http://informahealthcare.com/plt ISSN: 0953-7104 (print), 1369-1635 (electronic) Platelets, Early Online: 1–4 ! 2014 Informa UK Ltd. DOI: 10.3109/09537104.2014.977243

ORIGINAL ARTICLE

Outcome of patients with positive heparin–platelet factor-4 antibodies: A retrospective multi-institutional observational study Umit Tapan1, Saritha Bolla2, Ebubekir S. Daglilar3, Shaoyu Chang4, & Olga Kozyreva2,3

Platelets Downloaded from informahealthcare.com by University of Ulster at Jordanstown on 03/06/15 For personal use only.

1

Division of Hematology/Oncology, Boston University Medical Center, Boston, MA, USA, 2Division of Hematology/Oncology, St. Elizabeth Medical Center, Boston, MA, USA, 3Department of Medicine, St. Elizabeth Medical Center, Boston, MA, USA, and 4Department of Clinical and Translational Sciences, Tufts University, Boston, MA, USA Abstract

Keywords

Studies show increased mortality with positive heparin–platelet factor-4 (H–PF4) antibodies, especially in hemodialysis patients. We aimed to compare mortality and thrombosis in hospitalized patients with positive, equivocal and negative H–PF4 antibody results. Information was collected on these patients using a multi-institutional retrospective electronic medical record review. Patients tested for H–PF4 antibodies by commercial ELISA during the years 2006 to 2010 were identified. We compared 30-day, 90-day and 1-year mortality in patients with negative, equivocal and positive H–PF4 test and evaluated the relationship between H–PF4 status and rate of thrombosis. Four hundred and seventeen patients had ELISA testing for H–PF4 antibodies. Forty-four patients had equivocal (optical density value 0.4–0.9) and 21 had positive (value 1) H–PF4 antibody test. There were no statistically significant differences in mortality between patients with negative, equivocal and positive results at all three time points (p ¼ 0.22, 0.27 and 0.38, respectively) even after excluding patients with thrombosis (p ¼ 0.22, 0.24 and 0.31, respectively). Age and Charlson score were associated with increased 30-day, 90-day and 1 year mortality. Odds ratio of having thrombosis was 23.1 for positive vs. equivocal results (p50.001); however, there was no statistically significant difference between equivocal vs. negative results (p ¼ 0.22). Our results revealed no association between H–PF4 status and mortality, as well as no difference in 1-year survival between the positive and negative groups.

Heparin, heparin-induced thrombocytopenia and thrombosis, heparin–platelet factor-4 antibody, mortality, thrombosis

Introduction Heparin-induced thrombocytopenia (HIT), mediated by heparin– platelet factor-4 antibody (H–PF4), is an important entity that can lead to arterial and venous thrombosis and can potentially be fatal [1]. HIT occurs between 5 and 14 days after heparin exposure in 1–5% of patients but only 30–50% of these patients develop thrombosis [2]. Presence of H–PF4 antibody along with appropriate clinical context is highly suggestive of HIT. Negative H–PF4 result can rule out HIT with 4 97% sensitivity [3]. However, its specificity is low because up to 50% of surgical and 20% of medical patients treated with heparin may develop H–PF4 antibodies by enzyme-linked immunosorbent assay (ELISA), and only 1–5% develop clinical HIT [4, 5]. ELISA by the GTI Diagnostics, Inc. (Brookfield, WI) is commonly used to detect H–PF4 antibodies. Results are reported as optical density (OD) values by most of the laboratories and OD value 0.4 is accepted as positive. Significance of positive H–PF4 antibody testing without clinical HIT is not known. Some patients with these antibodies might develop clinical HIT later on (especially if they have higher OD values) whereas some will remain free of thrombosis or other sequelae. There are also studies showing increased mortality with positive H–PF4 antibodies, especially in

Correspondence: Dr. Umit Tapan, MD, 820 Harrison Avenue, FGH Building 1st floor, Division of Hematology/Oncology, Boston University Medical Center, Boston, MA 02118, USA. Tel: +1-617-638-6428. Fax: +1-617-638-7530. E-mail: [email protected]

History Received 17 July 2014 Revised 9 October 2014 Accepted 11 October 2014 Published online 10 November 2014

hemodialysis patients; however, this association was not evident in some others [6, 7]. In current study, we aimed to compare mortality and thrombosis in patients with positive, equivocal and negative H–PF4 antibody results.

