Hypobaric Hypoxia Causes Elevated Thrombin ...

Cellular Haemostasis and Platelets

Hypobaric Hypoxia Causes Elevated Thrombin Generation Mediated by FVIII that is Balanced by Decreased Platelet Activation Cécile H. Kicken1,2 Marisa Ninivaggi2 Joke Konings2 Martijn Moorlag3 Dana Huskens2 Jasper A. Remijn2,4 Saartje Bloemen3 Marcus D. Lancé5 Bas De Laat2,3 1 Department of Anaesthesiology, Maastricht University Medical

Centre, Maastricht, The Netherlands 2 Synapse Research Institute, Maastricht, The Netherlands 3 Department of Biochemistry, Maastricht University, Cardiovascular Research Institute Maastricht, Maastricht, the Netherlands 4 Department of Clinical Chemistry and Haematology, Gelre Hospitals, Apeldoorn, The Netherlands 5 Department of Anaesthesiology, Intensive Care and Perioperative Medicine, Hamad Medical Centre, Doha, Qatar

Address for correspondence Cécile H. Kicken, MD, Department of Anaesthesiology, Maastricht University Medical Centre, P.O. Box 5800, 6202 AZ Maastricht, The Netherlands (e-mail: [email protected]).

Abstract

Keywords

► hypoxia ► thrombin generation ► platelet activation

received December 6, 2017 accepted after revision March 2, 2018

Introduction Epidemiological studies suggest that hypobaric hypoxia at high altitude poses a risk for developing venous thromboembolism. The cause of this observed hypercoagulability remains unclear. Therefore, this study aimed to investigate the effect of hypobaric hypoxia at 3,883 m above sea level on thrombin generation and platelet activation. Methods After complying with medical ethical procedures, 18 participants were recruited, of whom 1 had to leave the study prematurely due to mild acute mountain sickness. Blood was drawn first at 50 m above sea level and second at 3,883 m altitude after gradual acclimatization for 6 days. Thrombin generation was measured in whole blood, platelet-rich plasma and platelet-poor plasma. Platelet activation was assessed using a whole blood flow-cytometric assay. Coagulation factor levels, D-dimer levels and markers of dehydration and inflammation were measured. Results Hypobaric hypoxia at 3,883 m altitude caused increased thrombin generation, measured as peak height and endogenous thrombin potential, in whole blood, platelet-rich and platelet-poor plasma without or at low tissue factor concentration. The elevated thrombin generation was mediated by increased factor VIII levels and not caused by dehydration or inflammation. In contrast, spontaneous and agonist-induced platelet activation was decreased at high altitude. Conclusion Hypobaric hypoxia causes increased factor VIII-mediated thrombin generation. The hypercoagulability was balanced by decreased platelet activation. These findings may explain why venous, and not arterial thrombotic events occur more frequently at high altitude.

Copyright © Schattauer GmbH

DOI https://doi.org/ 10.1055/s-0038-1641566. ISSN 0340-6245.

Downloaded by: Universiteit Maastricht. Copyrighted material.

Thromb Haemost

Kicken et al.

Introduction Hypoxia is defined as oxygen deprivation at tissue level, which occurs when oxygen levels needed for cellular respiration are insufficient, leading to increased anaerobic metabolism. High altitude causes hypoxia due to decreased inspired oxygen pressure (pO2).1 Ambient air contains 20.9% O2, resulting in an inspired pO2 of 19.6 kPa at sea level. Whereas the oxygen percentage does not change at high altitude, the atmospheric pressure decreases almost linearly, resulting in a lower pO2. For instance, at 4,000 m above sea level the inspired pO2 decreases to 12.9 kPa.2 Epidemiological studies suggest that high altitude poses a risk for developing venous thromboembolism (VTE). Lowlanders stationed at high altitude were found to have a 30 to 44 times higher risk of presenting in hospital with a deep vein thrombosis (DVT), compared with patients at low altitude.3,4 Patients undergoing a knee arthroscopy at 1,200 m or higher were found to have a 3.8 times higher risk of developing a VTE, compared with patients below 300 m elevation.5 Additionally, the normally rare disease of cerebral venous sinus thrombosis has been reported at high altitude in case studies.6 Interestingly, these studies found an increased risk for venous thrombosis, whereas arterial thrombosis does not appear to be more prevalent at high altitude. Over the last two decades, several groups have investigated the influence of various degrees and duration of hypoxia on blood coagulation, but conflicting results were obtained. In a previous study, our group investigated the effect of passive and active ascent from 50 to 3,900 m above sea level on thrombin generation (TG), and found that TG in whole blood (WB) rose with increasing altitude.7 Additionally, activation of coagulation was found in healthy volunteers after exposure to mild hypobaric hypoxia, equivalent to an 8-hour flight at cruising altitude.8 A triple crossover study found similar results, mainly in subjects with factor-V-Leiden who used oral contraceptives.9 However, others did not find an effect of simulated hypoxia comparable to cruising altitude, nor did prolonged profound hypoxia affect coagulation parameters.10–12 Moreover, even decreased thromboelastographic parameters at high altitude, and slightly decreased plasmatic coagulation following a short exposure to a very low peripheral oxygen saturation (SpO2) of 70% were found.13,14 Most studies have used surrogate parameters for TG, like prothrombin fragment 1 þ 2 or thrombin–antithrombin complex, and conventional clotting time tests, e.g. prothrombin time (PT) and activated partial thromboplastin time (APTT), to study the influence of hypoxia on coagulation. Other than these tests, TG assays can visualize the propagation phase, during which the majority of thrombin is generated via feedback loops on factors V, VIII and XI, and the termination phase, during which thrombin formation is shut down by anticoagulant pathways and all thrombin activity is inhibited by plasma protease inhibitors.15 This study aimed to investigate the effect of hypobaric hypoxia at 3,883 m above sea level on TG and platelet Thrombosis and Haemostasis

activation. Additionally, potential confounders such as dehydration and inflammation were tested.

