The authors report the CYP2D6 inhibitory effects of fluoxetine, paroxetine, sertraline, and venlafaxine in an open‐label, multiple‐dose, crossover design. Twelve CYP2D6 extensive metabolizers were phenotyped, using the dextromethorphan/dextrorphan (DM/DX) urinary ratio, before and after administration of fluoxetine 60 mg (loading dose strategy), paroxetine 20 mg, sertraline 100 mg, and venlafaxine 150 mg. Paroxetine, sertraline, and venlafaxine sequences were randomized with 2‐week washouts between treatments; fluoxetine was the last antidepressant (AD) administered. Comparing within groups, baseline DM/DX ratios (0.017) were significantly lower than DM/DX ratios after treatment (DM/DX AD ) with fluoxetine (0.313 , p < 0.0001) and paroxetine (0.601 , p < 0.0001) but not for sertraline (0.026 , p = 0.066) or venlafaxine (0.023 , p = 0.485). Between groups, DM/DX AD ratios were significantly higher for fluoxetine and paroxetine compared to sertraline and venlafaxine. No differences between DM/DX AD ratios were found for fluoxetine and paroxetine although more subjects phenocopied to PM status after receiving the latter (42% vs. 83%; χ 2 = 4.44, p = 0.049 , df = 1). Similarly, no differences between DM/DX AD ratios were found for sertraline and venlafaxine. Of note, the DM/DX AD for 1 subject was much lower after treatment with paroxetine (0.058) compared to fluoxetine (0.490), while another subject exhibited a much lower ratio after treatment with fluoxetine (0.095) compared to paroxetine (0.397). Significant correlations between AD plasma concentration and DM/DX AD were found for paroxetine ( r 2 = 0.404 , p = 0.026) and sertraline ( r 2 = 0.64 , p = 0.002) but not fluoxetine or venlafaxine. In addition, DM/DX AD correlated with baseline isoenzyme activity for paroxetine, sertraline, and venlafaxine groups. These results demonstrate the potent, but variable, CYP2D6 inhibition of fluoxetine and paroxetine compared to sertraline and venlafaxine. CYP2D6 inhibition may be related, in part, to dose, plasma concentration, and baseline isoenzyme activity, and these correlations merit further investigation .
