JHOC platelet transfusion service.

System Overview Management of patient platelet transfusion requirements is a repetitive cycle of six sequential steps (Figure 1). The challenge in this flow is to match the clinical transfusion requirement with product availability. When only 5 or 6 patients are being supported at one time, conventional manual methods of information flow can support timely decision making. However, the JHOC services daily over 50 patients potentially requiring a platelet transfusion. These patients are geographically distributed in five inpatient services and three outpatient departments. In this setting, computer systems are essential for the accurate formatting of data for timely and appropriate decisions.

Step 1. Analysis of Transfusion Outcome There are two major considerations in evaluating the outcome of a platelet transfusion: the extent to which the patient’s platelet count is increased (increment) and the length of time the transfused platelets persist in the patient’s circulation (survival). The observed increment (posttransfusion platelet count — pretransfusion platelet count) can be normalized for the number of platelets transfused (# of units) and the volume of the patient’s vascular system [estimated by the patient’s body surface area in square meters (m2)]:

Figure 2. Sample comprehensive platelet transfusion history.

(platelet increment) (m2) =

(# of units) ’

Thus, immediately after the transfusion, 1 unit of platelets should raise the platelet count of a 1 m2 patient by 10,000 platelets/mm3. By normalizing the post transfusion platelet count in this manner, the outcome of platelet transfusion can be compared among various patients of varying sizes receiving varying doses of platelets. Using a definition of 1 platelet unit as 5.5 x 1010 platelets, our “normal” posttransfusion increment is 10,000 ± 5000 (±1 SD).

By monitoring patients’ platelet counts in relation to transfusion, we can determine whether we have achieved the desired platelet count and, after considering the number of units transfused, whether the observed increment in platelet count was that which was expected. Failure to achieve the expected immediate posttransfusion increment can result from several factors, including formation of antibodies to tissue (HLA) antigens on the platelets, organomegaly, or intravascular coagulation.

Similarly, survival of transfused platelets can be calculated: 18 to 24 hours after the transfusion the corrected increment should be greater than 2500. Failure to observe a normal posttransfusion platelet survival can be caused by a variety of clinical variables, including fever, infections, and bleeding.

The complex clinical and laboratory database required to interpret logically the posttransfusion platelet increment and survival is organized and maintained on line in OCIS as “the platelet transfusion history” (Figure 2). Pertinent data for evaluating platelet transfusion outcome include each patient’s peak daily temperature (TEMP) and bleeding and infection status (B/I). Comments on other clinical variables, such as splenomegaly or disseminated intravascular coagulation, are noted in the “header” or margin, that is, in the upper left-hand corner or right-hand margin of the report. Product variables that affect transfusion outcome, such as ABO type (ABO), storage age (A), HLA type, and white cell (LE9) and platelet content (#U), are displayed. Posttransfusion corrected platelet increments are calculated for each transfusion and formatted with the clinical and laboratory information.

The example shown in Figure 2 is that of a 63-year-old white woman with Wal-dendrom’s Macroglobulinemia. She was treated with combination chemotherapy on 13 January 1988 and was on protocol day (PD) 43 posttherapy on 24 February 1988. At 9:10 a.m. her platelet count (PL) was 12,000/mm3, white count (WBC) 40,200/mm3, and hematocrit (HC) 32%. At 12:00 p.m. that day she received 12 units (P12) of platelets with HLA type A29, A30, B7, BW6, 6. One hour later her platelet count was 9000/mm3, an obvious failure. At 7:05 a.m. the next day her platelet count was 5000/mm3, and at 11:00 a.m. she received a second transfusion of 8 units of platelets with HLA type A3, A24, B7, B22, BW 6, 6. Two hours (12:50 p.m.) after transfusion her platelet count was 73,000/mm3. This gave a normalized platelet increment of 14,110 two hours after the transfusion (14,110/2). Clearly a successful transfusion. Data not shown in this example implicated antibodies to A29 and A30 as the cause of the first transfusion failure.

Partial HLA matching program for patient with HLA Type A3, A30, B18, B Figure 3. Partial listing from HLA matching program. This list contains donor matches for a patient with HLA Type A3, A30, B18, B8.

Figure 4. Sample morning donor room schedule. This is the schedule for the week of 4 April 1988. Ten cell separators are available. Donors (indicated by their HLA type) are scheduled to arrive between 7:30 a.m. and 8:30 a.m.

