THERAPEUTIC PLASMA EXCHANGE FOR MANAGEMENT OF HEPARIN-INDUCED THROMBOCYTOPENIA: RESULTS OF AN INTERNATIONAL PRACTICE SURVEY

Abstract

Introduction:

Anti-heparin/platelet factor 4 antibody immune complexes resulting from heparin-induced thrombocytopenia (HIT) are removed by therapeutic plasma exchange (TPE). We sought to define TPE in HIT practice patterns using an international survey.

Methods:

A 31-item online survey was disseminated through the American Society for Apheresis. After institutional duplicate responses were eliminated, a descriptive analysis was performed.

Results:

The survey was completed by 94 respondents from 78 institutions in 18 countries. Twenty-nine institutions (37%) used TPE for HIT (YES Cohort) and 49 (63%) did not (NO Cohort). Most NO respondents (65%) cited “no requests received” as the most common reason for not using TPE. Of the 29 YES respondents, 10 (34%) gave incomplete information and were excluded from the final analysis, leaving 19 responses. Of these, 18 (95%) treated ≤ 10 HIT patients over a 2-year period. The most common indications were cardiovascular surgery (CS; 63%) and HIT-associated thrombosis (HT; 26%). The typical plasma volume processed was 1.0 (63% CS and 58% HT). For CS, the typical replacement fluid was plasma (42%) and for HT, it was determined on an individual basis (32%). For CS, patients were treated with a set number of TPE procedures (37%) or laboratory/clinical response (37%). For HT, the number of TPE procedures typically depended on laboratory/clinical response (42%).

Conclusion:

In a minority of responding institutions, TPE is most commonly used in HIT to prophylactically treat patients who will undergo heparin re-exposure during cardiovascular surgery. Prospective studies are needed to more clearly define the role of TPE in HIT.

INTRODUCTION

Heparin-induced thrombocytopenia (HIT) is a life-threatening thrombotic disorder commonly caused by the formation of anti-heparin/platelet factor four (PF4) antibody immune complexes.¹ Once diagnosed, the widely accepted standard of care involves cessation of the offending anticoagulant and initiation of a parenteral direct thrombin inhibitor (DTI);² however, despite DTI use, HIT complications continue to occur with high rates of death, limb amputation, thrombosis and bleeding.³ To minimize HIT-related complications, additional therapies are needed that address the underlying pathophysiology of formation of anti-heparin/ PF4 antibody immune complexes.

These large intravascular anti-heparin/PF4 antibody complexes can be removed using therapeutic plasma exchange (TPE). In fact, TPE has been demonstrated, in case reports and case series, to be effective in decreasing anti-heparin/PF4 antibodies, treating acute HIT complications, and facilitating heparin re-exposure in patients with acute or subacute HIT.^4–9

Although TPE is used in the management of HIT, its extent of use and treatment practice patterns are unknown. Therefore, the HIT Subcommittee of the American Society for Apheresis (ASFA) Research Committee conducted a practice patterns survey to determine the extent of TPE use in HIT and to understand the parameters by which TPE is applied.

MATERIALS AND METHODS

Survey Development

Members of the ASFA HIT Subcommittee met monthly over a 6-month period to develop a survey to evaluate current apheresis management practices for patients with HIT. Survey items were developed using a combination of consensus expert opinion and review of the literature. Once questions were developed, they were entered into an online survey tool, SurveyMonkey (SurveyMonkey, Inc., San Mateo, CA). The survey was then tested by the HIT Subcommittee members with questions added, removed, or modified for clarity. The final product was a 31-item survey with the following characteristics: seven questions dedicated to demographic information; 21 questions designed to capture details of TPE in HIT treatment; and 3 questions designed to capture details of respondents that do not use TPE in HIT. The full text of the survey questions is detailed in the Appendix. The study was approved by the Duke University School of Medicine Institutional Review Board.

Survey Dissemination

The study population included ASFA registered members and ASFA-interested parties that subscribe to the ASFA listserv. Containing 3,667 email addresses, the ASFA listserv was used to disseminate the survey. The survey invitation’s initial distribution was done on April 23, 2018, two days before the 2018 ASFA Annual Meeting. Second and third requests were made via email one and two weeks following the initial invitation. In addition to the electronic solicitation, ASFA members attending the 2018 Annual meeting received an informational flyer in their meeting materials. The online survey remained open for 21 days.

