Journal of Veterinary Emergency and Critical Care CO-oximetry measurements and antioxidant effects of ascorbic acid and methylene blue in equine methaemoglobinaemic blood Journal: Journal of Veterinary Emergency and Critical Care Manuscript ID JVECC-19-02-0008.R2 Fo Wiley - Manuscript type: Original Study Keyword: CO-oximeter, methaemoglobin, horse, methylene blue, ascorbic acid rP ee rR ev iew Page 1 of 22 Journal of Veterinary Emergency and Critical Care 3 1 Summary 2 Objectives: The purpose of this study was to determine the effects of time after 3 sampling on CO-oximetry measurements of equine blood samples and the effects 4 of adding ascorbic acid (AscAc) and methylene blue (MetBlue) to samples with in 5 vitro induced methaemoglobinaemia. 6 Design: Experimental study. 7 Setting: University teaching hospital. 8 Animals: Thirty healthy adult horses assigned to 5 different groups. Fo 9 Interventions: Repeated CO-oximetry determinations were performed on venous rP 10 (n=6) and arterial blood samples (n=7) stored at 0°C (ice water) for 48h. 11 Methaemoglobinaemia was induced in vitro in 17 additional blood samples. Six ee 12 were used as untreated controls, six had AscAc and five had MetBlue added. Total rR 13 haemoglobin, oxyhaemoglobin (O2Hb), carboxyhaemoglobin, methaemoglobin ev 14 (MetHb) and oxygen saturation of haemoglobin were measured. Samples were 15 kept in 1ml airtight plastic syringes. iew 16 Measurements and main results: Oxyhaemoglobin and SO2 increased from 17 69.8±10.2% and 90±3% to 82.8±7.9% and 99±3% respectively after 8h in venous 18 blood (mean±SD, p<0.001). There was an effect of treatment (p=0.032) and of time 19 (interaction p=0.003) on MetHb% in methaemoglobinaemic samples. The 20 difference in absolute MetHb% from time 0 were as follows: 7.0% [IQR:-1.2, 20.0], 21 -0.2% [IQR:-2.0, 1.5] and -4.4% [IQR:-7.0, -1.8] at 48h in Control, AscAc and 22 MetBlue groups respectively (p<0.05). There was no effect of time on MetHb% in 23 the AscAc group (23% [IQR:9.2, 61.8] at time 0 to 23.2% [IQR:7.0, 63.6] after 48h). 3 Journal of Veterinary Emergency and Critical Care Page 2 of 22 4 24 Conclusions: Storage of blood in ice water to determine HbO2 and SO2 using a 25 CO-oximeter should not exceed 4h. Measurement of MetHb% could be delayed by 26 up to 48h if AscAc is added to the sample. Methylene blue significantly decreased 27 MetHb% over time. The limitations of this study include the fact that the antioxidant 28 effects of AscAc and MetBlue were evaluated in vitro and not in vivo. Further 29 studies are needed to evaluate different storage temperatures and syringe types. 30 31 Abbreviations Fo 32 AscAc Ascorbic acid rP 33 COHb Carboxyhaemoglobin 34 Hb Haemoglobin ee 35 MetBlue Methylene blue rR 36 MetHb Methaemoglobin ev 37 O2Hb Oxyhaemoglobin 38 SO2 percentage oxygen saturation of haemoglobin iew 39 TotHb Total haemoglobin 4 Page 3 of 22 Journal of Veterinary Emergency and Critical Care 5 40 INTRODUCTION 41 CO-oximetry is the gold standard method of measurement of oxygen saturation of 42 haemoglobin and provides the spectrophotometric measurement of total 43 haemoglobin as well as the percentages of the main haemoglobin derivatives, 44 which include oxyhaemoglobin (O2Hb), deoxyhaemoglobin, carboxyhaemoglobin 45 (COHb) and methaemoglobin (MetHb).1 CO-oximetry is mainly used in the critical 46 care setting and is indicated in patients in which hypoxaemia does not improve with 47 oxygen therapy, when there is a history consistent with toxin exposure or when the Fo 48 clinician suspects dyshaemoglobinaemias (i.e.: methaemoglobinaemia or rP 49 carboxyhaemoglobinaemia).1 50 Measurement of oxygen status and haemoglobin (Hb) species is problematic, as ee 51 oxygenation parameters as well as percentages of different Hb species change rR 52 after the blood has been sampled. Rapid changes in oxygenation parameters are ev 53 reported in human and animals after blood