Materials and methods Data collection method Study protocol was reviewed and approved by the Institutional Review Board (IRB) prior to data collection. Adult patients who had polyspecific H–PF4/ELISA antibody testing at Steward Carney Hospital and St. Elizabeth Medical Center between 1 January 2006 and 31 December 2010 were included in the study. Patient characteristics are outlined in the Table I. Patients who carried a previous diagnosis of HIT were excluded. Information was collected on these patients using retrospective chart review. Results of each study subject’s H–PF4 antibody test were classified into one of three categories based on the OD range of the result. OD values of 1 were classified as positive, values of 0.4–0.9 were classified as equivocal and values 50.4 were classified as negative. Patient charts were analyzed by two reviewers, one from each institution, and all charts were adjudicated by a single reviewer. Statistical analysis In order to assess primary aim of comparing mortality in patients with positive, equivocal and negative H–PF4 results, three 2 tests

2

U. Tapan et al.

Platelets, Early Online: 1–4

Table I. Age, gender, occurrence of thrombosis, initial platelet count, nadir platelet count, Charlson score and 4T score of patients with negative, equivocal and positive H–PF4 test results.

Platelets Downloaded from informahealthcare.com by University of Ulster at Jordanstown on 03/06/15 For personal use only.

Characteristic

Negative (OD50.4) (n ¼ 352)

Age (mean ± SD) 70.6 ± 13.88 Gender, n (%) Male 195 (55.4) Female 157 (44.6) Thrombosis, n (%) Yes 37 (10.5) No 315 (89.5) Initial plt (mean ± SD) 182.2 ± 85.9 Nadir plt (mean ± SD) 78.2 ± 36.1 Charlson score (mean ± SD) 3.7 ± 2.42 4T score (mean ± SD) 2.8 ± 1.37

Equivocal (OD ¼ 0.4–0.9) (n ¼ 44)

Positive (OD  1) (n ¼ 21)

p Value

68.5 ± 14.10

70.5 ± 13.50

0.645

29 (65.9) 15 (34.0)

7 (33.3) 14 (66.6)

0.047

2 (4.5) 42 (95.4)

11 (52.3) 10 (47.6)

0.0001

research articles related with ELISA H–PF4 testing. Reviews and research articles that were published before 2000 were not included.

Results

206.0 ± 114.3

208.5 ± 98.6

0.158

77.0 ± 37.3

74.6 ± 62.7

0.916

4.2 ± 2.66

3.3 ± 2.68

0.354

3.5 ± 1.65

5.0 ± 1.89

0.0001

SD, standard deviation; plt, platelet.

were performed for each of the three time points: during hospital stay when testing was ordered (30-day was used instead of inhospital mortality), 90 days after testing and 1 year after testing. All study subjects were classified as alive (yes/no) at each of these three time points, and the proportions that were alive at each time point were compared between three groups. Vital status of the patients was determined by chart review and the social security death index. Survival over 1 year was also compared between subjects with positive, equivocal and negative H–PF4 test results using the Kaplan–Meier method and log-rank test. Difference in survival between the three groups was modeled using Cox proportional hazards regression. This model was adjusted for age, sex and Charlson comorbidity score. A descriptive analysis was used to assess the clinical appropriateness of H–PF4 antibody testing based on 4T score. A 4T score was determined for all subjects, followed by classification into groups based on the 4T score as follows:  3 (low probability for HIT), 4–5 (intermediate probability for HIT) and  6 (high probability for HIT). Proportions of patients in each of these three categories who had H–PF4 antibody testing were calculated and reported with 95% confidence intervals (CI). Patients who were free from thrombosis at the time of H–PF4 antibody testing and who developed thrombosis within 30 days of H–PF4 antibody testing were further analyzed. Radiologic reports within 30 days of ELISA testing were used to adjudicate thrombosis. A 2 test was used to compare the proportions of patients with positive and equivocal H–PF4 results who experienced thrombosis within 30 days of testing. In addition, logistic regression was used to calculate an odds ratio (with 95% CI) comparing the odds of diagnosed thrombosis within 30 days between equivocal and the negative result groups and between the positive and equivocal result groups. This model was adjusted for age, sex, nadir platelet and Charlson comorbidity score which is used to predict mortality for a patient who may have comorbidities such as myocardial infarction, AIDS or malignancy via assigning a score of 1–6 depending on the significance of the comorbid condition [8]. The association between 4T score and H-PF4 result was examined by Fisher’s exact test. Statistical Analysis System (SAS) version 9.3 (SAS Institute Inc., Cary, NC) was used for statistical analysis. p50.05 considered significant (two tailed). Literature review National Center for Biotechnology Information, US National Library of Medicine website was analyzed to obtain reviews and

