Effect of Amitriptyline on Symptoms in Treatment Naïve Patients With Interstitial Cystitis/Painful Bladder Syndrome

Results 

Subjects were recruited into the study from February 2005 to October 2007. Of the 319 subjects who agreed to participate in the study 271 were randomized and 231 (85%) completed 12 weeks of followup (fig. 2). A total of 23 subjects (17%) withdrew from the amitriptyline group and 17 (13%) withdrew from the placebo group (p = 0.26). Study participants were predominantly female (83%) and white (74%), with a median age of 38 years (table 1). Overall baseline symptoms were moderate to severe, and treatment groups were comparable across all baseline measures evaluated.

View Large Image Download to PowerPoint

Figure 2. Consort diagram for flow of subjects through study phases by treatment arm.

Table 1. Baseline characteristics

		Amitriptyline	Placebo
	No. subjects randomized⁎	135	136
	No. female (%)	115(85)	111(82)
	Mean ± SD age	38.0±13.8	39.9±14.0
	No. race (%):
	White/Caucasian	96(71)	104(77)
	Black/African-American	18(13)	15(11)
	Multiracial, Asian, other	21(16)	17(13)
	No. ever diagnosed with IC/PBS (%)	62(46)	75(55)
	Mean ± SD yrs since diagnosis (if applicable)	1.4±4.0	1.2±3.0
	Mean ± SD yrs since onset of initial symptoms	6.3±9.0	6.5±9.9
	Mean ± SD pain score (0–10)	5.8±1.5	6.0±1.8
	Mean ± SD urgency score (0–10)	6.4±1.8	6.4±1.8
	Mean ± SD frequency score (0–10)	6.8±1.7	6.8±1.8
	Mean ± SD 24-hr voiding frequency	13.9±5.2	13.7±5.7
	Mean ± SD nighttime voiding frequency	2.4±1.8	3.0±2.8
	Mean ± SD IC Symptom Index	12.2±3.1	12.0±3.6
	Mean ± SD IC Problem Index	11.1±2.8	11.2±3.0
	Mean ± SD Wisconsin Symptom Survey	25.9±8.6	26.7±8.3

For the symptom scores shown the measurements at the 2 baseline visits were averaged to give the overall baseline score.

⁎ Approximately 9 subjects in each arm were missing data on voiding frequency and 8 in each arm were missing data on years since symptoms began. At most 1 subject was missing data for all other measures.

For the primary outcome there was no statistically significant difference between treatment groups. Of the subjects assigned to amitriptyline plus EBMP 55% (74 of 135) reported they were markedly or moderately improved vs 45% (61 of 136) of those assigned to placebo plus EBMP (difference of 10% [−2, 22], p = 0.12). Consistent with intent to treat analytic strategies participants who did not provide data at 12 weeks, including 24 (18%) on amitriptyline and 17 (13%) on placebo, were considered treatment nonresponders.

Table 2 summarizes changes in selected symptom outcomes from baseline to 12 weeks by treatment arm for those with available data at those points, and provides confidence intervals for these differences. Results for these secondary efficacy end points were qualitatively similar with all outcomes demonstrating greater improvement for amitriptyline than placebo. Additional secondary outcomes related to quality of life and sexual function demonstrated no differences between the 2 arms (data not shown).

Table 2. Selected secondary symptom outcomes

		Mean ± SD Amitriptyline	Mean ± SD Placebo	Difference (95% CI)
	Pain score	−2.6±2.5	−2.3±2.4	−0.4 (−1.0,0.3)
	Urgency	−3.0±2.5	−2.5±2.6	−0.5 (−1.2,0.1)
	Frequency	−3.5±2.3	−2.6±2.5	−0.9 (−1.5,−0.3)
	24-Hr voiding frequency	−4.4±5.4	−2.0±3.9	−2.4 (−3.7,−1.0)
	Nighttime voiding frequency	−0.9±2.4	−0.9±2.4	−0.01 (−0.7,0.7)
	IC Symptom Index	−4.8±3.7	−3.3±3.9	−1.5 (−2.5,−0.5)
	IC Problem Index	−5.2±3.9	−3.9±4.0	−1.2 (−2.3,−0.2)
	University of Wisconsin Symptom Survey	−13.1±10.2	−9.3±9.3	−3.8 (−6.3,−1.3)

The analysis of secondary end points was not per intent to treat due to withdrawals and missing data. The sample size for the voiding frequency measures was approximately 92 for amitriptyline and 95 for placebo. At most 2 subjects were missing data for all other measures.

Adherence to protocol treatment, to study drug and EBMP, which all subjects received, was also evaluated. For the 241 subjects evaluable for EBMP adherence, the rate at 6 weeks was 75% (181 of 241) for symptom management, 83% (201 of 241) for fluid management, 82% (196 of 240) for diet modification and 71% (172 of 241) for bladder training. For the drug therapies 64 (47%) subjects attained and remained on the highest dose in the amitriptyline arm vs 105 (77%) in the placebo arm (p <0.001). Subjects were classified into 3 groups based on the maximum dose obtained at 6 and 12 weeks (table 3). Only 46% of subjects on amitriptyline achieved a dose of at least 50 mg per day and maintained it throughout the 12 weeks of followup, whereas 72% of subjects on placebo maintained this equivalent dose. Furthermore, a number of subjects in both treatment arms (20% for amitriptyline, 15% for placebo) who achieved a dose of at least 50 mg per day by 6 weeks decreased their dose during the second 6 weeks of followup.