samples have been drawn from these 54 patients.2-5 Most studies in animals on the effects of time of storage on blood iew 55 oxygenation parameters have used standard blood gas machines. These calculate 56 oxygen haemoglobin saturation (SO2) based on the PO2 instead of measuring SO2 57 or O2Hb directly. CO-oximetry measures O2Hb percentage and total Hb allowing 58 for an accurate determination of the SO2. It has been shown that CO-oximetry 59 provides more precise SO2 and O2Hb measurements compared to standard blood 60 gas machines.6 Although the effects of exercise on SO2 or O2Hb as measured by 61 CO-oximetry have been determined on equine whole blood stored at 0°C in (ice 5 Journal of Veterinary Emergency and Critical Care Page 4 of 22 6 62 water),7 the effect of time of storage on equine blood CO-oximetry values has not 63 been previously described. 64 Methaemoglobin concentrations also vary after sampling due to the instability of 65 MetHb.8 The efficacy of ascorbic acid (AscAc) and methylene blue (MetBlue) for 66 the treatment of horses with methaemoglobinaemia has been questioned. 9 In 67 addition it has been shown that in contrast to other species, equine MetHb is more 68 sensitive to oxidation10 and has a decreased reduction rate by MetBlue using 69 equine blood samples. 11 In a previous in vitro study using equine washed red Fo 70 blood cells, MetBlue did not have an effect on MetHb%.11 However the antioxidant rP 71 effects of AscAc and MetBlue on equine whole blood and the effects of time after 72 sampling on samples to which these compounds were added is unknown. ee 73 rR 74 Objectives ev 75 The purpose of this study was to determine the effects of time after sampling on 76 CO-oximetry measurements of equine blood samples and the effects of adding iew 77 AscAc and MetBlue to samples with in vitro induced methaemoglobinaemia. We 78 hypothesized that 1) there would be an effect of a prolonged delay-in-time from 79 sampling to determination of CO-oximetry measurements in arterial and venous 80 blood from healthy horses and that 2) there would be an antioxidant effect of 81 methylene-blue and ascorbic-acid on equine blood samples with in vitro induced 82 methaemoglobinaemia. A tertiary hypothesis was that 3) there would be an effect 83 of a prolonged delay-in-time from sampling to determination of methaemoglobin in 84 methaemoglobinaemic blood samples. 85 6 Page 5 of 22 Journal of Veterinary Emergency and Critical Care 7 86 MATERIALS AND METHODS 87 Animals 88 Venous blood was collected from the jugular vein of 23 healthy horses, and arterial 89 blood samples were collected from the transverse facial artery in another seven 90 horses (4-25 years of age and mean age of 11.6, 17 males and 13 females). The 91 effect of time after sampling on CO-oximetry measurements was evaluated in the 92 arterial blood from seven and venous blood from six healthy horses respectively. 93 Venous blood from the remaining 17 horses was used to determine the effects of Fo 94 delay-in-time from sampling to determination of MetHb in methaemoglobinaemic rP 95 equine venous blood (6 horses) and to evaluate the antioxidant effects of AscAc (6 96 horses) and MetBlue (5 horses) on whole venous blood with in vitro induced ee 97 methaemoglobinaemia. Animals were randomly allocated to the 3 different groups. rR 98 Sample size was calculated following a resource equation approach, using a ev 99 design specific for group comparison, one between and one within factor, 100 repeated-measures ANOVA, as previously described.13 iew 101 Quality controls for the CO-oximeter included an initial calibration with all liquid 102 controls (RNA CO-Oximeter controlsa). Weekly testing of one different level of 103 liquid control was performed thereafter. In addition, daily optical quality controls 104 were performed during the whole testing period. 