Search

Ketamine produces rapid antidepressant effects in treatment-resistant depression (TRD), but the magnitude of response varies considerably between individual patients. Brain-derived neurotrophic factor (BDNF) has been investigated as a biomarker of treatment response in depression and has been implicated in the mechanism of action of ketamine. We evaluated plasma BDNF and associations with symptoms in 22 patients with TRD enrolled in a randomized controlled trial of ketamine compared to an anaesthetic control (midazolam). Ketamine significantly increased plasma BDNF levels in responders compared to non-responders 240 min post-infusion, and Montgomery-Åsberg Depression Rating Scale (MADRS) scores were negatively correlated with BDNF (r=-0.701, p = 0.008). Plasma BDNF levels at 240 min post-infusion were highly negatively associated with MADRS scores at 240 min (r = -0.897, p=.002), 24 h (r = -0.791, p = 0.038), 48 h (r = -0.944, p = 0.001) and 72 h (r = -0.977, p = 0.010). No associations with BDNF were found for patients receiving midazolam. These data support plasma BDNF as a peripheral biomarker relevant to ketamine antidepressant response.

We evaluated a new portable instrument, the PPG StatPal II pH and Blood Gas Analysis System, designed for "point-of-care" measurements of blood gases and pH. Inaccuracy (% of target value) and imprecision (CV%) were assessed by blood tonometry and comparison with a Corning 178. Within-day results for PCO2 inaccuracy and imprecision ranged from 98.2% to 102.9% and 3.3% to 3.9%, respectively; for PO2, these were 95.5% to 102.3% and 2.3% to 3.0%, respectively. Between-day results for PCO2 inaccuracy and imprecision ranged from 99.2% to 99.3% and from 2.9% to 3.2%, respectively; for PO2, the ranges were 96.2% to 98.2% and 2.6% to 3.0%, respectively. Two PCO2 outliers (in 645 samples = 0.3%) were observed. In general, tonometry recovery, measurement stability, and pH bias results for the StatPal II and Corning 178 were comparable. We conclude that the StatPal II performs within acceptable ranges of inaccuracy and imprecision.

We measured total hemoglobin (CtHb) and carboxyhemoglobin (COHb) in 100 patients' blood samples by using five specialized spectrophotometers (CO-oximeters)--IL 482 CO-Oximeter, Corning 2500 CO-oximeter, Radiometer OSM 3 Hemoximeter, Corning 270 CO-oximeter, and the AVL 912 CO-Oxylite--and compared the results with those obtained with the manual cyanmethemoglobin method and a gas-chromatographic (GC) method, respectively. For the CtHb measurements, the differences between the cyanmethemoglobin method and the CO-oximeters were not clinically important for any model. For the blood COHb measurements, the direction of the bias relative to GC was dependent on COHb concentration. In general, the CO-oximeters underestimated COHb concentration for COHb > 2.5% of total hemoglobin but overestimated COHb concentration for COHb < or = 2.5%. We conclude that all five CO-oximeters compared favorably with the reference methods for CtHb and for high concentrations of COHb. However, the inaccuracy of CO-oximeters for low-concentration (< or = 2.5%) COHb measurements may make these instruments unsuitable for some applications.

We evaluated a new commercially available partially purified reduced bovine hemoglobin solution (RBHS) and a new tonometer for use in the quality control of blood gas analyzers. RBHS is manufactured in three different formulations, each corresponding to three different bicarbonate-buffering capacities and concentrations of total hemoglobin (ctHb). The P50 for each formulation of RBHS was determined to be as follows: 30.6 mmHg (4.08 kPa) for Level 1, 29.1 mmHg (3.88 kPa) for Level 2, and 28.2 mmHg (3.76 kPa) for Level 3. When RBHS and human blood samples were tonometered at three clinically significant values of PO2 and PCO2, the recovered values for PO2 and PCO2 in RBHS were comparable with those of the tonometered whole blood. Each level of RBHS also produced precise pH (SD < or = 0.006 pH units) and ctHb (SD = 1.0 g/L) values. In addition, when the temperature of a sample chamber was intentionally altered, the changes in RBHS blood gas values closely approximated the changes seen with human blood. RBHS shows more thermal sensitivity than either perfluorocarbon emulsions or aqueous buffers, which are currently being used as quality-control and proficiency testing materials.

Tonometered whole-blood and plasma specimens were tested in plastic and glass syringes to determine whether clinically significant changes in gas tensions occur during sample storage. When whole blood was tonometered with 60 and 100 mL/L (6% and 10%) oxygen and then stored for 30 min in iced plastic syringes, the pO2 of the samples remained stable (mean change = +0.4 and +0.8 mmHg, respectively). However, for 140 mL/L (14%) oxygen tonometry, the pO2 increased significantly (mean change = +8.4 mmHg; P less than 0.0001). When tonometered plasma was stored in iced plastic syringes, the pO2 increased progressively at all three concentrations, with the smallest change occurring at 140 mL/L (mean change = +12.6 mmHg) and the greatest at 60 mL/L oxygen (mean change = +20.9 mmHg). In contrast, when iced glass syringes were used for storing plasma or whole blood, no clinically significant changes in pO2 were found at any of the tonometered oxygen values for 60 min. When whole blood was stored in plastic syringes at ambient temperature for 30 min, again no clinically significant changes in pO2 were found at these tonometry conditions. Apparently, some blood gas samples stored in iced plastic syringes may yield clinically significant errors in oxygen tension.

We measured the concentration of carboxyhemoglobin (HbCO) in original and CO-tonometered blood samples from 53 intensive-care patients with IL 282 and IL 482 CO-Oximeters and by gas chromatography (GC), finding very strong correlations among the three methods for HbCO concentrations greater than 2.5%. For concentrations within the normal reference interval (less than or equal to 2.5%), however, the correlation between CO-Oximetry and GC is poor (r2 less than 0.26). The capillary mode of the IL 482 has a consistently lower correlation with any other method or mode. The correlations of measurements between the CO-Oximeters for total hemoglobin and oxyhemoglobin were excellent (r2 greater than or equal to 0.98). Correlations for methemoglobin were lower, owing to its low concentrations in the samples. We conclude that the IL 482 and the IL 282 are analytically equivalent for all analytes measured, but that both instruments differentiate poorly between HbCO values that fall within the reference interval.