Materials and Methods This study was approved by the Maastricht University medical ethical research committee (METC azM/UM, reference NL49890.068.14), was monitored by the Clinical Trial Center Maastricht, was registered in the Dutch Trial register (reference NTR4806) and met all standards of the Declaration of Helsinki (version 10, 2013). First, group size was calculated using data from a previous study conducted at 3,900 m altitude in 2013 (mean WB endogenous thrombin potential [ETP] 749 nM·min [range, 544–906] at baseline, compared with 1,035 nM·min [range, 822–1,298] at high altitude, estimated standard deviation 220 nM·min).7 To achieve a power of 90% using α ¼ 0.05, at least 10 subjects needed to be recruited. A total of 18 healthy participants aged between 21 and 51 years were included in this study. Exclusion criteria were cardiovascular disease, pulmonary disease, impaired mobility and taking medication interfering with coagulation (heparins, vitamin K antagonists [VKAs], direct oral anticoagulants or non-steroidal anti-inflammatory drugs[ NSAIDs]). After informed consent, but before inclusion, all participants passed a medical assessment by an independent medical doctor, consisting of an electrocardiogram, vital signs (SpO2, heart rate, blood pressure) and auscultation of heart and lungs. Subjects ascended from 50 to 3,883 m above sea level. All participants were transported passively by car, train and cable car and acclimatized for 6 days following a schedule. Day 1 and 2 were spent at 1,930 m, day 3 and 4 at 3,030 m and day 5 at 3,883 m. The volunteers descended to 1,600 m on day 6, after a stay overnight at 3,883 m. Blood was drawn twice: on day 0 at 50 m (baseline) and on day 6 at 3,883 m, before descent to the base camp at 1,600 m. Vital signs (SpO2 and heart rate) were recorded.

Lake Louise Acute Mountain Sickness Questionnaire To record any signs of acute mountain sickness (AMS), participants filled out a questionnaire based on the Lake Louise Consensus on the Definition of Altitude Illness16 on day 0 (50 m altitude), on day 6 after staying overnight at 3,883 m altitude and again on day 6, 5 hours after return to base camp. AMS was defined as the presence of headache (0–3 points) and additionally at least a score of 3 or more points. Symptom score: gastrointestinal symptoms (0–3 points), fatigue (0–3 points), dizziness (0–3 points) and sleeping disturbance (0–3 points). Clinical score: change in mental status (0–3 points), ataxia (0–4 points) and peripheral oedema (0–3 points).

Blood Collection and Testing Blood was collected by venipuncture of the antecubital vein assisted by a tourniquet. The blood was aseptically drawn in Vacutainer tubes (Greiner Bio-One) containing 3.2% sodium citrate (9 volumes blood, 1 volume anticoagulant). Any particularities such as bleeding, bruising, difficult procedure


Hypobaric Hypoxia Increases Thrombin Generation


Blood Count Red blood cell (RBC) count, haemoglobin levels (Hb), haematocrit (Ht), leucocyte count (L) and platelet count (PC) were determined with a Coulter counter within 3 hours of blood collection in citrated WB (Beckman Coulter, Woerden, the Netherlands). For the high altitude measurement, aliquots of WB were transported down to the base camp, keeping the samples at room temperature, and measured after gentle reconstitution. RBC, Hb, Ht, L and PC were corrected for 10% dilution by citrate.

Coagulation Factor Analysis Prothrombin (FII) levels, antithrombin (AT) levels, factor VIII concentration (FVIII:C), Von Willebrand factor antigen (VWF: Ag), fibrinogen levels and D-dimer levels were measured in PPP by batch analysis using the STA-R Evolution (Stago Diagnostica, Paris, France).

Biochemical Markers Albumin as a marker for dehydration, creatinine and urea as markers for kidney function, lactate as a marker for anaerobic metabolism and C-reactive protein as a marker for inflammation were measured with an ARCHITECT ci8200 (Abbott Diagnostics, Hoofddorp, the Netherlands). The method has been validated in plasma for albumin bromocresol purple [BCP] (ref 7D54), creatinine (ref 3L81–22), urea nitrogen (ref 7D75), lactic acid (ref 9D89) and CRP Vario (ref 6K26–30). For this study, another validation against serum was completed and found to be within acceptable ranges (see ►Supplementary Fig. S1, available in the online version).

Whole Blood Thrombin Generation The WB TG assay was performed at 50 and 3,883 m altitude, 30 minutes after blood collection, after addition of 0.5 pM tissue factor (TF; Innovin, Dade Behring, Marburg, Germany). The method has been described previously.17 For the high altitude measurement, the necessary equipment was installed at 3,883 m altitude in a climate-controlled room, to allow measurement in fresh undisturbed blood. In short, 30 µL citrated WB was mixed with 10 µL rhodamine solution (thrombin-specific fluorogenic substrate, 1.8 mM) and activated with 20 µL of a TF/CaCl2 (1.5 pM/50 mM) mixture.

As a calibrator, α2macroglobulin-thrombin complex (in-house prepared α2M-T, 300 nM thrombin activity) was used. After mixing the sample, 5 µL was transferred immediately onto a paper disk (Whatman GmbH) and covered with 40 µL mineral oil (USB Corporation) to prevent evaporation of the sample. TG was measured in triplicate in a 96-well plate fluorometer (Ascent reader, Thermolabsystems OY, Finland) equipped with a 485/538 nm filter set (excitation/emission). Samples were measured for 40 minutes at 37°C. Raw data were converted into thrombograms as described previously.18 Data were expressed as the lag time (LT, minutes), time to peak (TTP, minutes), peak height (Peak, nM) and ETP (nM·min).