Four hundred and seventeen patients had ELISA testing for H–PF4 antibodies. Patient characteristics are summarized in Table I. Among total of 417 patients, 44 had equivocal (OD value 0.4–0.9) and 21 had positive (OD value 1) H-PF4 antibody test result. Thirty-day, 90-day and 1 year mortalities were 13.6% (95% CI: 10.02–17.18), 33.2 (95% CI: 28.28–38.12) and 46.3 (95% CI: 41.09–51.51), respectively for HPF4 negative patients, 4.5 (95% CI: 1.63–10.63), 45.4 (95% CI: 30.69–60.11) and 56.8 (95% CI: 42.16–71.44), respectively for patients with equivocal HPF4 status and 14.3 (95% CI: 0.67–29.27), 33.3 (95% CI: 13.14–53.46) and 42.8 (95% CI: 21.64–63.96), respectively for H-PF4 positive patients. There were no statistically significant differences in mortality between patients with negative, equivocal and positive results at all three time points (p ¼ 0.22, 0.27 and 0.38, respectively). This was the case also after excluding the patients with thrombosis (p ¼ 0.22, 0.24 and 0.31, respectively). As shown in Figure 1, Kaplan–Meier plot of 1-year survival curves failed to demonstrate significant difference between the three H-PF4 test result groups (log-rank test p ¼ 0.51). Charlson score was associated with increased 30-day, 90-day and 1-year mortality, with odds ratios of 1.17, 1.24 and 1.28, respectively (all p50.05). Age was associated with increased 3 months and 1 year mortality with odds ratios of 1.034 and 1.040, respectively (p50.05). Multiple regression analysis showed that age, nadir platelet and Charlson score had no statistically significant effect on the occurrence of thrombosis in negative, equivocal or positive groups. There was a significant correlation between the occurrence of thrombosis and H–PF4 result (Figure 2). Using logistic regression analysis, odds ratio of having thrombosis was 23.1 for positive vs. equivocal results (p50.001); however, there was no statistically significant difference between equivocal vs. negative results (p ¼ 0.22) (Table II). The correlation between 4T score and H-PF4 result was as shown in Table III. The assignment of low probability according to 4T score was strongly associated with a negative H-PF4 result, while high probability according to 4T score was associated with a positive H-PF4 result (Fisher’s exact p50.001).

Discussion Significant progress has been made in diagnosis and management of HIT since the first published report of HIT complicated by thrombosis in 1969 [9]. Diagnosis of HIT relies on clinical assessment and laboratory testing (e.g. detection of H–PF4 antibody by ELISA) [10]. The 4Ts is a pretest clinical scoring system that incorporates timing and degree of thrombocytopenia, presence of thrombosis or other sequelae and likelihood of other possible causes of thrombocytopenia [11]. Depending on total score 4Ts system provides low, intermediate and high pretest probability for HIT [12]. A recent review and meta-analysis by Cuker et al. [10] found that NPV of a low probability 4T score was 0.998; positive predictive value (PPV) of an intermediate and high probability 4T score was 0.14 and 0.62, respectively. Authors suggested that low probability 4T score appeared to be helpful to exclude HIT. In current study, high probability 4T score was associated with positive H–PF4 test result and only 2.7% of patients with low probability 4T score had positive H–PF4 result. Detection of H–PF4 antibodies by ELISA has been widely performed in USA and reported either as continuous OD values or negative, weakly positive and positive according to OD cutoff

Heparin–platelet factor-4 antibody and survival

DOI: 10.3109/09537104.2014.977243

3

Platelets Downloaded from informahealthcare.com by University of Ulster at Jordanstown on 03/06/15 For personal use only.

Figure 1. Kaplan–Meier plot of 1-year survival curves failed to demonstrate significant difference between the three H-PF4 test result groups (log-rank test p ¼ 0.51). Result: 0 ¼ negative, 1 ¼ equivocal and 2 ¼ positive.

Figure 2. Multiple regression analysis showed that age, nadir platelet and Charlson score had no statistically significant effect on the occurrence of thrombosis in negative, equivocal or positive groups. There was a significant correlation between the occurrence of thrombosis and H–PF4 result (composite OR ¼ 2.476).

Table II. Effects of H–PF4 result, age, gender and Charlson score on the thrombosis risk.

