Withdrawal from paroxetine

Preliminary evidence that this issue is an important one for clinicians derives from post hoc analyses of the double-blind, placebo-controlled paroxetine withdrawal study (Geller et ah, 2001b), in which no strict exclusionary criteria were applied. In that study, 193/ 335 (58%) patients had at least one psychiatric disorder in addition to OCD and 102/335 (30%) had multiple ( 2) other disorders. The response rates in patients with comorbid ADHD, tic disorder, or oppositional defiant disorder (56%, 53%, and 39%, respectively) were significantly less than in patients with OCD only (75%, [ITT LOCF] p < 0.05). Psychiatric comorbidity was also associated with a greater... 

By the mid-1990s, there were also reports of severe withdrawal from Paxil and Zoloft. Debattista and Schatzberg (1995) reported on physical symptoms associated with a case of paroxetine withdrawal with vomiting, headache, and tremulousness, which they compared to a similar report concerning sertraline withdrawal (Louie et al., 1994). 

Extensive databases [see Table 24-1) have now shown that paroxetine and sertraline can reduce panic attacks to zero and prevent relapse. Paroxetine studies constitute the largest data set more than 700 patients have been treated for periods ranging from 10 to 36 weeks. In the placebo-controlled comparisons with clomipramine, paroxetine had an earlier onset of action and was better tolerated than clomipramine. Paroxetine was significantly better than placebo from week 4 onward, whereas no separation was seen between clomipramine and placebo until the end of the study. Fewer withdrawals occurred as a result of adverse events with paroxetine (7.3%) than with either clomipramine (14.9%) or placebo (11.4%). The minimum dose shown to be superior to placebo was 40 mg/day. 

Paroxetine at low concentration is dependent on CYP 2D6 for its clearance. However, this enzyme is almost completely saturated by paroxetine at low concentrations, which accounts for the nonlinear pharmacokinetics of paroxetine and why its half-life goes from 10 to 20 hours when the dose is advanced from 10 to 20 mg per day. At higher concentrations, paroxetine is most likely dependent on CYP 3A3/4 for its clearance. This dose-dependent change in the clearance of paroxetine probably accounts for the higher incidence of withdrawal reactions with this SSRI than might otherwise be expected for a drug with a half-life of 20 hours at steady-state on 20 mg per day (296, 297). 

This infant s symptoms were similar to those described in other neonates whose mothers took SSRIs shortly before delivery, although in this case a contributory effect from olanzapine was also possible. The authors made the point that it can be difficult to decide from the clinical presentation whether neonatal problems, such as those described above, represent SSRI withdrawal (5HT deficiency) or SSRI toxicity (5HT excess). The fact that no paroxetine was detected in the infant s plasma led them to conclude that SSRI withdrawal was responsible for the symptoms in this case. 

Of 93 cases of neonatal symptoms associated with the use of SSRIs in mothers around the time of delivery 64 were associated with paroxetine but reactions were also reported in infants whose mothers had taken citalopram, fluoxetine, and sertraline (87). It is unclear from these data whether paroxetine is actually most likely to provoke the neonatal syndrome, but in adults its use is associated with more severe withdrawal reactions than other SSRIs. It should also be noted that it is not clear whether the syndrome described in neonates is due to SSRI withdrawal or a form of serotonin toxicity. 

A 55-year-old woman who was taking terbinafine and paroxetine presented with fever, diarrhea, and vomiting. A bone marrow biopsy showed overall reduced ceUularity, and the aspirate showed a profound shift toward the production of immature myeloid cells, consistent with maturation arrest. Treatment consisted of withdrawal of all outpatient medications, hydration, intravenous fluids, broad-spectrum antibiotics, and G-CSF 5 pg/kg for 5 days. Mature granulocytes appeared in the peripheral blood on the fifth day in hospital, and she was discharged on the seventh hospital day with an absolute neutrophil count of 6.2 x 10 /1. Paroxetine was resumed weeks after discharge from hospital without hematological toxicity over 6 months. 

---