Calibrated Automated Thrombinography The calibrated automated thrombinography (CAT) assay was performed in PRP at 50 and 3,883 m altitude, 30 minutes after centrifuging WB to PRP but within 1 hour, and later in PPP by batch analysis. For the measurement of PRP TG at high altitude, the necessary equipment was installed at 3,883 m altitude in a climate-controlled room. The method has been described previously.19 TG was determined in PRP in triplicate in response to 1 pM TF (PRP-reagent, Thrombinoscope, the Netherlands) and in PPP in triplicate in response to 0 and 1 pM TF with addition of 4 μmol/L phospholipid vesicles (MP-reagent and PPP-low reagent, Thrombinoscope, the Netherlands), according to the manufacturer’s instructions, at 37°C. As a calibrator, α2-macroglobulin-thrombin complex (α2M-T, 600 nM thrombin activity, Thrombinoscope, the Netherlands) was used. As fluorogenic substrate, Z-GlyGly-Arg-AMC (FluCa kit, Thrombinoscope, the Netherlands) was used. When measuring TG in PPP, on each 96-well plate a sample of normal pooled plasma (NPP, pooled PPP collected from around 100 volunteers) was added under the same testing conditions. The thrombograms were measured in a 96-well plate fluorometer equipped with a 390/460 nm filter set (excitation/emission) and a dispenser. Immulon 2HB, round-bottom 96-well plates (Dynex) were used. Dedicated software (Thrombinoscope, the Netherlands) calculated the TG curves. For PPP, the NPP results were used for normalization of the peak and ETP.20 Data are expressed as the LT (minutes), TTP (minutes), Peak (nM or % of NPP) and ETP (nM/min or % of NPP).

Factor VIII-Thrombin Generation Assay Contribution of an excess of FVIII to TG was determined. The CAT assay was performed as described earlier in NPP (1 and 5 pM TF). The samples were analysed with or without added recombinant FVIII (Kogenate, Bayer Pharmaceuticals, Germany) at final concentrations of 0, 1, 2 and 5 IU/mL. The plasma thrombograms were calculated by dedicated software (Thrombinoscope, the Netherlands). Data are expressed as the LT (minutes), TTP (minutes), Peak (nM) and ETP (nM·min).

Tissue Factor Pathway Inhibitor-Thrombin Generation Assay Contribution of TFPI to TG was measured by a modified CAT assay according to Maurissen et al.21 This assay measures TFPI contribution to TG by blocking TFPI with anti-TFPI Thrombosis and Haemostasis


and haemolysis were noted. The blood was kept at room temperature (21°C) and all measurements were performed within 2 hours after collection. Platelet-rich plasma (PRP) was obtained by centrifuging the blood at 240 g for 15 minutes and used to perform measurements within 1 hour (see below). Platelet-poor plasma (PPP) was obtained by centrifuging the blood twice at 2,830 g for 10 minutes and immediately snap-frozen on dry ice for later analysis. A centrifuge was placed at 3,883 m altitude to allow rapid processing of the blood. The following tests were performed on the collected blood samples: blood count, coagulation factor analysis, biochemical markers, TG in WB, PRP and PPP, contribution of FVIII and tissue factor pathway inhibitor (TFPI) to the TG, as well as a WB platelet activation test.

Kicken et al.

Kicken et al.

antibodies. The ratio of thrombin peaks determined in the absence and presence of anti-TFPI antibodies, respectively, is a measure of the overall activity of the TFPI system. A dosefinding study for determining the amount of anti-TFPI C-terminus antibodies needed to block all TFPI activity was performed (see ►Supplementary Fig. S2, available in the online version). Supramaximal inhibition was found to occur at 40 μM anti-TFPI C-terminus domain antibodies; therefore, this concentration was used in the experiments. Thermostable inhibitor of the contact activation system (TICA) was added to inhibit contact activation.22 The anti-TFPI CAT assay was performed on all samples as follows: 5 µL anti-TFPI C-terminus domain antibodies (final concentration 40 µM) or 5 µL hydroxyethyl piperazineethanesulfonic acid (HEPES) was added to 80 µL PPP, together with 5 µL TICA (final concentration 0.5 mg/mL) and incubated for 15 minutes at 37°C. Next, 10 µL TF/PL mix (final concentrations 1 pM TF/4 µM PL) was added to the plasma and incubated for 10 minutes at 37°C. As a calibrator, α2-macroglobulinthrombin complex (α2M-T, 600 nM thrombin activity, Thrombinoscope, the Netherlands) was used. Coagulation was started by adding 20 μL FluCa (final concentrations 16.7 mM CaCl2/300 µM Z-Gly-Gly-Arg-AMC) to the sample. The CAT assay was measured as described earlier and the dedicated software (Thrombinoscope, the Netherlands) calculated the TG curves. The ratio of peak height in the absence and presence of anti-TFPI antibodies was calculated to determine contribution of TFPI activity to the TG.

Platelet Activation Test in Unprocessed Blood The platelet activation test in unprocessed blood (PAcT-UB) has been described before23 and was performed at 50 and 3,883 m altitude. WB was added in a 1:20 ratio (5 µL diluted WB (one-fourth) in 20 µL assay) to pre-prepared reaction kits consisting of an agonist, and 87.5 µL/mL fluorescein isothiocyanate (FITC)-conjugated anti-fibrinogen (Dako, Glostrup, Denmark), 87.5 µL/mL phycoerythrin (PE)-conjugated antiP-selectin and 12.5 µL/mL activated protein C (APC)-conjugated anti-CD42b (all purchased from BD Biosciences, Franklin Lakes, United States) in HEPES-buffered saline (HBS; 10 mmol/L HEPES, 150 mmol/L NaCl, 1 mmol/L MgSO4, 5 mmol/L KCl, pH 7.4). Agonists used in the reaction kits were adenosine diphosphate (ADP, 01897, Sigma-Aldrich, Zwijndrecht, the Netherlands), thrombin receptor-activating peptide (TRAP-6, Bachem, Weil am Rhein, Germany) or collagen-related peptide (CRP, a kind gift from Professor Farndale, University of Cambridge, United Kingdom) in final concentrations of 30 µM, 30 µM and 2 µg/mL, respectively. Spontaneous activation was assessed without addition of an agonist. After 20 minutes of incubation at room temperature, reactions were stopped by adding 250 µL fixation solution (137 mmol/L NaCl, 2.7 mmol/L KCl, 1.12 mmol/L NaH2PO4, 1.15 mmol/L KH2PO4, 10.2 mmol/L Na2HPO4, 4 mmol/L ethylenediaminetetraacetic acid (EDTA), 0.5% formaldehyde). Flow cytometry was used to distinguish between platelets and other cells on forward and sideward scatter pattern and by gating on the CD42b positive cells. Fluorescent intensity in the PE gate was selected to determine Thrombosis and Haemostasis

P-selectin density. Fluorescent intensity in the FITC gate was used to determine fibrinogen binding, which indicates αIIbβ3 activation. Results are expressed as median fluorescence intensity (MFI) in arbitrary units (AUs). For the high altitude measurement, the flow cytometer was installed at the base camp and fixed samples were measured there on the day of blood donation.