Treatment of Duplicate Responses

Although individual responses were collated, only one response per institution was recorded. When duplicates arose, the preferred respondent was determined based on the following hierarchy: 1) Apheresis Medical Director; 2) Apheresis Attending Physician; and 3) Nurse Manager. In situations where more than one respondent met equivalent hierarchal criteria (e.g., data from two attending physicians), we reviewed the date and time of survey completion and retained data only from the earliest respondent with a completed survey. In accordance with the above criteria, institutional duplicate responses were discarded and not included in the final analysis.

Cohort Determination

Respondents were separated into two distinct cohorts based on their response to Question Seven: “Has your Center performed TPE for HIT in the last five years?” Respondents answering in the affirmative were designated the “YES TPE” cohort, while respondents answering in the negative were designated the “NO TPE” cohort. Respondents in the “YES TPE” cohort answered an additional 21 questions beyond the initial seven, while respondents in the “NO TPE” cohort answered an additional three questions (See survey items in the Appendix).

Statistical Analysis

Descriptive statistical analysis was performed using Statistical Analysis System 9.4 (SAS Institute Inc., Cary, NC). Results are presented as the count and percent (%) of respondents for each question. Due to rounding, frequency may not equal exactly 100%.

RESULTS

Overall Results

The survey was completed by a total of 94 respondents from 78 institutions in 18 countries. The majority of responses came from the United States (USA), which had 70 respondents from 55 institutions in 28 states. Demographics of all survey respondents are shown in Table 1.

Table 1.

Demographics of All Survey Respondents without Duplicates Removed

	Number	%
Institution Location
United States	70	74
Outside the United States	24	26
Medical Subspecialty
Blood banking/Transfusion medicine	53	56
Hematology/Oncology	18	19
Nephrology	12	13
Anesthesia/Critical Care	3	3
Apheresis unspecified	3	3
Neurology	2	2
Other	3	3
Type of Provider
Apheresis Medical Director	37	39
Nurse/Nurse manager	36	38
Apheresis Attending Physician	14	15
Resident/Fellow	3	3
Other	4	4
Adult or Pediatric Provider*
Adult only	26	29
Pediatric only	11	12
Both adult and pediatric	54	59

^*For this question, n = 91

As demonstrated in Figure 1, once duplicates were removed, the number of unique institutional respondents was 78. Twenty-nine respondents (37%) answered in the affirmative with regard to TPE use (the YES TPE cohort) while 49 respondents (63%) indicated that TPE was not used (the NO TPE cohort). Respondents in the YES TPE cohort represented 29 institutions in 7 countries including Canada (1), Greece (1), Iraq (1), Russia (1), United Kingdom (1), USA (23), and Vietnam (1). Respondents in the NO TPE cohort represented 49 institutions from 14 countries, including Argentina (1), Australia (1), Brazil (1), Canada (2), China (1), Egypt (1), India (1), Israel (1), Italy (3), Lebanon (1), Qatar (1), South Africa (1), United Kingdom (2), and USA (32). Differences between the “YES” and “NO” TPE cohorts are outlined in Table 2.

Figure 1.

The total number of survey respondents was 94. Based on pre-determined criteria (outlined in the methods), 16 duplicate responses from the same institution were removed. The resulting number of unique institutional responses was 78. Twenty-nine institutions indicated that TPE was used for the management of HIT, and 49 indicated that TPE was not used for HIT. Of the 29 that indicated TPE use for HIT, 10 provided incomplete responses, leaving a total number of 19 respondents’ data for analysis.

Table 2.

Survey Respondent Characteristics After Duplicates Removed

	YES Cohort n = 29		NO Cohort n = 49
	Number	%	Number	%
Institution Location
United States	23	79	32	65
Northeast	4	14	8	16
Midwest	9	31	10	20
South	6	21	8	16
West	4	14	6	12
Non-US Regions	6	21	17	35
Africa	0	–	2	4
Asia	2	7	5	10
Europe	3	10	4	8
North America	1	3	2	4
South America	0	–	3	6
Oceania	0	–	1	2
Estimated Number of Cardiovascular Surgery Procedures Performed Annually
0	4	14	6	12
< 50	1	3	1	2
50–100	1	3	7	14
101–500	5	17	12	25
500–1000	5	17	4	8
> 1000	6	21	5	10
Unsure	7	24	14	29
Non-TPE Therapies Used*
Parenteral Direct thrombin inhibitors (DTI)	15	–	17	–
Direct oral anticoagulants (DOACs)	9	–	11	–
Fondaparinux	9	–	10	–
Intravenous immune globulin (IVIG)	7	–	10	–
Other†	2	–	0	–

N = number; TPE = Therapeutic plasma exchange;

^*Percentages were not calculated due to respondents’ ability to select all that apply.

^†Steroids and warfarin in the YES TPE cohort.