105 106 Procedures 107 Effects of time after sampling on percentage O2Hb, COHb, MetHb and SO2 as 108 measured by CO-oximetry (Avoximeter 4000b) were evaluated in arterial and 109 venous equine blood samples. Aliquots of 1 ml of blood were stored in air tight 7 Journal of Veterinary Emergency and Critical Care Page 6 of 22 8 110 sealed plastic syringes placed in ice water and CO-oximetry measurements 111 repeated at 0, 1, 2, 4, 8 and 24 hours after sampling. Blood samples were drawn in 112 1ml plastic syringes (ENFA 1 ml. Tub. 3C-Luer cen.c) through 22G 2.5cm (SOL- 113 MTM hypodermic needled) needles previously washed with liquid heparin (5000 114 IU/ml) left to fill the dead space. A rubber cup was immediately placed at the end of 115 the needle to create an airtight environment. Syringes were filled to total capacity 116 and then inverted (at least 20 times) and rolled during 60 seconds after sampling. 117 To determine the effects of delay-in-time from sampling to determination of MetHb Fo 118 percentage (MetHb%) and to evaluate the antioxidant effects of MetBlue and rP 119 AscAc on methaemoglobinaemic equine blood samples, methaemoglobinaemia 120 was induced in vitro as described elsewhere.14 Briefly, a 0.1M stock solution of ee 121 sodium nitrite was prepared. A variable percentage of MetHb was induced in vitro rR 122 by adding equal amounts (0.1 ml) of sodium nitrate solution using a 2-fold serial ev 123 dilution factor up to a 64-fold total dilution of the original stock solution in 1 ml of 124 blood. Venous blood was incubated with sodium nitrate for 20 minutes at room iew 125 temperature. Serial CO-oximetry measurements were performed at 0, 8, 24 and 126 48h after incubation for the 3 groups namely: {1} Control group (Control), {2} a 127 group with added AscAc and {3} a group with added MetBlue. 128 Briefly, for the AscAc group, 1 ml of blood with in vitro induced 129 methaemoglobinaemia was exposed to 0.1 ml (0.03 uM) of a freshly prepared 130 solution of AscAc. For the MetBlue group, 1 ml of blood with in vitro induced 131 methaemoglobinaemia was exposed to 0.1 ml (10 uM) of a freshly prepared 132 MetBlue solution. Control group samples were composed of methaemoglobinaemic 133 samples to which 0.1 ml of normal saline solution was added. All samples were 8 Page 7 of 22 Journal of Veterinary Emergency and Critical Care 9 134 stored in airtight sealed plastic syringes placed in ice water and kept until the 135 parameters were evaluated by CO-oximetry. 136 137 Data analysis 138 The data were analysed using Sigmastat® 3.0.1e using repeated measures 139 analysis of variance (ANOVA) as assumptions of normal data distribution using a 140 Kolmogorov-Smirnov test with Lilliefor´s correction were met. Methaemoglobin was 141 logarithmically transformed to meet these criteria and converted back to original Fo 142 units. All data are presented as mean±SD except for MetHb% that is presented as rP 143 median and interquartile range 25-75 [IQR]. The reduction of MetHb over time for 144 the 3 different groups is expressed as the absolute reduction rate (%) from time 0 ee 145 at 8, 24 and 48h. Multiple comparisons within treatment and time were performed rR 146 by the Holm-Sidak method. Differences were considered statistically significant ev 147 when p≤0.05. 148 iew 149 RESULTS 150 Oxygenation status 151 There was a significant increase in O2Hb and SO2 over time from 69.8±10.2% and 152 90±3% to 82.8±7.9% and 99±3% respectively at 8h (mean±SD, p<0.001) in venous 153 blood samples stored at 0°C (ice water). There were no statistically significant 154 changes observed in arterial blood samples for any of the CO-oximetry 155 measurements (Table 1). 