Data Analysis Data were analysed using SPSS version 23 (SPSS Inc., Chicago, Illinois, United States). Normality of the data was assessed using the D’Agostino K-squared test. Either the paired t-test, or the Wilcoxon signed ranks test in case of a non-parametric distribution, was used to determine statistical significance within participants. The Pearson’s correlation coefficient, or Spearman’s correlation coefficient in case of non-parametric distribution, was used to determine if potential confounders (blood count, dehydration) correlated with WB-TG parameters. A p-value of < 0.05 was considered statistically significant. Data are represented as mean standard deviation (SD) or median interquartile range (IQR) in case of nonparametric distribution. Figures were generated using Prism version 6 (GraphPad Software Inc., La Jolla, United States).

Results Eighteen adult participants were recruited and passed the medical check-up (ratio male:female 11:7, mean age 34.6 years, range, 21–50 years). One participant had to leave the study prematurely (see below). Therefore, results of 17 participants are shown.

Acute Mountain Sickness At 50 m above sea level, none of the participants exhibited signs of AMS (mean score 0.21 0.42 points). At 3,030 m altitude, one participant was excluded due to presence of headache and vomiting, which resolved quickly after descent to 1,600 m. After 24 hours at 3,883 m altitude, participants were experiencing more discomfort (mean score 2.58 2.27 points, p < 0.0001). Five of 17 participants scored positive for headache plus three or more points on other items and were escorted back to base camp immediately after blood withdrawal. After return to 1,600 m, all symptoms resolved quickly (mean score 0.17 0.38 points) and none of the participants required additional medical care.

Vital Signs, Blood Count and Biochemical Markers At high altitude, barometric pressure is lower than at sea level, causing a significant drop in SpO2 after 6 days of incremental exposure to high altitude (p < 0.0001) accompanied by a compensatory increase in heart rate (p ¼ 0.0001) (►Table 1, panel A). The haemoglobin levels, haematocrit and RBC count rose slightly but significantly with p-values of 0.004, 0.0002 and 0.0002, respectively (►Table 1, panel B). This increase was not due to dehydration, as albumin and creatinine were unchanged (p ¼ 0.073 and p ¼ 0.16, respectively), and urea was significantly lower at high altitude (p ¼ 0.0003) (►Table 1, panel C). Leucocyte count rose




Kicken et al.

Table 1 Effect of high altitude on vital signs, blood count, biochemical markers and coagulation factor levels 50 m

3,883 m

A. Vital signs SpO2 (%)

98 –1

Heart rate (min )

88b

1.1

3.3

b

67

11

79

9.2

0.67

9.5b

15

B. Blood count Hb (mmol/L) Ht (L/L)

0.43 12

0.03

0.73

0.47

b

0.04

b

0.54

RBC (·10 /L)

5.02

0.42

5.46

L (·109/L)

6.5

1.00

7.4b

2.00

290

52

284

55

9

PC (·10 /L) Albumin (g/L)

36.1

2.25

34.5

3.64

Creatinine (μmol/L)

68.7

9.24

65.1

13.0

Urea (mmol/L)

4.41

1.19

3.42b

0.94

0.9

0.78a

C-reactive protein (mg/L)

0.7

0.64

a

1.39

0.38

2.10

FII (%)

104

7.2

104

Lactate (mmol/L)

b

0.46

D. Coagulation factor levels 8.0

AT (%)

115

8.2

115

5.1

VWF:Ag (%)

107

23

107

28

FVIII:C (%)

109

28

118c

34

Fibrinogen (g/L)

2.92

0.43

2.94

0.47

D-dimers (µg/mL)

0.34

0.11

0.35

0.14

Abbreviations: AT, antithrombin; FII, prothrombin; FVIII:C, factor VIII concentration; Hb, haemoglobin level; Ht, haematocrit; L, leukocyte count; PC, platelet count; RBC, red blood cell count; SpO2, peripheral oxygen saturation; VWF:Ag, Von Willebrand Factor antigen. Note: Vital signs, blood count, biochemical markers and coagulation factor levels were measured at 50 and 3,883 m above sea level (n ¼ 17). Data are expressed as mean standard deviation. a Non-parametric distribution. b p < 0.001. c p < 0.05.

slightly at high altitude (p ¼ 0.002), but was not accompanied by a significant increase in C-reactive protein (p ¼ 0.15). PC did not change (p ¼ 0.55). Lactate levels were slightly elevated (p < 0.0001) (►Table 1, panel C).

Coagulation Factor Levels Results considering coagulation factor levels are listed ►Table 1(panel D). Compared with sea level, at high altitude only FVIII:C was increased (p ¼ 0.0075). There were no significant changes in FII levels (p ¼ 0.84), AT (p ¼ 0.62), VWF:Ag levels (p ¼ 0.79), fibrinogen (p ¼ 0.71) or D-dimer levels (p ¼ 0.74).