Practice Patterns in the NO TPE Cohort

In the NO TPE cohort (n=49), reasons given for not using TPE, in order of decreasing frequency (respondents were asked to select all that apply) were as follows: No requests received (n=32, 65%); institutional preference (n=18, 37%); preference for alternative therapies (n=8, 16%); lack of data (n=5, 10%); and denial of requests (n=1, 2%). The NO TPE respondents used the following HIT therapies in decreasing order of frequency, as shown in Table 2: parenteral direct thrombin inhibitors (DTI, n=16; 33%); direct oral anticoagulants (n=11, 23%); fondaparinux (n=10, 21%); and intravenous immune globulin (n=10, 21%).

Practice Patterns in the YES TPE Cohort

Of the 29 respondents that reported TPE use in the treatment of HIT, 10 (34%) did not provide additional information. Therefore, the below results attributed to the YES TPE Cohort below are based on information provided by the remaining 19 respondents (66%).

Number of Patients Treated Over a 24-month Period.

With regard to the number of patients treated over a 2-year period, 95% of respondents (n=18) indicated that they treated 10 or fewer patients. The most frequent range of patients treated was 2–5 (n=10; 53%); followed by 0–1 (n=4; 21%); 6–10 (n=4; 21%); and 11–15 (n=1; 5%).

Clinical Diagnostic Tests.

Respondents reported that a number of parameters were used to initiate TPE (respondents were asked to select all that apply), including the following: positive anti-heparin/PF4 antibody test result (n=15; 79%); HIT predictive score showing high probability (n=8; 42%); positive serotonin release assay (n=8; 42%); HIT predictive score showing moderate probability (n=5; 26%); clinical suspicion alone (n=5; 26%); HIT predictive score showing low probability (n=2; 11%); positive HIT aggregation test (n=1; 5%); and hematology referral (n=1; 5%). The most commonly used predictive score was the 4Ts score¹⁰ (n=14; 74%) followed by using both the 4Ts score and the HIT Expert Probability Score¹¹ (n=3; 16%).

TPE Treatment Indication.

In decreasing order of frequency (respondents were asked to select all that apply), respondents reported using TPE for any of the following indications: patients with HIT undergoing cardiovascular surgery (n=13; 68%); patients with HIT and thrombosis (n=10; 53%); patients with HIT without thrombosis (n=2; 11%); patients with HIT undergoing non-cardiovascular surgery (n=1; 5%); and patients with HIT and renal complications (n=1; 5%). When asked, “What is your Center’s most frequent indication?”, the most common choice selected by respondents was patients with HIT undergoing cardiovascular surgery (n=12; 63%), followed by HIT with thrombosis (n=5; 26%).

TPE Treatment Parameters.

Four respondents indicated that they have a standard operating procedure for HIT TPE treatment (n=4; 21%), but most respondents reported that they do not (n=15; 79%). TPE treatment details of number of procedures, plasma volumes processed, and replacement fluid utilized are shown in Table 3. Respondents indicated that the most common volume processed was 1.0 plasma volume. With regard to the number of TPE procedures, most respondents (n=18; 95%) indicated that they based TPE number on laboratory or clinical parameters. Laboratory parameters used, in decreasing order of frequency (respondents were asked to select all that apply), included the following: anti-heparin/PF4 antibody results (n=14; 78%); platelet counts (n=8; 44%); serotonin release assay (n=4; 22%); other immunologic or functional assay (n=4; 22%); and active thrombosis and additional heparin exposure (n=1; 6%).

Table 3.

Parameters of TPE Treatment in Acute HIT with Thrombosis and Prophylactic Treatment Prior to Surgery (n = 19)

	Acute HIT with thrombosis		Prophylactic Treatment Prior to Surgery
	Number	%	Number	%
Number TPE Procedures
1 TPE	0	0	2	11
2–4 TPE	1	5	3	16
5 TPE	4	21	1	5
> 5 TPE	1	5	1	5
No set number (treat to laboratory parameter or clinical response)	8	42	7	37
Other*	5	26	5	26
Plasma Volumes Processed
1.0	11	58	12	63
1.5	5	26	5	26
Other*	3	16	2	11
Typical Replacement Fluid
Albumin only	1	5	1	5
Plasma only	4	21	8	42
Albumin and plasma	4	21	7	37
Determined on an individual patient basis	6	32	1	5
Other*	4	21	2	11
Timing of Procedure	–	–
Pre-operatively only	–	–	9	47
Both pre- and intra-operatively	–	–	3	16
Intra-operatively (in the operating room) only	–	–	2	11
Both intra- and post-operatively	–	–	1	5
Pre- and post-operatively only	–	–	1	5
Pre-, intra-, and post-operatively	–	–	1	5
Other*	–	–	2	11

TPE = therapeutic plasma exchange;

^*This category included two respondents who indicated that TPE was not used in either scenario but did not further clarify.