156 9 Journal of Veterinary Emergency and Critical Care Page 8 of 22 10 157 Antioxidant effects of ascorbic acid and methylene-blue on MetHb% 158 Methaemoglobinaemic samples from each group had variable percentages of 159 MetHb ranging from 4 to 73% at time 0 (Table 2). There was a significant effect of 160 treatment (p=0.032) and of time (interaction p=0.003) on MetHb%. There was a 161 significant increase in absolute MetHb% in samples from the control group at 24 162 and 48 hours (7.0% [IQR:-1.2, 20.0]) (p<0.001) (Figure 1A). There was no 163 significant effect of time on MetHb% in the AscAc group at any point in time (23% 164 [IQR:9.2, 61.8] at time 0 to 23.2% [IQR:7.0, 63.6] after 48h) (Figure 1B). There was Fo 165 an effect of time on absolute MetHb% in the MetBlue group at 48h (17.2% rP 166 [IQR:9.8, 32.8] at time 0h to 7.6% [IQR:3.5, 28.4] after 48h; p=0.003) (Fig 1C). 167 In samples with added AscAc and MetBlue, changes in MetHb% were as follows: - ee 168 0.25% [IQR:-2.0, 1.5] and -4.4% [IQR:-7.0, -1.8] at 48h for the AscAc and MetBlue rR 169 groups respectively (p<0.039) (Fig 2). ev 170 171 DISCUSSION iew 172 Our results revealed that O2Hb and SO2 increased significantly beyond 4 hours of 173 storage in airtight plastic syringes placed ice water. Addition of AscAc to 174 methaemoglobinaemic samples had a negligible antioxidant effect but was very 175 effective in preventing further oxidation of MetHb during the 48 hours study period. 176 Methylene blue resulted in a significant reduction on MetHb concentration over 177 time. 178 179 CO-oximetry is the gold standard for the determination of oxygenation status in 180 patients. However, CO-oximetry is not widely available hence the need to store 10 Page 9 of 22 Journal of Veterinary Emergency and Critical Care 11 181 blood for later determination at reference laboratories. Storage of blood samples in 182 ice water is a practical and standard way to keep samples refrigerated until arrival 183 to the laboratory. Previous work performed in healthy dogs has shown that CO- 184 oximetry measurements in arterial blood samples were stable for up to 4h in plastic 185 syringes placed in ice water12 and study measurements were performed at 2, 4 and 186 48h after sampling. The frequency of sampling in the present study was adapted to 187 determine if parameters would remain stable at 1, 2, 4, 8 and 24h. These collection 188 times were chosen to determine if testing could be delayed beyond 4h in horses. In Fo 189 addition to arterial blood, venous samples were also included in order to evaluate rP 190 whether samples with lower O2Hb and SO2 levels would be affected by storage 191 conditions. Air was removed from all samples and a rubber cap was placed at the ee 192 end of the needle to create an airtight environment. In the present study SO2 and rR 193 O2Hb values in arterial samples remained unchanged throughout the study period. ev 194 This may be explained by the high SO2 and O2Hb percentages found in normal 195 arterial blood samples, thus these parameters remained stable (Table 1). iew 196 In the present study, venous blood SO2 and O2Hb had a 9 and 10% absolute 197 increase respectively at 8h of storage in ice water (Table 1). The results of the 198 present study are in agreement with those obtained by Rezende in dogs. In their 199 study oxygenation status variables remained stable for up to 4h post-sampling.10 200 However, these results were only observed on samples with lower oxygen 201 saturation of haemoglobin (venous blood), while no changes were seen on 202 samples with very high oxygen saturation of haemoglobin (arterial samples). It is 203 likely that due to a higher oxygen saturation of haemoglobin, a modest increase in 204 partial pressures of oxygen within the sample did not affect the percentage of 11 Journal of Veterinary Emergency and Critical Care Page 10 of 22 12 205 oxyhaemoglobin. Increasing the sampling frequency confirmed that blood samples 206 stored under these conditions suffered spurious increases in SO2 and O2Hb after 207 4h of storage. 