Thrombin Generation The contribution of blood cells to TG was determined by measuring TG in WB, in PRP (where platelets and coagulation factors are present) and in PPP (where coagulation factors and microparticles are present). Hb, Ht and RBC, potential confounders of a WB assay, did not correlate significantly

with WB-TG parameters, therefore the WB-TG results were not corrected for Hb, Ht or RBC. In WB initiated with 0.5 pM TF (►Fig. 1A), mean Peak increased from 146 to 187 nM (p ¼ 0.0004) and mean ETP increased from 684 to 860 nM·min (p ¼ 0.0002), whereas LT and TTP remained unchanged (3.8–3.6 minutes, p ¼ 0.06 and 7.4–7.4 minutes, p ¼ 0.7). In PRP triggered by 1 pM TF (►Fig. 1B), Peak increased from 118 to 145 nM (p ¼ 0.0006) and ETP increased from 1403 to 1885 nM·min (p < 0.0001). Again, LT did not change (9.8–10.4 minutes, p ¼ 0.2) and TTP was slightly prolonged from 18.9 to 20.8 minutes (p ¼ 0.03). The increased TTP can be explained by the velocity index that did not increase significantly (14 4.65 nM/min at sea level to 14.9 4.5 nM/min at 3,883 m altitude, p ¼ 0.17, not shown in figure), while ETP was increased. This means that the rate of thrombin formation was not increased, whereas more thrombin was formed, leading to a prolonged TTP. Thrombosis and Haemostasis


C. Biochemical markers

Kicken et al.

Fig. 1 High altitude increases thrombin generation in whole blood (WB), platelet-rich plasma (PRP) and platelet-poor plasma (PPP). Thrombin generation was measured at 50 and 3,883 m above sea level in WB (0.5 pM TF, A), in PRP (1 pM TF, B) and in PPP with 4 μmol/L phospholipid vesicles in the absence of TF (C), 1 pM TF (D) and 5 pM TF (E). Parameters derived from the thrombin generation curve are lag time (LT), time to peak (TTP), peak height (Peak) and endogenous thrombin potential (ETP). Results from PPP Peak and ETP were normalized against normal pooled plasma (NPP) and are expressed as % of NPP. Data are expressed as mean standard deviation (SD) or median interquartile range (IQR) in case of non-parametric distribution. 5, p < 0.05, 55, p < 0.001. 1, non-parametric distribution.

In PPP triggered by 1 pM TF (►Fig. 1D), an increased Peak (from 195 to 272%, p < 0.0001) and increased ETP (209– 245%, p ¼ 0.0003) were found at high altitude, and LT and TTP were unchanged compared with sea level (5.9–6.2 minutes, p ¼ 0.087 and 11.8–11.6 minutes, p ¼ 0.38, respectively). TG was also measured in PPP without addition of TF, an assay sensitive for increased coagulation caused by, i.e. microparticles exposing TF or negatively charged phospholipids (►Fig. 1C). This assay also showed increased Peak (from 153 to 305%, p 0.0001) and increased ETP (159–225%, p ¼ 0.0003), while LT and TTP were unchanged compared with sea level (19.3–19.8 minutes, p ¼ 0.83 and 24.1–23,9 minutes, p ¼ 0.89, respectively). To test whether the extrinsic pathway is responsible for the effects found, TG was measured in PPP triggered by 5 pM TF (►Fig. 1E). LT and ETP were unchanged compared with sea level (from 2.91 to 2.93 minutes, p ¼ 0.67 and from 152 to 143%, p ¼ 0.08, respectively). Contrary to TG triggered with 1 pM TF, the TTP increased (from 6.08 to 6.38 minutes, Thrombosis and Haemostasis

p ¼ 0.0011) and Peak decreased (190 to 166%, p ¼ 0.0003) at high altitude. Spearman’s correlation was used to see if FVIII levels correlated with TG results. A significant correlation was found between FVIII levels and PPP Peak at 50 m altitude (0 pM TF: r ¼ 0.55 p ¼ 0.023; 1 pM TF: r ¼ 0.80 p ¼ 0.0003; and 5 pM TF: r ¼ 0.57 p ¼ 0.019, respectively) and at 3,883 m altitude (0 pM TF: r ¼ 0.55 p ¼ 0.03; 1 pM TF: r ¼ 0.65 p ¼ 0.005; and 5 pM TF: r ¼ 0.53 p ¼ 0.03, respectively). There was a weak to no correlation between FVIII levels and PPP ETP. FVIII levels did not correlate with Peak or ETP in PRP and WB.

Contribution of an Excess of Factor VIII to Thrombin Generation Following the aforementioned results, increased TG at high altitude appears to be mediated by increased FVIII:C. Therefore, the contribution of an excess of FVIII to TG was tested by spiking NPP with 0, 1, 2 or 5 IU/mL recombinant FVIII (►Fig. 2). Incremental concentrations of FVIII had the most




Kicken et al.

effect on TG induced by 1 pM TF; 5 IU/mL FVIII shortened the LT from 4 to 3 minutes, and shortened the TTP from 8 to 5 minutes. When TG was induced by 5 pM of TF, spiking with FVIII had a minimal effect and only a small reduction of LT from 1.67 to 1.33 minutes and of TTP from 3.67 to 3 minutes was seen with 5 IU/mL FVIII. Additionally, an incremental elevation of Peak and ETP was seen with increasing FVIII concentrations. Again, the effect of FVIII was more pronounced when TG was triggered with 1 pM TF; 5 IU/mL FVIII increased the Peak from 105 to 417 nM and ETP from 785 to 1,511 nM·min. When TG was triggered with 5 pM TF, the Peak and ETP increased from 335 to 450 nM and from 1256 to 1513 nM·min, respectively.

Contribution of Tissue Factor Pathway Inhibitor to Thrombin Generation A modified anti-TFPI TG assay was performed to study whether a decreased activity of TFPI was responsible for the elevated TG at high altitude (see ►Supplementary Fig. S3, available in the online version). No significant difference between samples drawn at 50 and 3,883 m altitude was observed (p ¼ 0.30), the mean Peak ratios were 0.23 0.13 at sea level and 0.24 0.13 at high altitude, respectively.

