The shape of human atrial action potential accounts for different frequency-related changes in vitro

We aimed at investigating frequency-related changes of human atrial action potential (AP) in vitro to see whether baseline AP shape might account for different responses to increasing stimulation rates. Human right atrial trabeculae (n = 48) obtained from adult (n = 38, mean age 59 +/- 8, range 45-72 years) consecutive patients (approximately equal to 30% of those operated upon by a single surgeon; 1.26 preparations per patient, range 1-2) were superfused in an organ bath with oxygenated (O2 content 16 ml/l) and modified (NaHCO3 25.7 mmol/l) Tyrode's solution at 31 degrees C. Baseline electrophysiology (pacing: 1 ms duration, 2-4 mA current intensity) at cycle length (CL) of 1000 ms was recorded in 90% (43 out of 48) of the preparations. The frequency-related protocol (CL from 1600 to 300 ms) was, however, undertaken in 23 (48%) preparations because 20 (42%) became pacing unresponsive immediately after baseline recordings. No statistical differences were seen when baseline electrophysiological parameters (mean +/- SD) were grouped according to late pacing responsiveness (n = 43 vs. n = 23): respectively, resting membrane potential (RMP) was -74 +/- 6 vs. -75 +/- 4 mV, maximal upstroke velocity (Vmax) 172 +/- 60 vs. 173 +/- 39 V/s, AP amplitude (APA) 89 +/- 11 vs. 91 +/- 8 mV and AP durations were at 30% (APD30%) 10 +/- 13 vs. 13 +/- 18 ms, 50% (APD50%) 45 +/- 79 vs. 62 +/- 91 ms and 90% (APD90%) 383 +/- 103 vs. 407 +/- 108 ms. To classify baseline AP shape, two criteria were adopted: criterion 1 ("objective"), based on APA (cut-off 90 mV) and APD90% (cut-off 500 ms) computed values and criterion 2 ("visual") derived from the literature. These criteria enabled us to differentiate three AP shape types: type 1 (spike and dome), type 3 (no dome) and type 4 (extremely prolonged). At baseline, the two criteria diagnosed different proportions of AP shape types. There were, however, no intra-type statistical differences among electrophysiological parameters. By criterion 1, analysis of variance (ANOVA) showed significant inter-type differences of RMP,Vmax, APA, APD50 and 90% and by criterion 2 of APA, APD30, 50 and 90%, respectively. To facilitate comparisons with previous published data, criterion 2 was selected to analyse frequency-related changes of AP shape types. At low stimulation rate, ANOVA for repeated measures (with Greenhouse-Geisser epsilon correction) showed inter-type differences for APD30, 50 and 90% (P = 0.00005). RMP, Vmax, APA and APD90% were overall frequency-related (P = 0.00005). Inter-type frequency-related differences were however seen only for APD90%. Human atrial AP durations (30, 50 and 90%) enable differentiation among AP shape types (1, 3 and 4). By a standardized use-dependent protocol overall RMP, Vmax, APA and APD90% are frequency-related. AP shape accounts for frequency-related changes of APD90% only. A type 4 AP shape with much prolonged AP duration had a flat frequency dependence. At high stimulation rates, adult type 1 and 3 AP shapes are indistinguishable. Use-dependent and pharmacological investigations in human atrial myocytes need to take AP shape into account.