Table III. Correlation between 4T score and H–PF4 result. H–PF4 test results

Variables H–PF4 (+) vs. equivocal Equivocal vs. H–PF4 () Gender Charlson score

Odds ratio (95% CI) 23.1 0.4 0.8 1.1

(4.4–121.1) (0.09–1.7) (0.45–1.58) (0.95–1.21)

p Value 50.0001 0.22 0.61 0.22

4T score Low probability Intermediate probability High probability

Negative

Equivocal

Positive

Total

255 89 8

24 16 4

8 5 8

287 110 20

Fisher’s exact p value ¼ 5.968  e0950.0001.

points (50.4, 0.4–0.9 and 1, respectively). Recent studies have raised two major concerns on reporting of H–PF4 antibody testing: first is the concept of ‘‘weak-positive’’ result and second is reporting as positive, weakly positive or negative and not the actual OD value. Baroletti et al. [13] showed that higher OD

predicted for greater probability of clinically evident thrombosis, pointing the importance of reporting the actual OD value. In another study by Zwicker et al. [2], thrombosis risk was nearly 6-fold higher in patients with OD41.0 vs. OD between 0.4 and 0.99. In the editorial accompanying Baroletti et al. [13],

Platelets Downloaded from informahealthcare.com by University of Ulster at Jordanstown on 03/06/15 For personal use only.

4

U. Tapan et al.

Warkentin [14] suggested that OD 51.0 should be considered against HIT. In our study, there was a significant correlation between H–PF4 result and thrombosis risk; odds ratio of having thrombosis was 23.1 for positive vs. equivocal results (p50.001); however, there was no statistically significant difference between equivocal vs. negative results (p ¼ 0.22). The available literature provides estimates of overall mortality rates for thrombocytopenic patients regardless of H–PF4 status. Oliveira et al. [15] observed 5.1% in-hospital mortality in a group of patients similar to our patient population with thrombocytopenia. Another study by Lopes at al. [16] looking at the same patient population found that 6-month mortality was similar among patients with normal thrombocyte counts and patients with in-hospital thrombocytopenia. In that study, 10% of these patients had died within 6 months. Zwicker et al. [2] retrospectively analyzed 63 patients with H–PF4 antibody; among these patients the 30-day in-hospital mortality rate was 39% for those with thrombosis (i.e. HITT) and 10% for those with isolated HIT. About 37.5% of patients with isolated HIT had postmortem examinations which did not show any thrombosis. Pena de la Vega et al. [17] reported an increase in mortality in hemodialysis patients with elevated H–PF4 antibody (adjusted HR ¼ 2.47) compared to patients with lower antibody levels. In the same study, hazard ratio for cardiovascular mortality was even higher (adjusted HR of 4.14). Carrier et al. [18] also showed that having this antibody was an independent predictor of mortality (adjusted HR 2.68). In our study, there were no statistically significant differences in mortality between patients with negative, equivocal and positive results at all three time points. In order to identify the effect of H–PF4 antibody on mortality, a second mortality analysis was performed after excluding patients with thrombosis and this analysis also did not detect any statistically significant difference. Our study carries the inherent limitations of a retrospective analysis including incomplete documentation or follow up which could have affected the results of the study. Another weakness is that H–PF4 testing is not reported as exact OD value in our institution. Chart reviewers were not blinded to the ELISA results; therefore, there could have been a bias while evaluating thrombosis and 4T score. We believe that our study is otherwise well designed and has included enough number of patients to detect the mortality difference and risk of thrombosis according to H–PF4 status. Further prospective studies could elucidate the effect of H–PF4 antibody (based on exact OD value) in patients who do not have thrombosis and help clinicians to manage this entity appropriately.

Conclusion H–PF4 antibody testing is commonly performed in patients with suspected HIT and has a high negative predictive value. Significance of a positive H–PF4 result in absence of thrombosis is less clear but available literature suggests that higher OD values are associated with higher thrombosis risk. In this study, there was no statistically significant mortality difference in patients with negative, equivocal or positive H–PF4 result in contrast to some studies suggesting an association. We found a significant correlation between the occurrence of thrombosis and positive H–PF4 result.

Declaration of interest The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.

