A Pilot Study on Probing of Imatinib Induced Platelet Dysfunction in Patients with Chronic Myeloid Leukemia-Chronic Phase and Absence of Associated Bleeding Manifestation: Trying to Solve an Enigma

Abstract

Imatinib, the first Tyrosine Kinase Inhibitor (TKI) used for the treatment of chronic myeloid leukaemia (CML) has revolutionized the management by inhibiting BCR-ABL tyrosine kinase. According to earlier reports there are concerns regarding the adverse effect of imatinib on haemostasis by causing platelet dysfunction. Here we studied platelet function using platelet aggregometry, in 19 CML chronic phase (CML-CP) patients on imatinib therapy, in complete haematologic response (CHR). The median duration of imatinib therapy before performing the test was 154 days. This study reveals that there are large inter-individual variations in platelet functions among imatinib treated patients and different levels of variability have been seen for different agonists. Most common aggregation abnormality (< 50% aggregation) was seen with low dose collagen (1 μg/ml) in 31.57% patients. Despite in-vitro platelet aggregation defects, none of the patients showed any bleeding symptoms. This enigma can possibly be explained by the fact that platelet specific agonists, epinephrine and collagen act in synergy for platelet aggregation compared against individual low dose agonists, supported by ex-vivo experiments in normal healthy control group (n = 5) (p value < 0.0004 for epinephrine, p value < 0.0001 for collagen). This experiment was also confirmed in a CML-CP patient. In future, more studies are needed to find out the exact mechanism of this inhibition.

Electronic supplementary material

The online version of this article (10.1007/s12288-020-01376-8) contains supplementary material, which is available to authorized users.

Introduction

Tyrosine kinase inhibitors (TKIs) have significantly improved the prognosis of Chronic Myeloid Leukemia (CML) by inhibiting BCR-ABL kinase. Imatinib is one of the first line and most widely used TKI therapy for CML [1]. It is a small signal transduction inhibitory molecule that targets BCR-ABL [2]. The efficacy of imatinib is well established [3], but there are few reports on its controversial effect on haemostasis by inhibiting platelet aggregation [4, 5] with occasional bleeding tendency. Recent findings have shown the inhibition in platelet function during TKI therapy is not usually clinically reflected through bleeding manifestations [6], the reason of this enigma is yet to be explored. However, detecting the extent of platelet dysfunction, among CML patients on imatinib, is important to assess the clinical risk of haemostatic complications, if any, for better patient management. More on that, to the best of our knowledge, the way imatinib affects platelet function in CML-CP patients has not been prospectively studied in detail for Indian population, except for very few case reports [7].

Thus, the main focus of this study is to assess the extent of platelet dysfunction in CML-CP patients on imatinib from eastern part of India and analyze the possible reason behind this paradoxical situation of impaired platelet aggregation without significant bleeding manifestation.

Materials and Methods

Patients and Control

CML-CP patients, diagnosed as per ELN guideline [8] by morphology, Cytogenetics, FISH and/or RT-PCR for BCR/ABL, above 18 years of age except pregnant individuals, under imatinib therapy were included in the study. Patients were followed up according to recommendations of standard guideline [8]. There were altogether 19 patients with CML-CP on imatinib therapy, analyzed for platelet functions. The patient population had male predominance (73.6%) with age ranging from 24 to 78 years. All patients were in complete haematologic response (CHR) following imatinib therapy at the time of assay (median duration of imatinib therapy 154 days). Blood taken from 10 normal healthy individuals were studied as appropriate control for platelet function test. Synergism experiment with collagen and epinephrine was performed in 5 healthy individuals and 1 CML-CP patient.

Platelet Aggregometry

The platelet optical aggregometry study was conducted as described in previously published reports [9–11]. Concisely, blood drawn from subjects was collected in 3.2% tri-sodium citrate anticoagulant (9:1) for platelet function study by ChronoLog aggregometer (Model 530BS, Havertown, USA) and in EDTA for platelet count using automated cell counter Sysmex SF3000 (Ramsey, USA). Citrated blood was centrifuged to prepare platelet rich plasma-PRP (200 g for 10 min) for platelet function test and platelet poor plasma-PPP (1500 g for 10 min) as standard negative control. Platelet aggregation study was assessed by adding different platelet specific agonists separately, i.e. ADP, epinephrine and collagen (CHRONO-PAR, US) to PRP at 37 °C temperature with a stirring rate of 1000 rpm. Synergism study was assessed by simultaneously adding collagen and epinephrine to PRP [10]. Synergism study was same as normal aggregometric study. Only exception was instead of adding one agonist, two agonists were added simultaneously to the PRP keeping all other parameters same. Here simultaneous addition means the addition of agonists was at not more than 5 s intervals. The details of agonists concentrations used in this study are as follows: ADP higher dose (10 µM), ADP lower dose (2 µM), collagen higher dose (4 µg/ml), collagen lower dose (1 µg/ml), epinephrine standard dose (10 µM). For the synergism study collagen (0.5 µg/ml) and epinephrine (0.2 µM) were used [10].