208 Previous studies evaluating the effects of time and conditions of storage performed 209 in horses and other animal species evaluated PaO2 and the calculated SO2 using 210 standard blood gas machines.2-5 Partial pressure of oxygen accounts for less than 211 3% of the oxygen content in blood, which can be a deficient reflection of the 212 remaining 97% total oxygen present in the blood. In case of Fo 213 dyshaemoglobinaemias PaO2 and PvO2 and a calculated SO2 as measured by rP 214 standard blood gas machines may remain within normal limits, while O2Hb and the 215 real SO2 could be in fact decreased. Previous studies have described the effect of ee 216 time and syringe type (glass versus plastic) on PaO2 and PaCO2 measurements rR 217 as well as the calculated SO2 in blood samples from horses. 2-5 In the present ev 218 study we used plastic syringes. The effects of time of storage on HbO2 and directly 219 measured SO2 have not been previously described in horses. iew 220 The in vitro experiments in the present study demonstrated a significant increase in 221 the concentration of MetHb over time in the control group. This is in agreement 222 with previous observations in humans in which MetHb oxidation continued over 223 time when whole blood methaemoglobinaemic samples were stored at room 224 temperature or at -30°C.15 This could be caused by a process known as 225 autoxidation of MetHb.16 It has been shown that oxidation of Hb into MetHb occurs 226 by different mechanisms, direct oxidation and autocatalytic oxidation.17 However, 227 only control samples with MetHb >20% in our study had an increase over time, 12 Page 11 of 22 Journal of Veterinary Emergency and Critical Care 13 228 whereas control samples with lower MetHb% would subjectively experience a 229 decrease over time. Perhaps, this effect was related to homeostatic mechanisms of 230 soluble and enzymatic antioxidant mechanisms that may be overwhelmed when a 231 given MetHb% is surpassed. 232 233 The MetBlue group showed a significantly decreased MetHb concentration when 234 compared to the control group at 48 hours. This is in contrast to a previous study in 235 which no effect of MetBlue on equine MetHb% was observed.11 In the study by Fo 236 Medeiros et al. CO-oximetry measurements were evaluated up to 24h, while in the rP 237 present study CO-oximetry measurements were evaluated for a longer period of 238 time (48h). For the present study whole blood was used instead of washed red ee 239 blood cells placed in a medium containing glucose and methylene blue as rR 240 described by Medeiros.11 Methylene blue is a pro-oxidant that needs to be first ev 241 reduced to leukomethylene blue in order to reduce methaemoglobin.18 It is likely 242 that some degree of oxidation due to methylene blue occurred at a similar rate than iew 243 red blood cell reduction causing concentrations to remain levelled throughout 244 Medeiros’ experiment. Although a decrease in MetHb was observed in this study, 245 the magnitude of the effect appeared to be less pronounced compared to the 246 reducing effect of MetBlue in cases of acquired methaemoglobinaemia in human 247 subjects.19 However, the effect was relevant since control samples had a large 248 increase in absolute percentages of MetHb while MetHb% in AscAc and MetBlue 249 treated samples either stabilized or decreased. 250 13 Journal of Veterinary Emergency and Critical Care Page 12 of 22 14 251 Ascorbic acid concentrations in this study were of similar molar concentration 252 described to be effective in reducing MetHb in methaemoglobinaemic samples in 253 an in vitro study in ruminants.20 In Atyabi’s experiment, blood was pretreated with 254 different vitamins at varying concentrations adding the AscAc before induction of 255 methaemoglobinaemia. Higher concentrations of AscAc resulted in a smaller 256 antioxidant effect when compared to a medium-low concentration.20 In this study 257 AscAc was added after incubation with sodium nitrate. 