Platelets, Early Online: 1–4

References 1. Amiral J, Bridey F, Dreyfus M, Vissoc AM, Fressinaud E, Wolf M, Meyer D. Platelet factor 4 complexed to heparin is the target for antibodies generated in heparin-induced thrombocytopenia. Thromb Haemost 1992;68:95–96. 2. Zwicker JI, Uhl L, Huang WY, Shaz BH, Bauer KA. Thrombosis and ELISA optical density values in hospitalized patients with heparin-induced thrombocytopenia. J Thromb Haemost 2004;2: 2133–2137. 3. Greinacher A, Juhl D, Strobel U, Wessel A, Lubenow N, Selleng K, Eichler P, Warkentin TE. Heparin-induced thrombocytopenia: A prospective study on the incidence, platelet-activating capacity and clinical significance of antiplatelet factor 4/heparin antibodies of the IgG, IgM, and IgA classes. J Thromb Haemost 2007;5:1666–1673. 4. Warkentin TE, Greinacher A, editors. Heparin-induced thrombocytopenia. New York: Marcel Dekker; 2000. 5. Trossaert M, Gaillard A, Commin PL, Amiral J, Vissac AM, Fressinaud E. High incidence of anti-heparin/platelet factor 4 antibodies after cardiopulmonary bypass surgery. Br J Haematol 1998;101:653–655. 6. Yusuf AM, Warkentin TE, Arsenault KA, Whitlock R, Eikelboom JW. Prognostic importance of preoperative anti-PF4/heparin antibodies in patients undergoing cardiac surgery. A systematic review. Thromb Haemost 2012;107:8–14. 7. Gluckman TJ, Segal JB, Schulman SP, Shapiro EP, Kickler TS, Prechel MM, Conte JV, Walenga JM, Shafique I, Rade JJ. Effect of anti-platelet factor-4/heparin antibody induction on early saphenous vein graft occlusion after coronary artery bypass surgery. J Thromb Haemost 2009;7:1457–1464. 8. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chronic Dis 1987;40:373–383. 9. Natelson EA, Lynch EC, Alfrey Jr CP, Gross JB. Heparin-induced thrombocytopenia. An unexpected response to treatment of consumption coagulopathy. Ann Intern Med 1969;71:1121–1125. 10. Cuker A, Gimotty PA, Crowther MA, Warkentin TE. Predictive value of the 4Ts scoring system for heparin-induced thrombocytopenia: A systematic review and meta-analysis. Blood 2012;120: 4160–4167. 11. Warkentin TE, Heddle NM. Laboratory diagnosis of immune heparin-induced thrombocytopenia. Curr Hematol Rep 2003;2: 148–157. 12. Lo GK, Juhl D, Warkentin TE, Sigouin CS, Eichler P, Greinacher A. Evaluation of pretest clinical score (4 T’s) for the diagnosis of heparin-induced thrombocytopenia in two clinical settings. J Thromb Haemost 2006;4:759–765. 13. Baroletti S, Hurwitz S, Conti NA, Fanikos J, Piazza G, Goldhaber SZ. Thrombosis in suspected heparin-induced thrombocytopenia occurs more often with high antibody levels. Am J Med 2012;125: 44–49. 14. Warkentin TE. Optical densities reduce odds of heparin-induced thrombocytopenia over-diagnosis. Am J Med 2012;125:3–4. 15. Oliveira GB, Crespo EM, Becker RC, Honeycutt EF, Abrams CS, Anstrom KJ, Berger PB, Davidson-Ray LD, Eisenstein EL, Kleiman NS, et al. Incidence and prognostic significance of thrombocytopenia in patients treated with prolonged heparin therapy. Arch Intern Med 2008;168:94–102. 16. Lopes RD, Ohman EM, Granger CB, Honeycutt EF, Anstrom KJ, Berger PB, Crespo EM, Oliveira GB, Moll S, Moliterno DJ, et al. Six-month follow-up of patients with in-hospital thrombocytopenia during heparin-based anticoagulation (from the Complications After Thrombocytopenia Caused by Heparin [CATCH] registry). Am J Cardiol 2009;104:1285–1291. 17. Pena de la Vega L, Miller RS, Benda MM, Grill DE, Johnson MG, McCarthy JT, McBane II RD. Association of heparin-dependent antibodies and adverse outcomes in hemodialysis patients: A population-based study. Mayo Clin Proc 2005;80:995–1000. 18. Carrier M, Rodger MA, Fergusson D, Doucette S, Kovacs MJ, Moore J, Kelton JG, Knoll GA. Increased mortality in hemodialysis patients having specific antibodies to the platelet factor 4-heparin complex. Kidney Int 2008;73:213–219.

Outcome of patients with positive heparin-platelet factor-4 antibodies: A retrospective multi-institutional observational study.

Studies show increased mortality with positive heparin-platelet factor-4 (H-PF4) antibodies, especially in hemodialysis patients. We aimed to compare ...
218KB Sizes 0 Downloads 5 Views