Treatment Parameters for Acute HIT with Thrombosis.

In patients with a diagnosis of acute HIT in the setting of confirmed thrombosis, most respondents indicated that the number of TPE procedures was adjusted based on laboratory parameters or clinical response (n=8; 42%). Decisions regarding albumin or plasma fluid replacement were most commonly determined on an individual patient basis (n=6; 32%). Respondents indicated that modification of the TPE treatment plan could occur if any of the following were present: clinical severity; new or worsening thrombosis; treatment refractoriness; increasing antibody levels; changing clinical status; or poor tolerance to prescribed volume.

Treatment Parameters for Prophylactic Treatment Prior to Surgery.

Treatment parameters for patients without a thrombotic episode who required re-exposure to heparin during surgery are outlined in Table 3. The same number of respondents indicated that either 1) the number of TPE procedures was pre-set (n=7; 37%) or 2) the number of procedures was determined by laboratory or clinical response (n=7; 37%). TPE treatment was most likely to occur only pre-operatively (n=9; 47%). The most common replacement fluid was plasma (n=8; 42%). In decreasing order of frequency (respondents were asked to select all that apply), plasma replacement was used for one or more of the following reasons: to replace coagulation factors (n=11; 58%); prevent coagulopathy (n=10; 53%); decrease thrombotic risk (n=7; 37%); minimize therapeutic anticoagulation-associated bleeding risk (n=1; 5%); and allow plasma IgG to compete with HIT antibodies for FcγRIIa binding/platelet activation (n=1; 5%).

Respondents noted that modification of the TPE treatment plan could occur with any of the following: in patients undergoing heart transplantation who had a high HIT antibody titer; intolerance to prescribed plasma volume processed; lack of treatment response; persistent anti-heparin/PF4 antibodies and/or positive SRA; amount of time remaining prior to surgery; PF4 ELISA or SRA results; increase in 4Ts score; continued positive HIT testing; treatment refractoriness; and ongoing elevation of anti-heparin/PF4 antibodies.

Additional Therapies.

Beyond TPE, additional therapies used in the YES TPE cohort are shown in Table 2. The most frequently used additional therapy (respondents were asked to select all that apply) was parenteral direct thrombin inhibitor (DTI; n=15; 79%), followed by fondaparinux (n=9; 47%) and direct oral anticoagulants (n=9; 47%).

DISCUSSION

Our international practice survey is the first of its kind to report treatment practice patterns in the use of TPE for HIT. Our findings demonstrate that TPE is used to treat HIT in at least 29 institutions internationally (37% of 78 surveyed sites). However, we discovered that the number of cases performed per institution is small, with most institutions performing 10 or fewer cases every 2 years. We also found that the most common TPE indication at treating institutions was prophylactic treatment for acute HIT patients undergoing cardiovascular surgery followed by acute HIT with thrombosis. Where TPE was used to treat acute HIT with thrombosis, patients were more likely to receive a combination of plasma and albumin replacement and have greater variation in the number of TPE procedures. Patients with HIT receiving TPE prior to cardiovascular surgery were most likely to be treated with plasma replacement fluid. For both cardiovascular surgery and acute HIT with thrombosis indications, the most common volume processed was 1.0 plasma volume. In institutions that did use TPE in the treatment of HIT as well as those that did not, the most common non-TPE therapy was parenteral DTI.

When we consider that the average number of patients treated over a 2-year period at most institutions was 2–5, our survey suggests that TPE is infrequently used in the treatment of HIT at any given site. This finding likely reflects the clinical observation that HIT is a rare thrombotic disorder with a low absolute risk of occurrence.¹² Therefore, since patients are less likely to develop the disorder, the role for TPE is diminished. However, once HIT is diagnosed, the primary course of action is to stop the offending anticoagulant and transition the patient to a parenteral DTI.² As DTIs are primarily used in the management of HIT, TPE only becomes relevant in selected circumstances such as when continued anticoagulation is contraindicated;⁵ there is ongoing clinical deterioration despite DTI therapy;⁹ or there is need for heparin re-exposure in the setting of cardiovascular surgery.^8,13

Our survey demonstrated that TPE is most commonly used as prophylactic treatment prior to heparin re-exposure in patients undergoing cardiovascular surgery. Because this finding is consistent with reports in the literature,^8,9 it is not surprising that for the cardiovascular surgery indication, TPE treatment parameters such as plasma volume processed, number of procedures performed, and timing of the procedure prior to surgery were more clearly defined than for acute HIT with thrombosis. Nevertheless, there is still some variation in the number of TPE procedures needed for effective antibody reduction.^5,8,13 Our survey findings appear to demonstrate equipoise between the number of cardiovascular surgery patients that are treated with a set number of TPE procedures vs. those treated to laboratory or clinical response.