Figure 5. Sample schedule for HLA matches. Patient’s (names deleted) location and HLA type indicated on left hand column.

Step 2. Selection of Platelet Products Required Each week a transfusion plan is established for all patients on platelet transfusion support. Frequency of transfusion requirement, product type (concentrates made from a single donor by apheresis or 6 to 10 units pooled from individual units of whole blood), HLA type, and CMV serology are considered. From these individual projections a master transfusion plan is prepared. Special requirements for products without risk for cytomegalovirus are also considered.

Step 3. Donor Selection The JHOC maintains a volunteer blood donor program of over 1200 members. Individual donor files are maintained and include information on HLA, ABO, and Rh type; cytomegalovirus serology; donation history, and availability for donations.

Patients requiring specific HLA types are matched with available donors (Figure 3). In this example a patient with HLA type A3, A30, B18, B8 is found to have two excellent matches (A and B1U) and many close matches (BIX, B2UX, and B2X). (Donor names and IDs have been deleted.) Subprograms then display the schedule for the hemapheresis donor room (Figure 4) and the transfusion coordinators (Figure 5).

Step 4. Platelet Production In order to support the platelet transfusion needs of the JHOC, the Hemapheresis Center operates a large platelet collection program. Platelets are harvested from donors using blood cell separators; each donation produces 8 to 10 units of platelets during a 2- to 3-hour donation.

Each product prepared is entered into the hemapheresis database and includes data on the product’s unique identifying number, volume, platelet content, white cell content, HLA type, red cell type, machine operator, and cell separator number. This database is then used for generating productivity statistics of the Hemapheresis Unit, as well as assisting in preventative maintenance of the cell collection equipment. Data collected on each cell separator are used to calculate average platelet yields and operating efficiency. (Figure 6). In the report shown, one can see that the CS3000 cell separators consistently produce fewer numbers of platelets (ranging from 3.8 to 4.1 X 1011) with a lower collection efficiency (ranging from 41% to 44%) than the Model-30 system, in terms of both absolute yield (ranging from 4.5 to 4.9 x 1011) and collection efficiency (ranging from 56% to 58%).

Step 5. Inventory All platelet concentrates produced are entered into a master on-line inventory (Figure 7). This is then used by the clinical transfusion coordinators to assign PLATELET YIELD - ALL 08/01/87 TO 08/31/87

(PI,PntWC DONATIONS NOT INCLUDED)

MACH. NAME

#

#DONS.

AVE.PLT. COUNT QUO”) PER DON.

STD.

DEV.

PLT.

EFF.

STD. # DEV.

CS3000

1070

8

3.9

0.7

0.41

0.06 8

CS 3000

1159

19

4.2

0.9

0.44

0.05 19

CS3000

1247

24

4.4

0.9

0.43

0.04 20

CS 3000

1605

31

3.8

1.0

0.40

0.04 30

CS 3000

1609

30

4.2

1.1

0.42

0.07 22

CS 3000

1613

41

3.8

0.8

0.43

0.05 38

CS 3000

1615

40 .

4.1

0.9

0.42

0.05 36

MODEL 30

1520

NOT USED

MODEL 30

130

NOT USED

MODEL30

238

NOT USED

MODEL 30

498

12

4.7

1.2

0.57

0.06 12

MODEL 30

590

25

4.8

1.2

0.58

0.08 25

MODEL 30

739

24

4.9

1.5

0.57

0.09 24

MODEL 30

975

9

4.5

1.0

0.56

0.06 8

-50

1090

NOT USED

-50

1555

16

3.5

1.0

0.42

0.13 16

Figure 6. Platelet yield and operating efficiency by machine.

products to recipients for final transfusion assignment. The inventory also operates as a central quality control step. All products, each assigned a unique sequential number, remain in inventory unless assigned to one of five outcomes:

1. Transfused into a patient, in which case the data are entered into the patient’s database and used to generate transfusion history data, as well as a billing file for administrative purposes.

2. Used for quality control and destroyed.

3. Destroyed because of a positive serology for hepatitis, AIDS, etc.

4. Outdated and destroyed. Data in this and the previous category are reported monthly.

5. Frozen for future use, in which case the product is entered into a master inventory of frozen platelet concentrates; this frozen inventory is also searched for each HLA matching program ordered.