258 The main limitations of this study are that only one type of syringe and one method Fo 259 of cooling of the blood samples were used. Further studies are needed to evaluate rP 260 the effects of different storage temperatures and types of syringes on CO-oximetry 261 parameters in horses. Another limitation is that the temperature of the water may ee 262 have fluctuated depending on the volume of ice and water components, however, rR 263 this parameter was not recorded. As the antioxidant effects of AscAc and MetBlue ev 264 were evaluated in vitro, it is not known if these results may be extrapolated for use 265 in clinical cases. iew 266 In conclusion, whole blood can be stored in regular airtight plastic syringes for up 267 to 4 hours before measurement of HbO2 and SO2 by CO-oximetry. 268 Methaemoglobin may continue to increase after sampling in equine whole blood. 269 The addition of freshly prepared AscAc solution may prevent this spurious pre- 270 analytical increase in MetHb. Methylene blue reduces MetHb in vitro on equine 271 whole blood and it should be further investigated in horses using in vivo models. 272 Manufacturers’ addresses 273 a RNA medical, Devens, MA, USA 274 b ITC international Technydine Corporation, Edison, NJ, USA 14 Page 13 of 22 Journal of Veterinary Emergency and Critical Care 15 275 c Sol-Millenium Medical Inc., Lawrenceville, GA, USA 276 dJiangsu Kanghua Medical Equipment Co. Ltd., Jiansu, P.R. China 277 e SYSTAT Software Inc., San José, CA, USA 278 Fo rP ee rR ev iew 15 Journal of Veterinary Emergency and Critical Care Page 14 of 22 16 279 Table 1: Mean (±SD) percentage oxyhaemoglobin (O2Hb) and oxygen saturation 280 (SO2) in arterial blood samples at 0, 1, 2, 4, 8 and 24h and venous blood samples 281 at 0, 1, 2, 4 and 8h as well as, magnitude of change (∆) of O2Hb and SO2 from time 282 0h for both types of samples (calculated as median of the differences). Results are 283 expressed as mean [SD], Repeated measures ANOVA, n=7 (arterial group), n=6 284 (venous group). There was a significant effect of time at 8h * p<0.01, as compared 285 to 0h for the venous blood group. T0 T1 T2 T4 T8 T24 Fo Arterial O2Hb 99.6 (±0.8) 99.1 (±1.9) 99.3 (±1.5) 99.3 (±1.5) 99.7 (±0.3) 99.7 (±0.4) rP Arterial SO2 99.6 (±0.6) 99.2 (±1.8) 99.3 (±1.5) 99.4 (±1.2) 99.8 (±0.1) 99.8 (±0.2) Venous O2Hb 69.8(10.2) 73.5(8.6) 75.5(7.5) 78.3(7.5) 82.7(7.8)* --- ee Venous SO2 70.1(10.3) 73.4(8.9) 75.3(7.8) 78.1(7.6) 82.4(8.0)* --- rR ∆T0-T1 ∆T0-T2 ∆T0-T4 ∆T0-T8 ∆T0-T24 Arterial O2Hb --- -0.4 (±1.0) -0.3 (±0.7) -0.3 (±0.6) 0.2 (±0.5) 0.2 (±0.5) ev Arterial SO2 --- -0.5 (±1.1) -0.3 (±0.9) -0.2 (±0.6) 0.2 (±0.5) 0.2 (±0.4) iew Venous O2Hb --- 3.7(4.8) 5.7(4.3) 8.5(5.7) 12.9(7.5)* --- Venous SO2 --- 3.3(4.7) 5.3(4.2) 8.1(5.7) 12.3(7.5)* --- 286 287 16 Page 15 of 22 Journal of Veterinary Emergency and Critical Care 17 288 Table 2. Median [IQR] methaemoglobin percentage (MetHb%) in venous blood 289 samples of the control, ascorbic acid (AscAc), and methylene blue (MetBlue) 290 groups with variable in vitro induced methaemoglobinaemia at 0, 8, 24 and 48h, as 291 well as, magnitude of change (∆) of MetHb% from time 0h (calculated as median of 292 the differences). Results are expressed as median [IQR], Repeated measures 293 ANOVA, n=6 (control and AscAc groups) n=5 (MetBlue group). There was a 294 significant effect of time and treatment at 24h and 48h *p<0.01, when compared to 295 0h for the control group, and §p<0.01 when compared to 0h for the MetBlue group. Fo T0 T8 T24 T48 rP Control 21.9 [11.0, 29.2] 25.5 [11.8, 42.3] 27.6 [11.7, 46.8]* 28.9 [9.1, 52.4]* ee AscAc 23.1 [9.2, 61.8] 24.8 [10.3, 62.5] 24.1 [8.6, 62.2] 23.2 [7.0, 63.6] MetBlue 17.2 [9.8, 32.8] 16.5 [7.3, 35.1] rR 12.8 [3.7, 31.9]§ 7.6 [3.5, 28.4]§ ev ∆T0-T8 ∆T0-T24 ∆T0-T48 Control -- 3.3 [0.7, 10.0] 5.7 [0.7, 14.4]* 7.0 [-1.2, 20.0]* iew AscAc -- 0.7 [0.25, 1.2] -0.5 [-0.7, 0.5] -0.2 [-2.0, 1.5] MetBlue -- -0.9 [-1.4, -0.7] -2.7 [-4.4, -1.2]§ -4.4 [-7.0, -1.8]§ 296 AscAc, ascorbic acid; MetBlue, methylene