Platelet Activation To test the effect of altitude on platelet activation, platelet Pselectin expression and αIIbβ3 activation after addition of agonists (ADP, CRP or TRAP) or without agonist was measured. Spontaneous platelet activation was lower at high altitude compared with sea level, measured by a lower Pselectin expression (MFI decreased from 111 to 86 AU, p ¼ 0.0031) and a lower αIIbβ3 activation (MFI decreased from 849 to 683 AU, p ¼ 0.0032). At high altitude, platelets activated with agonists expressed less P-selectin with MFI values at sea level/high altitude of 5,813/4,085 AU (p < 0.0001), 8,739/7,496 AU (p < 0.0001) and 8,000/6,849 AU (p < 0.0001) for ADP, TRAP and CRP, respectively. Also, a

lower expression of activated αIIbβ3 was observed at high altitude with MFI values at sea level/high altitude of 71,961/ 57,052 AU (p < 0.0001), 64,530/53,340 AU (p ¼ 0.001) and 72,164/68,397 AU (p < 0.0001) for ADP, TRAP and CRP, respectively. Results are shown in ►Fig. 3.

Discussion There is an on-going debate about the influence of hypoxia at high altitude on venous thromboembolic risk. The risk of developing a VTE is influenced by the presence of risk factors, such as age, sex, previous VTE, coagulation disorders, obesity, recent surgery, immobilization and cancer.24 Some authors tentatively comment that hypoxic stress can be an additional risk factor.25 Others are confident enough to state that hypoxia deserves a place in the famous Virchow’s triad.26 Opposing authors conclude that altitude does not cause VTE, due to the absence of elevated D-dimers in Mount Everest climbers at an altitude of 5,340 m.27 Interestingly, VTE but not arterial thrombosis is a point of debate. We found that exposure to hypobaric hypoxia caused elevated TG, measured in WB, PRP and PPP without or at low TF. However, in PPP triggered with a high concentration of TF, a slightly prolonged TTP and decreased Peak were found. Furthermore, FVIII levels were significantly higher at high altitude compared with levels at sea level, whereas other coagulation factor levels and D-dimers remained unchanged. The FVIII levels at 50 and 3,883 m altitude correlated significantly with Peak in PPP, but not in PRP or WB. These results are in agreement with recently published work, that found that FVIII levels are a determinant for TG in PPP, but not for TG in PRP or WB with added TF.28 Together, these findings point towards increased activation of the intrinsic coagulation pathway at high altitude, an effect that is abolished by high TF concentration. As shown in vitro, FVIII has a larger effect on the TG when induced by 1 pM TF compared with 5 pM TF. This Thrombosis and Haemostasis


Fig. 2 An excess of FVIII increases thrombin generation, predominantly at a low tissue factor (TF) concentration. Normal-pooled plasma (NPP) was spiked with or without recombinant FVIII (0, 1, 2 and 5 U/mL) and thrombin generation was triggered with 1 pM TF (A) and 5 pM TF (B).

Kicken et al.

Fig. 3 High altitude decreases both spontaneous and agonist-induced platelet activation. Whole blood was incubated for 20 minutes at room temperature with adenosine diphosphate (ADP), thrombin receptor-activating peptide (TRAP-6) and collagen-related peptide (CRP), or without agonist (control). Thereafter, platelet activation was measured by P-selectin expression and αIIbβ3 activation and expressed as median fluorescence intensity (MFI) in arbitrary units (AU). Individual data points are shown, the error bars depict mean standard deviation (SD) or median interquartile range (IQR) in case of non-parametric distribution. 5, p < 0.05, 55, p < 0.001. 1, non-parametric distribution.

could explain the discrepancy between low and high TFinduced TG in PPP at high altitude. TG is determined not just by procoagulant factors such as FVIII, but also by the endogenous anticoagulant system, consisting of antithrombin, protein C, protein S and TFPI.15 Antithrombin levels did not change due to high altitude. Additionally, the activity of TFPI was studied using a modified TG assay; however, TFPI activity did not change at high altitude. APC levels or activity were not measured due to shortage of PPP from high altitude, but may be another interesting focus. APC is the naturally occurring inhibitor of FVIII, therefore a decrease of APC level or activity could contribute to the increased TG at high altitude. This could be assessed by performing a thrombomodulin-modified TGassay. Of note, the possible confounders dehydration and inflammation were not observed at high altitude. The current expedition was a follow-up study, after a previous high-altitude expedition carried out in 2013 by our group.7 In this previous study, TG was studied at different altitudes in both an actively climbing group and a passively ascending group. In both groups, increased TG at high altitude was found in WB (passive group: mean ETP from 712 to 1,053 nM·min at 3,883 m, p ¼ 0.0015), but not in PPP (passive group: mean ETP from 794 to 850 nM·min at 3,883 m, p ¼ 0.19). Forthcoming, activation of coagulation was attributed to an influence of hypoxia on an unknown part of cellular haemostasis. The current results are largely, but not entirely in agreement with results from the previous expedition. Although PPP TG did not increase in the previous study, FVIII levels did rise at 3,883 m altitude (passive group: Thrombosis and Haemostasis

mean FVIII from 110 to 120%, p ¼ 0.0071), which is in agreement with our current findings. It is unknown why this increase in FVIII did not increase PPP TG then, but does so now. At high altitude, the healthy volunteers suffered from tissue hypoxia, as shown by the decreased SpO2 and the small albeit consistent increase in lactate levels. Hypoxia may influence coagulation via alteration of the redox status of the blood. In a recent study, healthy subjects were exposed to normobaric hypoxia (12% O2) for 6 hours.29 In that study, hypoxia caused increased reactive oxygen species (ROS), which was accompanied by a shortened APTT, whereas international normalized ratio (INR) did not change significantly. The authors suggested that elevated ROS causes activation of coagulation via the intrinsic pathway. This study is limited by inappropriate use of the INR, since the INR is an optimized measurement for monitoring VKA therapy, and was not developed for ruling out extrinsic coagulation activation. Additionally, in another study it was found that hypoxia causes increased TG due to elevated FVIII levels and activity.30 In that study, healthy volunteers were exposed to different degrees of normobaric hypoxia or normoxia for 2 hours, with or without pre-treatment with the antioxidant vitamin E. It was found that severe hypoxia (12% O2) caused FVIII-mediated elevation of TG, but not in the group treated with vitamin E. The authors concluded that hypoxia causes elevated FVIII-mediated TG via increased oxidative stress in the blood. Combining these findings with the current results, we conclude that the hypercoagulability observed at high altitude is presumably due to




What is known about this topic? • Venous thromboembolic events occur more frequently at high altitude. • High altitude inhabitants have a lower mortality due to arterial thrombotic events than lowland inhabitants.