Statistical Analysis

Statistical analysis was performed using GraphPad Prism 5 software (CA, USA). Results were presented as mean ± standard error of mean (SEM). Comparison between two sets of results were analyzed by t-test (unpaired). The P value was calculated for each set where n = 10 for normal healthy individual and n = 19 for CML-CP patients and value < 0.05 was taken as statistically significant. P value = 0.0443 and 0.0051 are represented respectively as *, ** in the corresponding graphs.

Results

Patient Demography

Among all CML-CP patients on imatinib therapy, 18 patients received at a dose of 400 mg/day and 1 patient at a dose of 300 mg/day. As per the past medical history 2 patients on imatinib therapy had thrombocytopenia (platelet count of 54,000/µl and 46,000/µl) for which imatinib was stopped for a stipulated time, and again restarted. Both the patients were on imatinib for 3 months prior to the test and platelet counts were above 1,20,000/µl during this study. All the 19 patients had platelet count well above the required count for platelet function test (minimum platelet count during the test was 90,000/µl). All patients were in complete haematologic response (CHR) following imatinib therapy at the time of assay. The median duration of imatinib therapy before performing the test was 154 days. Two patients had past history of gum bleeding associated with thrombocytopenia and one patient had history of bleeding PR from external haemorrhoids. None of the patients had any complaint of bleeding when CHR was achieved on imatinib. All the patients received uninterrupted imatinib therapy for a minimal duration of one month.

Variability in the Agonists Induced Platelet Aggregation

Three platelet agonists (ADP, collagen and epinephrine) were used with different doses (ADP 2 µM and 10 µM; collagen 1 µg/ml and 4 µg/ml and epinephrine 10 µM) to evaluate platelet function among 19 CML-CP patients (Fig. 1, Supplement Fig. 3). There were some levels of inter-patient variability of platelet aggregation to all agonists, which is in agreement with a recently published research [12]. It was found, that the variation was maximum for low dose collagen 1 µg/ml and epinephrine 10 µM (Fig. 1). Such cases showed the significant inhibition of platelet aggregation compared to the control group e.g., low dose collagen (p ≤ 0.01) and epinephrine (p ≤ 0.05) (Fig. 1). ADP (for both the concentrations 2 µM and 10 µM) activation did not show any significant change in platelet aggregation pattern (Fig. 1). High dose of collagen showed lowering of platelet aggregation for imatinib treated patients though it was not statistically significant compared to the control group ((Fig. 1). As mentioned before, the significant platelet inhibition was seen both for collagen (1 µg/ml) and epinephrine (10 µM) but the inhibition potency was more for collagen (1 µM) compared to epinephrine (10 µM). According to the Supplement Fig. 1, 52.63% of the total population has shown ‘normal aggregation’ (when final aggregation is ≥ 70%) in response to epinephrine (10 µM), whereas for collagen 1 µg/ml, it was only 26.31%. Likewise, 21.05% of total population was within the zone of ‘diminished aggregation’ (≥ 50% to < 70% of final aggregation) for epinephrine (10 µM) and for collagen 1 µg/ml, it was 42.10%. For the ‘impaired aggregation’ zone (when final aggregation is < 50%), the epinephrine (10 µM) and collagen (1 µg/ml) induced aggregation were 26.31% and 31.57% respectively.

Fig. 1.

Altered platelet aggregation in CML patients on imatinib therapy. Agonists used a ADP 10 μM, b ADP 2 μM, c Epinephrine 10 μM, d Collagen 1 μg/ml, e Collagen 4 μg/ml. Patient = 19, control group = 10. (*) is P value 0.0443 and ** = P value 0.0051

Collagen Epinephrine Synergy

Experiment for collagen-epinephrine synergy was done in ex-vivo condition where, very low dose collagen (0.5 µg/ml) and epinephrine (0.2 µM) could not induce any platelet aggregation individually but combination of the aforesaid agonists with same doses (collagen 0.5 µg/ml + epinephrine 0.2 µM), used simultaniously, showed a sharp aggregation with a steep slope. This is comparable with the higher doses of same individual agonists (collagen 4 µg/ml and epinephrine 10 µM). This experiment was done in PRP of normal healthy volunteers (n = 5, Fig. 2) as well as a CML-CP patient (Supplement Fig. 2). The difference between the individual agonist, (collagen (0.5 µg/ml), epinephrine (0.2 µM)) and synergistic action of both agonists were satistically significant, (p value < 0.0001, mean of difference = 68, 95% confidence interval = 62.45 to 73.55) for collagen, and (p value < 0.0004, mean of difference = 62, 95% confidence interval = 45.18 to 78.19) for epinephrine (Fig. 2).

Fig. 2.