blue. 297 298 17 Journal of Veterinary Emergency and Critical Care Page 16 of 22 18 299 Figure legends: 300 Figure 1. Median [IQR] (solid line [error bars]) and individual horses’ (dashed lines) 301 methaemoglobin % in venous blood samples with in vitro induced 302 methaemoglobinaemia from (A) the control group, (B) the ascorbic acid (AscAc) 303 group, and (C) the methylene blue (MetBlue) groups. The interquartile range is 304 depicted by the error bars arising from the solid lines (median methaemoglobin 305 concentration) at each time point. The methaemoglobin concentrations in (A) the 306 control group increased over time while the (B) AscAc group showed virtually no Fo 307 changes during the study period. The methaemoglobin concentration of (C) the rP 308 MetBlue group showed a decrease over time. The median (solid line) of figure C 309 obscures one individual horses’ methaemoglobin % as both (median and horse) ee 310 have the same values. Asterisks and § signs identify statistically significant changes rR 311 from baseline at specific times. Repeated measures ANOVA, n=6 (control and ev 312 AscAc groups) n=5 (MetBlue group), * 24h and 48h statistically different (p=0.001), iew 313 when compared tofrom 0h for the control group, § 24h and 48h statistically different 314 (p=0.003) when compared tofrom 0h for the MetBlue group. 315 316 Figure 2. Median [IQR] of change (∆) in methaemoglobin % from time 0h in venous 317 blood samples with in vitro induced methaemoglobinaemia of the control (dashed 318 line with circles), ascorbic acid (solid line with rectangles) and methylene blue 319 (dashed line and dots with triangles) groups. 320 18 Page 17 of 22 Journal of Veterinary Emergency and Critical Care 19 321 Supporting information 322 Figure 3. Individual sample’s magnitude of change (∆) in methaemoglobin % 323 (MetHb%) from time 0h, in venous blood samples with in vitro induced 324 methaemoglobinaemia of the control, ascorbic acid (AscAc) and methylene blue 325 (MetBlue) groups respectively. Fo rP ee rR ev iew 19 Journal of Veterinary Emergency and Critical Care Page 18 of 22 20 326 References 327 (1) Mack E. Focus on diagnosis: co-oximetry. Pediatr Rev 2007 Feb;28(2):73-74. 328 (2) Deane JC, Dagleish MP, Benamou AE, et al. Effects of syringe material and 329 temperature and duration of storage on the stability of equine arterial blood gas 330 variables. Vet Anaesth Analg 2004 Oct;31(4):250-257. 331 (3) Kennedy SA, Constable PD, Sen I, Couetil L. Effects of syringe type and 332 storage conditions on results of equine blood gas and acid-base analysis. Am J 333 Vet Res 2012 Jul;73(7):979-987. Fo 334 (4) Kules J, Mayer I, Rafaj RB, et al. Co-oximetry in clinically healthy dogs and rP 335 effects of time of post sampling on measurements. J Small Anim Pract 2011 336 Dec;52(12):628-631. ee 337 (5) Picandet V, Jeanneret S, Lavoie JP. Effects of syringe type and storage rR 338 temperature on results of blood gas analysis in arterial blood of horses. J Vet ev 339 Intern Med 2007 May-Jun;21(3):476-481. 340 (6) Subramanian G, Anitha VP, Ranjit S. Comparison of central venous saturation iew 341 by standard ABG machine versus co-oximeter: Is 18 carat as good as the 24 342 carat gold standard? Indian J Crit Care Med 2013 Mar;17(2):82-86. 343 (7) Fenger CK, McKeever KH, Hinchcliff KW, Kohn CW. Determinants of oxygen 344 delivery and hemoglobin saturation during incremental exercise in horses. Am 345 J Vet Res 2000 Oct;61(10):1325-1332. 346 (8) Hegesh E, Gruener N, Cohen S, et al. A sensitive micromethod for the 347 determination of methemoglobin in blood. Clin Chim Acta 1970 Dec;30(3):679- 348 682. 20 Page 19 of 22 Journal of Veterinary Emergency and Critical Care 21 349 (9) Smith JE, Beutler E. Methemoglobin formation and reduction in man and 350 various animal species. Am J Physiol 1966 Feb;210(2):347-350. 