Greater variation in the number of TPE procedures is seen for patients who are treated for the indication of acute HIT with acute thrombosis. For this indication, more respondents indicated they would treat based on laboratory or clinical response than with a set number of procedures. This variation likely reflects limited data available in the literature to guide decision making.

Nevertheless, it was clear that most respondents used laboratory or clinical parameters to help define treatment endpoints. The most frequent parameter used, which had a higher frequency of use than platelet counts, was a positive anti-heparin/PF4 antibody result. In the treatment of HIT, anti-heparin/PF4 antibodies appear to be the most common laboratory parameter used. However, caution should be exercised in use of antibody results to guide management because some early data suggest that there may be a discrepancy between anti-heparin/PF4 antibody and functional assay results.¹⁴ More data are needed to identify the best method to assess TPE responsiveness and possibly allow for fewer TPE procedures.

With 37% of surveyed sites demonstrating TPE use in HIT, we observed that TPE for the treatment of HIT is not a rare therapeutic intervention. Thus, it is somewhat surprising that fewer than half of these institutions that perform TPE for HIT have an established standard operating procedure (SOP). The paucity of centers with defined SOPs indicate that there is an opportunity to better define the role of TPE in the treatment of HIT. It may also suggest the presence of clinical equipoise to support prospective studies in this field.

The primary limitation of our survey is the absence of additional demographic data describing the kinds of institutions at which TPE is used for treating HIT. The literature would suggest, and practicality dictate, that the institutions most likely to use TPE for HIT are large inpatient facilities with high output cardiovascular procedures which tend to have a higher number of cases.³ Our retrospective evaluation suggested that, if we had excluded centers where HIT is not routinely diagnosed, the overall percentage of TPE use may be increased. Furthermore, there are likely to be systematic differences between responding and non-responding institutions.

Additionally, a practice patterns survey may not accurately reflect institutional practice. A better measure of practice may be actual data gathered from treatment centers. Recall bias may also exist if the respondents did not base their responses on chart reviews or otherwise had incomplete information on institution-wide practices. Therefore, a retrospective or prospective analysis of HIT patients treated with TPE may be more informative. To counter this problem, the HIT Subcommittee is working towards initiating a multi-center retrospective (and subsequently prospective) study of TPE use in HIT. Such data will give us a better understanding of TPE in HIT practice patterns and give the apheresis community additional insights into interventional studies that may be relevant to perform prospectively.

Such prospective studies could address questions such as, “What is the minimum number of TPE procedures needed for maximum clinical benefit?” The need for this question is supported by data that show that ultra large size of pathogenic heparin-PF4 antibody complexes (> 670 kDa)¹⁵ may require fewer than the typical 5–6 procedures that are needed for smaller IgG antibodies (~150 kDa).¹⁶ Another question for investigation could be, “What is the appropriate replacement fluid for treatment?” In support of this question, in vitro data suggest that normal plasma may inhibit HIT antibody-mediated platelet activation.¹⁷ However, it is not known whether this effect translates clinically or how much plasma may be needed to inhibit antibodies. Therefore, prospective studies to determine the appropriate replacement fluid are also needed to better understand the use of TPE therapy in HIT.

This survey was the first step toward determining the feasibility of prospective studies to evaluate the use of TPE for management of HIT. Given the number of active institutions and small numbers of patients, prospective studies will need to be performed as multi-institutional trials. This feat would require the support of as many participating ASFA member sites as possible. We encourage those within the ASFA community and beyond to participate in both retrospective and prospective studies of TPE in HIT. The practice survey has served an important first step to help our HIT Subcommittee understand the therapeutic landscape in HIT therapy. Our Subcommittee’s next steps will include undertaking a multi-institutional retrospective study to understand and better define our actual clinical practice, to more clearly define our clinical and laboratory endpoints, and to identify which patient populations may be best targeted in prospective studies.

CONCLUSION

TPE is used to treat HIT in a minority of survey site respondents. In this group, the most common indication is prophylactic treatment in patients who will undergo heparin re-exposure during cardiovascular surgery. These findings highlight the need for prospective studies to more clearly define the role of TPE in HIT.