What does this paper add? • Hypobaric hypoxia causes elevated thrombin generation mediated by increased factor VIII levels. • Platelet activation is consistently decreased at high altitude. • These findings may explain why venous and not arterial thrombotic events occur more frequently at high altitude.

Funding This work was supported by departmental funds from Synapse Research Institute.

Conflict of Interest None. Acknowledgments The authors are greatly indebted to Iris Thuis, Edward and Ilse Bekker for their invaluable logistical assistance. Furthermore, the authors thank Leslie In ’t Panhuis for assisting with data analysis. Last but not the least, the authors are grateful to all volunteers who generously made time to participate in this study.

References 1 West JB. Physiological effects of chronic hypoxia. N Engl J Med

2017;376(20):1965–1971 2 Peacock AJ. ABC of oxygen: oxygen at high altitude. BMJ 1998;317

(7165):1063–1066 3 Anand AC, Jha SK, Saha A, Sharma V, Adya CM. Thrombosis as a

4

5

6

7

8

complication of extended stay at high altitude. Natl Med J India 2001;14(04):197–201 Anand AC, Saha A, Seth AK, Chopra GS, Nair V, Sharma V. Symptomatic portal system thrombosis in soldiers due to extended stay at extreme altitude. J Gastroenterol Hepatol 2005;20(05):777–783 Tyson JJ, Bjerke BP, Genuario JW, Noonan TJ. Thromboembolic events after arthroscopic knee surgery: increased risk at high elevation. Arthroscopy 2016;32(11):2350–2354 Zavanone C, Panebianco M, Yger M, et al. Cerebral venous thrombosis at high altitude: a systematic review. Rev Neurol (Paris) 2017;173(04):189–193 Ninivaggi M, de Laat M, Lancé MD, et al. Hypoxia induces a prothrombotic state independently of the physical activity. PLoS One 2015;10(10):e0141797 Bendz B, Rostrup M, Sevre K, Andersen TO, Sandset PM. Association between acute hypobaric hypoxia and activation of coagulation in human beings. Lancet 2000;356(9242):1657–1658 Thrombosis and Haemostasis


increased FVIII levels, which may be elevated due to oxidative stress in the systemic circulation. The elevated TG was accompanied by decreased platelet activation at high altitude, both spontaneous and agonistinduced. Strikingly, this has been found before in various studies. For instance, already in 1982 it was found that acute induction to 3,700 m caused decreased platelet aggregation up until 5 days after ascent.31 More recently, a stay at 4,100 m or higher up until 13 months duration was found to significantly reduce maximal platelet aggregation induced by ADP, epinephrine and collagen.32 Moreover, in a recent in vitro study, severe hypoxia (1 and 8% O2) caused reversible decreased platelet aggregation mediated by the fibrinogen αIIbβ3 receptor.33 Interestingly, VTE seems to be more prevalent at high altitude, whereas mortality from arterial thrombotic events, such as myocardial infarction and stroke, is lower in high altitude inhabitants.3,4,34,35 Arterial thrombotic events are characterized by rupture of an atherosclerotic plaque with a “white” platelet-rich thrombus, whereas a VTE is traditionally described as a “red” thrombus that is low in platelet content.36 Decreased platelet activation at high altitude may therefore be protective of arterial thrombotic events in the hypoxia-induced hypercoagulable state. Platelet function may be influenced by physiological changes that occur during adaptation to high altitude. For instance, adenosine is released in response to hypoxia, via soluble ecto-5′-nucleotidase, a key enzyme to generate extracellular adenosine.37 The adenosine causes regional vasodilation, hereby enhancing blood delivery to hypoxic tissues. Interestingly, it is also an inhibitor of platelet aggregation.38 Furthermore, elevation of adenosine induces activation of the erythrocyte A2B adenosine receptor. Downstream of the A2B adenosine receptor, adenosine monophosphate (AMP)-activated protein kinase activates diphosphoglycerate mutase, which produces 2,3 diphosphoglycerate (2,3-DPG).37 The 2,3-DPG is an allosteric modulator of haemoglobin-O2 affinity. It facilitates O2 release from haemoglobin, hereby augmenting O2 delivery to hypoxic tissues. At the same time 2,3-DPG inhibits platelet aggregation.39 These two mechanisms likely cause decreased platelet activation at high altitude. In conclusion, we found that exposure to hypobaric hypoxia at 3,883 m above sea level causes FVIII-mediated increased TG in healthy volunteers. Via this mechanism the risk of VTE may be increased at high altitude. On the other hand, we found decreased platelet activation at high altitude. Interestingly, decreased mortality from an arterial thrombotic event at high altitude is observed, an effect that is likely mediated by decreased platelet activation. Our findings corroborate previous experimental work,7–9 although several other studies showed no effect of hypoxia on blood coagulation.10–14 There is a wide variability in the degree and duration of exposure to the hypoxic environment in these experimental studies. Additional research is needed to determine the exact degree of tissue hypoxia and duration of exposure that cause clinically relevant hypercoagulability.

Kicken et al.

Kicken et al.