351 (10) Alward A, Corriher CA, Barton MH, et al. Red maple (Acer rubrum) leaf 352 toxicosis in horses: a retrospective study of 32 cases. J Vet Intern Med 2006 353 Sep-Oct;20(5):1197-1201. 354 (11) Medeiros LO, Nurmberger R,Jr, Medeiros LF. The special behavior of equine 355 erythrocytes connected with the methemoglobin regulation. Comp Biochem 356 Physiol B 1984;78(4):869-871. Fo 357 (12) Rezende M, Haskins S, Hopper K. The effects of ice‐water storage on blood rP 358 gas and acid–base measurements. Journal of veterinary emergency and 359 ee critical care 2007;17(1):67-71. 360 (13) Arifin WN, Zahiruddin WM. Sample Size Calculation in Animal Studies Using rR 361 Resource Equation Approach. Malays J Med Sci 2017 Oct;24(5):101-105. ev 362 (14) Shihana F, Dissanayake DM, Buckley NA, Dawson AH. A simple quantitative 363 bedside test to determine methemoglobin. Ann Emerg Med 2010 iew 364 Feb;55(2):184-189. 365 (15) Sato K, Tamaki K, Okajima H, Katsumata Y. Long-term storage of blood 366 samples as whole blood at extremely low temperatures for methemoglobin 367 determination. Forensic Sci Int 1988 Apr;37(2):99-104. 368 (16) Faivre B, Menu P, Labrude P, Vigneron C. Hemoglobin 369 autooxidation/oxidation mechanisms and methemoglobin prevention or 370 reduction processes in the bloodstream. Literature review and outline of 21 Journal of Veterinary Emergency and Critical Care Page 20 of 22 22 371 autooxidation reaction. Artif Cells Blood Substit Immobil Biotechnol 1998 372 Jan;26(1):17-26. 373 (17) Doyle MP, Herman JG, Dykstra RL. Autocatalytic oxidation of hemoglobin 374 induced by nitrite: activation and chemical inhibition. J Free Radic Biol Med 375 1985;1(2):145-153. 376 (18) Beutler E, Balauda MC. Methemoglobin Reduction. Studies of the Interaction 377 between Cell Populations and of the Role of Methylene Blue. Blood 1963 378 Sep;22:323-333. Fo 379 (19) Dotsch J, Demirakca S, Kratz M, et al. Comparison of methylene blue, rP 380 riboflavin, and N-acetylcysteine for the reduction of nitric oxide-induced 381 methemoglobinemia. Crit Care Med 2000 Apr;28(4):958-961. ee 382 (20) Atyabi N, Yasini SP, Jalali SM, Shaygan H. Antioxidant effect of different rR 383 vitamins on methemoglobin production: An in vitro study. Vet Res Forum 2012 ev 384 Spring;3(2):97-101. iew 22 Page 21 of 22 Journal of Veterinary Emergency and Critical Care Fo rP ee rR ev iew Figure 1. Median [IQR] (solid line) and individual horses’ (dashed lines) methaemoglobin % in venous blood samples with in vitro induced methaemoglobinaemia from (A) the control group, (B) the ascorbic acid (AscAc) group, and (C) the methylene blue (MetBlue) groups. The interquartile range is depicted by the error bars arising from the solid lines (median methaemoglobin concentration) at each time point. The methaemoglobin concentrations in (A) the control group increased over time while the (B) AscAc group showed virtually no changes during the study period. The methaemoglobin concentration of (C) the MetBlue group showed a decrease over time. The median (solid line) of figure C obscures one individual horses’ methaemoglobin % as both (median and horse) have the same values. Asterisks and § signs identify statistically significant changes from baseline at specific times. Repeated measures ANOVA, n=6 (control and AscAc groups) n=5 (MetBlue group), * 24h and 48h statistically different (p=0.001), when compared to 0h for the control group, § 24h and 48h statistically different p=0.003 when compared to 0h for the MetBlue group. 80x181mm (300 x 300 DPI) Journal of Veterinary Emergency and Critical Care Page 22 of 22 Fo rP ee rR ev iew Page 23 of 22 Journal of Veterinary Emergency and Critical Care Fo rP Figure 2. Median [IQR] of change (Δ) in methaemoglobin % from time 0h in venous blood samples with in vitro induced methaemoglobinaemia of the control (dashed line with circles), ascorbic acid (solid line with ee rectangles) and methylene blue (dashed line and dots with triangles) groups. 80x45mm (300 x 300 DPI) rR ev iew