EXPIRES 04/07/88

PRODUCT

TIME

UNITS

LE9

VOL

RECIPIENT

BT/CM

HLA

R24714P

950A

8

287

0+/

A28,19 B7,40 BW6,6

R24717PX0

145P

6

135

B+/-

A25,31 B18,60 BW6,6

R24718P

200P

9

279

0+/

A2,11 B8,14 BW6,6

SS18189PX

1053A

6.73

0.3

221

0+/+

A1,26 B8,45 BW6,6

SS18191PAB

1115A

5.09

0.6

142

A-/+

A3, B7,44 BW6,4

SS18192PAD

1135A

6.18

0.04

430

A+/-

A2,29 B8,14 BW6,6

SS18193PX

145P

8.73

0.5

223

A+/+

A1,2 B17,62 BW4,6

SS18194PX

144P

8.36

0.3

222

A+/-

A2,11 B44, BW4,4

EXPIRES 04/08/88

PRODUCT

TIME

UNITS

LE9

VOL

RECIPIENT

BT/CM

HLA

EXPIRES 04/09/88

PRODUCT

TIME

UNITS

LE9

VOL

RECIPIENT

BT/CM

HLA

R24699PY0

1105A

11

222

A+/-

A29,31 B22,44 BW6,4

R24701PY0

225P

5

160

0+/-

A2,3 B35,60 BW6,6

EXPIRES 04/10/88

PRODUCT

TIME

UNITS

LE9

VOL

RECIPIENT

BT/CM

HLA

R24710PY0

255P

6

180

B+/-

A2,3 B7,8 BW6,6

EXPIRES 04/11/88

PRODUCT

TIME

UNITS

LE9

VOL

RECIPIENT

BT/CM

HLA

R24713PY0

945A

6

215

A+/-

A1,24 B18,37 BW6,4

R24715PY0

1026A

9

219

B+/-

A2,24 B13,27 BW4,4

EXPIRED Master Inventory Figure 7.

With this control step any delay in data entry or other nonstandard outcome is rapidly known: Products not assigned a destination are easily detected as they remain in inventory. Similarly, products not entered into the system are detected, as a sequential number is missing.

Figure 8.

Step 6. Transfusion All platelet products produced locally or ordered from other bloos centers are stored in the Hospital’s Central Blood Bank. The patient’s individual physician is responsible for ordering the transfusion of the product and specifying any secondary processing, such as leukocyte depletion to prevent febrile transfusion reactions. To assist the physician in ordering platelets, on-line files are maintained with specific recommendations for transfusion (Figure 8). Files also include transfusion reaction history and recommendations for any pretransfusion medications needed.

Step 7. Reanalysis of Transfusion Outcome With completion of each transfusion, analysis of the outcome is again reviewed as in Step 1. Unexpected outcomes are then identified and alternate transfusion approaches ordered.

Quality Control of Platelet Transfusion Outcome The American Hospital Commission (AHC) requires review of all transfusions on a regular basis. At the JHOC’s current rate of nearly 8000 transfusions per year, this would be a logistical nightmare. However, with daily review of the computer-based transfusion histories, the AHC standards can be met in a timely manner. More important, modification of transfusion practice can be achieved in real time rather than retrospectively.

Transfusion practices are also traditionally monitored by review of departmental utilization statistics over time. However, departmental trend analyses have two major deficiencies: First, they do not adjust utilization for volume. A doubling of platelet utilization could be accounted for by a doubling of use per patient or a doubling of the number of patients transfused. Second, utilization statistics do not address the question of the quality of care. Platelets are transfused to prevent or stop bleeding. In order to evaluate the appropriateness of a given number of transfusions, one must consider whether hemostasis was maintained.

To address this problem, we have implemented the case-adjusted statistical evaluation (CASE). Patients are grouped by diagnosis and treatment; for example, all patients with acute myelocytic leukemia (AML) treated with a particular combination of drugs are grouped, and then daily platelet counts, along with platelet transfusion and bleeding status data, are acquired from the OCIS database. Mean daily platelet counts, the mean number of units of platelets transfused per patient per day, and the mean bleeding status of the group are reported. Trends of higher platelet counts with fewer transfusions and less bleeding would be desirable. For this evaluation all patients are scored daily on a scale of 0 to 4 (Table 1) for bleeding status.