Platelet aggregation in presence of collagen and epinephrine in normal healthy individuals (n = 5). shows very low concentration of collagen (0.5 μg/ml) and epinephrine (0.2 μM) individually didn't induce any platelet aggregation. Whereas, simultaneous activation with collagen 0.5 μg/ml and epinephrine 0.2 μM induced sharp platelet aggregation ( synergistic effect). **** = P value < 0.0001 and *** = P value = 0.004

Discussion

Imatinib is a first line therapy for CML-CP treatment. There are few reports on its controversial effect on haemostasis. As per our knowledge this is the first dedicated prospective study from India addressing the effect of imatinib on platelet function in CML-CP patients. The study reveals a large inter-patient variability in platelet aggregation pattern among imatinib treated patients, with varying doses of different platelet specific agonists (Fig. 1). This is in line with a recent finding [12] where TKI response in in-vitro condition showed the same variation.

In this study, platelet function test of CML-CP patients revealed a significant inhibition in platelet aggregation in presence of both epinephrine 10 µM and collagen 1 µg/ml (Fig. 1, Supplementary Figs. 1, 3). The mechanisms behind the extent of inhibitions are distinct for different agonists. Epinephrine being a secondary agonist (low potency) is susceptible to any mild inhibition in platelet function, whereas collagen being a strong agonist can override any inhibitory effect of imatinib in conventionally used dose (4 µg/ml). Rather, collagen at lower concentration (1 µg/ml) is pretty sensitive to the mild inhibition of platelets in imatinib treated patients (Supplementary Fig. 1).

Considering final aggregation of ≥ 70% as ‘normal aggregation’, ≥ 50% to < 70% as ‘diminished aggregation’ and < 50% as ‘impaired aggregation’, as mentioned in other related article [6], collagen 1 µg/ml induced aggregation has been shown to be inhibited much more than epinephrine 10 µM induced aggregation (Supplement Fig. 1) in the imatinib treated CML-CP patients. This has been also reflected in the t test, where significant differences compared to the control, was seen more for collagen 1 µg/ml (p ≤ 0.01) than epinephrine 10 µM (p ≤ 0.05) (Fig. 1). The similar effect was observed in a recently published study, where in in-vitro TKI treatment, collagen related peptide (CRP) induced platelet activation was mostly inhibited [12]. Now why platelets showed more inhibition with collagen (lower concentration) than standard epinephrine dose, may be due to the inhibitory effect of imatinib on Src family kinase (SFK) [13, 14] which is one of the most important components in glycoprotein-VI pathway (collagen mediated) [15]. However, a future study in this aspect will help in the validation of this phenomenon, which is beyond the scope of this study.

The interesting fact is that, though some imatinib treated patients showed significant lowering in platelet aggregation both for collagen—low dose and epinephrine but none of these patients reported any associated bleeding manifestation. This observation is also in agreement with a recent research [6]. The reason behind this clinically asymptomatic platelet inhibition is probably due to the GP-VI (collagen) and alpha adrenergic (epinephrine) pathway synergy in platelet aggregation. Individual epinephrine and low dose collagen induced platelet aggregation is susceptible to the inhibitory effect of imatinib, but the simultaneous addition of said agonists even in ultra low concentrations can override this inhibition due to synergism effect (Fig. 2, Supplement Fig. 2). As these agonists (collagen and epinephrine) are present in physiological system, their synergistic action [10, 16, 17] most probably prevents clinically significant bleeding manifestation even in presence of mild inhibition induced by imatinib.

Conclusion

In conclusion platelet function inhibition among imatinib treated CML-CP patients due to inhibition of GP-VI pathway can be best diagnosed by platelet aggregation study with low dose collagen (1 µg/ml). Due to the synergistic action of alpha adrenergic receptor and GP-VI pathway this inhibition is not reflected in clinically significant bleeding manifestations. Therefore, imatinib can be considered as a safe drug for CML-CP patients in respect to platelet aggregation as far as bleeding manifestations are concerned. Future studies are needed to find out the underlying mechanism of platelet inhibition and inter-individual variations in CML-CP patients on imatinib therapy.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

CML

Chronic Myeloid Leukemia

CML-CP

Chronic Myeloid Leukemia Chronic Phase

TKI

Tyrosine Kinase Inhibitor

FISH

Fluorescence in situ hybridization

RT PCR

Real Time PCR

EDTA

Ethylenediaminetetraacetic acid

PRP

Platelet Rich Plasma

PPP

Platelet Poor Plasma

ADP

Adenosine diphosphate

SEM

Standard Error of Mean

PFT

Platelet Function Test

ELN

European Leukemia Net

Author contributions

RD: concept, clinical data analysis and writing; RC: data analysis and writing; TKD: clinical data analysis; BB: experiment; MMI: statistical analysis; PC: analysis; SD: Concept, experiments, data analysis and writing.

Compliance with Ethical Standards

Conflict of interest

The authors declare that they have no conflict of interest.

Human and Animal Rights

All performed experiments in this study involving human participants were in accordance with the ethical standards of the institutional ethical committee No/NMC/4655 dated 27/08/2016) of NRS medical college, Kolkata, India and with due consent of the participating individuals (Funded by DST young scientist grant (YSS/2015/002101/dated 26/09/16), DST, India.