9 Schreijer AJM, Cannegieter SC, Meijers JCM, Middeldorp S, Büller

10

11

12

13

14

15

16

17

18 19

20

21

22

23

HR, Rosendaal FR. Activation of coagulation system during air travel: a crossover study. Lancet 2006;367(9513):832–838 Toff WD, Jones CI, Ford I, et al. Effect of hypobaric hypoxia, simulating conditions during long-haul air travel, on coagulation, fibrinolysis, platelet function, and endothelial activation. JAMA 2006;295(19):2251–2261 Crosby A, Talbot NP, Harrison P, Keeling D, Robbins PA. Relation between acute hypoxia and activation of coagulation in human beings. Lancet 2003;361(9376):2207–2208 Venemans-Jellema A, Schreijer AJM, Le Cessie S, Emmerich J, Rosendaal FR, Cannegieter SC. No effect of isolated long-term supine immobilization or profound prolonged hypoxia on blood coagulation. J Thromb Haemost 2014;12(06):902–909 Martin DS, Pate JS, Vercueil A, Doyle PW, Mythen MG, Grocott MP; Caudwell Xtreme Everest Research Group. Reduced coagulation at high altitude identified by thromboelastography. Thromb Haemost 2012;107(06):1066–1071 Redford DT, Paidy SR, Steinbrenner EB, Nielsen VG. Effects of profound hypoxemia on coagulation & fibrinolysis in normal individuals. Blood Coagul Fibrinolysis 2016;27(02):228–231 Hemker HC, Al Dieri R, De Smedt E, Béguin S. Thrombin generation, a function test of the haemostatic-thrombotic system. Thromb Haemost 2006;96(05):553–561 Roach RC, Bärtsch P, Hackett PH, et al. The Lake Louise Consensus on the Quantification of Altitude Illness. Hypoxia Mol Med Burlington, VT: Queen City Press; 1993:272–274 Ninivaggi M, Apitz-Castro R, Dargaud Y, de Laat B, Hemker HC, Lindhout T. Whole-blood thrombin generation monitored with a calibrated automated thrombogram-based assay. Clin Chem 2012;58(08):1252–1259 Hemker HC, Kremers R. Data management in thrombin generation. Thromb Res 2013;131(01):3–11 Hemker HC, Giesen P, Al Dieri R, et al. Calibrated automated thrombin generation measurement in clotting plasma. Pathophysiol Haemost Thromb 2003;33(01):4–15 Spronk HMH, Dielis AWJH, De Smedt E, et al. Assessment of thrombin generation II: validation of the calibrated automated thrombogram in platelet-poor plasma in a clinical laboratory. Thromb Haemost 2008;100(02):362–364 Maurissen LFA, Castoldi E, Simioni P, Rosing J, Hackeng TM. Thrombin generation-based assays to measure the activity of the TFPI-protein S pathway in plasma from normal and protein Sdeficient individuals. J Thromb Haemost 2010;8(04):750–758 Hackeng TM, Suijlen DPL, Hemker HC, et al. Thermostable inhibitors of activation of the blood clotting system through contact with foreign surfaces. Available at: http://www.google.com/ patents/WO2013028069A1?cl¼en. Accessed July 31, 2017 Roest M, van Holten TC, Fleurke G-J, Remijn JA. Platelet activation test in unprocessed blood (Pac-t-UB) to monitor platelet concentrates and whole blood of thrombocytopenic patients. Transfus Med Hemother 2013;40(02):117–125

Thrombosis and Haemostasis

24 Rosendaal FR. Venous thrombosis: a multicausal disease. Lancet

1999;353(9159):1167–1173 25 Gupta N, Ashraf MZ. Exposure to high altitude: a risk factor for

26

27

28

29

30

31

32 33

34

35

36 37

38

39

venous thromboembolism? Semin Thromb Hemost 2012;38(02): 156–163 Hackett TR, Godin JA. Editorial commentary: should the Virchow triad have been a quartet? Is high altitude a risk factor for deep venous thrombosis after knee arthroscopy?. Arthroscopy 2016;32 (11):2355–2356 Zafren K, Feldman J, Becker RJ, Williams SR, Weiss EA, Deloughery T. D-dimer is not elevated in asymptomatic high altitude climbers after descent to 5340 m: the Mount Everest Deep Venous Thrombosis Study (Ev-DVT). High Alt Med Biol 2011; 12(03):223–227 Bloemen S, Huskens D, Konings J, et al. Interindividual variability and normal ranges of whole blood and plasma thrombin generation. J Appl Lab Med AACC Publ 2017;2:150 Fall L, New KJ, Evans KA, Bailey DM. Arterial hypoxaemia and its impact on coagulation: significance of altered redox homeostasis. J Clin Pathol 2015;68(09):752–754 Wang J-S, Cheng M-L, Yen H-C, Lou BS, Liu HC. Vitamin E suppresses enhancement of factor VIII-dependent thrombin generation by systemic hypoxia. Stroke 2009;40(02):656–659 Chatterji JC, Ohri VC, Das BK, et al. Platelet count, platelet aggregation and fibrinogen levels following acute induction to high altitude (3200 and 3771 metres). Thromb Res 1982;26(03):177–182 Vij AG. Effect of prolonged stay at high altitude on platelet aggregation and fibrinogen levels. Platelets 2009;20(06):421–427 Kiouptsi K, Gambaryan S, Walter E, Walter U, Jurk K, Reinhardt C. Hypoxia impairs agonist-induced integrin αIIbβ3activation and platelet aggregation. Sci Rep 2017;7(01):7621 Faeh D, Gutzwiller F, Bopp M; Swiss National Cohort Study Group. Lower mortality from coronary heart disease and stroke at higher altitudes in Switzerland. Circulation 2009;120(06): 495–501 Ezzati M, Horwitz MEM, Thomas DSK, et al. Altitude, life expectancy and mortality from ischaemic heart disease, stroke, COPD and cancers: national population-based analysis of US counties. J Epidemiol Community Health 2012;66(07):e17 Diamond SL. Systems analysis of thrombus formation. Circ Res 2016;118(09):1348–1362 Liu H, Zhang Y, Wu H, et al. Beneficial role of erythrocyte adenosine A2B receptor-mediated AMP-activated protein kinase activation in high-altitude hypoxia. Circulation 2016;134(05): 405–421 Johnston-Cox HA, Yang D, Ravid K. Physiological implications of adenosine receptor-mediated platelet aggregation. J Cell Physiol 2011;226(01):46–51 Iatridis PG, Hadd H, Kotrotsou M, Ling-Indeck L, Iatridis SG. Structure-function relationship of 3-phosphoglycerate analogues with platelet aggregation and thromboxane A2 formation. Haemostasis 1987;17(1-2):8–13