If comparisons among patient populations are to be made, platelet transfusion requirements must also be adjusted for patient size: Larger patients require more platelets than smaller patients. This can be accomplished by dividing utilization by mean patient size (m2). Figure 9 displays the quality control report for adult patients undergoing induction chemotherapy (Ac-D-AMSA) for acute myelocytic

Grade

Code

0

1

2

3

4

Bleeding status

No symptoms

Petechia

Epistaxis, venipuncture sites

More than 1 unit per day due to hemorrhage

Life-threatening symptoms

Infection status

No symptoms

Complete response—on antibiotics, afebrile, sites stable

Partial response—on antibiotics, previous plus culture now, or decreasing T max

No change

Progression

Figure 9. Case report for patients treated for AML with chemotherapy regimen 18410 1984-87. Data on first line of each year includes all data points: Data on second line includes data only for days WBC<500/mm3.

Figure 10. Individual bleeding and infection profiles: number of days at each score/number of patients. Acute days are days WBC<500/mm3.

leukemia (AML). From 1984 to 1987 total platelet use per patient per square meter fell from 125.6 to 99.4 units pt/m2; at the same time the mean bleeding score for all hospital days remained unchanged at 0.4 + 0.7(1 SD). Such “average” bleeding scores, however, could mask a small number of serious bleeding episodes. In Figure 10 data are reported by individual bleeding score: Of the 51 patients hospitalized for a total of 1930 patient days in 1987, all 51 had at least

1 day of grade 0 bleeding, and life-threatening bleeding (grade 4) was limited to

2 patients, each having 1 day of potentially life-threatening bleeding. Only 6 of 51 patients experienced grade 3 or worse bleeding.

Such data still are not controlled for the actual need for platelet transfusion. In general, patients receiving intensive treatments for hematologic malignancies develop thrombocytopenia as the result of bone marrow failure. It is during these periods of marrow failure that platelet transfusion is routinely required to prevent hemorrhage. Therefore, one would like to evaluate platelet utilization on the basis of platelet transfusion need or, in the case of patients treated for hematologic malignancies, on the degree and duration of marrow failure. In theory, the degree and duration of marrow failure could be estimated by measuring the number of days a patient has a platelet count below a given number. However, this is not a true estimate of platelet transfusion need, as platelet transfusion abolishes thrombocytopenia. Therefore, we have used the daily white blood cell count, which is not affected by transfusion, to estimate days of bone marrow failure. Following chemotherapy, when patients have fewer than 500 leukocytes/mm3 (termed acute days), they usually have severe enough marrow failure to require platelet transfusion. The platelet utilization per patient can then be corrected for the days of marrow failure by dividing platelet use by patient days of fewer than 500 leukocytes.

Controlling for such “acute days” of aplasia, we note that the average number of units of platelets used per patient per square meter per day on which the white blood cell count is less than 500 has fallen from 3.6 ± 2.8(1 SD) in 1984 to 2.9 ± 2.1 in 1987 (Figure 9). Again, the mean bleeding score remained unchanged at 0.5 ±0.8, and the incidents of serious (grade 4) bleeding were few in number: In 1987, 2 patients out of 51 at risk experienced one day each of grade 4 hemorrhage (Figure 10). Thus it would appear that improved transfusion efficiency (fewer units) has been achieved without significantly increased toxicity (grade 3 or 4 hemorrhage).

Conclusion Traditionally blood banking medicine has concentrated on product quality, serologic matching, and inventory control. Increasingly, however, blood banking is becoming a complex transfusion science, with increasingly effective transfusion options being applied to complex medical treatments. This requires that an accurate and timely database be maintained in order that appropriate decisions be made and that an ongoing evaluation of transfusions be performed to ensure the delivery of the right care.

In the early 1970s platelets were selected only with regard to how many units were needed. Inventory was usually low and triage of units was common. Transfusion monitoring was limited to evaluation of next-day platelet counts.

Today’s platelet transfusion science delivers a safer and more effective product. Concentrates prepared by hemapheresis can be HLA matched to the donor. Secondary product washing and/or leukocyte depletion can reduce the risk of transfusion reaction. Serologic screening can reduce the risk of disease transmission. Effective use of this powerful tool is one of the challenges of transfusion medicine in the 1990s. The availability of a timely and accurate laboratory and clinical database and an ongoing evaluation of transfusion practices are essential. In the case of JHOC, an expansion of OCIS to meet these special needs has allowed us to provide effective service to a large number of patients at a reasonable cost using limited resources.

Рекомендуем к просомтру

www.kievoncology.com благодарны автору и издательству, которые способствует образованию медицинских работников. При нарушении авторских прав, сообщите нам и мы незамедлительно удалим материалы.