Table 3. Response rates at 12 weeks by measures of adherence to therapy

		Amitriptyline	Placebo
	No. subjects evaluable for adherence at 12 wks (%)	112 (83)	119 (88)
	No. highest dose at 6 + 12 wks (%):
	50 mg or Greater achieved + maintained at 12 wks	62 (46)	98 (72)
	50 mg or Greater achieved but reduced to less than 50 mg by 12 wks	27 (20)	20 (15)
	50 mg or Greater not achieved	46 (34)	18 (13)
	No./total No. response rates at 12 wks (%):
	50 mg or Greater achieved + maintained at 12 wks	48/62 (77)	52/98 (53)
	50 mg or Greater achieved but reduced to less than 50 mg by 12 wks	11/27 (41)	3/20 (15)
	50 mg or Greater not achieved	15/46 (33)	6/18 (33)

We also evaluated the GRA response with respect to achieved drug dose, although we recognize that there may be numerous biases in such an analysis. The overall GRA response rate was approximately 57% for adherers in any of the 4 EBMP categories, whereas it was lower for nonadherers. This rate neared statistical significance for diet modification, which demonstrated only a 41% response rate among nonadherers (p = 0.051). In the subgroup of subjects (160) who were able to achieve and maintain a daily study drug dose of at least 50 mg, the responder rate was significantly greater in the amitriptyline group (77%) compared to placebo (53%, p <0.001).

Overall 80% (217 of 271) of participants reported at least 1 AE, classified primarily as mild (32%, 87 of 271) or moderate (42%, 114 of 271). The overall AE rate was 88% in the amitriptyline arm and 72% in the placebo arm (p = 0.0013, table 4). Within systems constitutional symptoms such as fatigue (p = 0.018), gastrointestinal problems (p = 0.0012), neurological symptoms including dizziness (p = 0.030) and renal/genitourinary problems (p = 0.025) were significantly more common in the amitriptyline arm. Five serious AEs were reported and none was classified as related to the study drug.

Table 4. Significant, cumulative AEs by body system and treatment group

		No. Amitriptyline (%)	No. Placebo (%)
	Highest AE grade	119 (88)	98 (72)
	Constitutional symptoms (primarily fatigue, malaise)	61 (45)	42 (31)
	Dermatology/skin	9 (7)	10 (7)
	Gastrointestinal (primarily dry mouth, constipation)	57 (42)	32 (24)
	Infection, fever	9 (7)	14 (10)
	Musculoskeletal	7 (5)	3 (2)
	Neurological (primarily dizziness, somnolence)	45 (33)	29 (21)
	Ocular, visual	3 (2)	7 (5)
	Pain (primarily headache)	43 (32)	48 (35)
	Pulmonary	15 (11)	13 (10)
	Renal/genitourinary	22 (16)	10 (7)

Discussion 

This is the first large scale, multicenter, randomized clinical trial of amitriptyline in IC/PBS to our knowledge. The primary intent to treat analysis of overall response rates demonstrated no statistically significant difference between amitriptyline (55%) and placebo (45%) for overall response rates in relatively treatment naïve patients with an aggressive EBMP. Although several secondary outcomes suggested a beneficial effect of amitriptyline, these findings are possibly subject to bias due to study withdrawals. Participants who were able to maintain a dose of at least 50 mg per day of amitriptyline reported a significantly higher response rate (77%) compared to an equivalent dose of placebo (53%). However, adherence to a higher dose on either treatment was also associated with higher response rates, as was adherence to the EBMP, and these do not represent randomized comparisons. Although a higher rate of AEs was experienced by subjects assigned to amitriptyline, most events were mild or moderate.

Ours is the first large scale clinical trial focused on treatment naïve subjects. Most prior trials in IC/PBS have evaluated treatment refractory cases, and it is possible that these cases have a more severe form of the disease. The widespread use of amitriptyline as a first line therapy in IC/PBS emphasizes the importance of its assessment in a group of treatment naïve subjects.

Another unique aspect of this study was the EBMP received by both treatment groups early in followup. While the design of the study did not permit assessment of the effect of this program compared to a control, it was implemented because investigators strongly believed that at least minimum care should be provided to participants of a trial involving a placebo, especially for newly diagnosed subjects. It is important to note that the observed overall response rate of 45% for subjects randomized to placebo is higher than the majority of previously published studies on IC/PBS. This may represent not only a possible placebo effect in the treatment naïve patient population, but also possible benefits of the EBMP.

Although amitriptyline is commonly used to treat patients with IC/PBS, the evidence on which this practice is based is limited. It has been 2 decades since the first report of treating patients with IC with this drug.3 In 25 patients with IC in whom hydrodistention and intravesical dimethylsulfoxide had failed a daily dose of amitriptyline was administered beginning with 25 mg and increasing to 75 mg during a 3-week period. Of the 20 patients who tolerated this dose 8 reported complete resolution of symptoms. Improvements in pain, daytime frequency, urgency and dyspareunia were also noted.

Despite these encouraging results it was 15 years later that the first randomized, placebo controlled clinical trial was reported.2 A total of 50 patients (88% women) who met the National Institute of Diabetes and Digestive and Kidney Diseases symptom criteria for IC were randomized to amitriptyline (up to 100 mg nightly) or placebo and followed for 4 months. Reduction in the O'Leary-Sant Symptom and Problem Index score was significantly greater in the amitriptyline group, with dry mouth being the most frequent side effect. Although the side effects in our study were primarily mild to moderate, less than 50% of subjects randomized to amitriptyline were able to achieve and maintain a dose of at least 50 mg per day throughout the treatment period. Despite our inability to show a beneficial effect of the drug in our primary intent to treat analysis, we did observe significant improvement in the GRA response rate in subjects who maintained higher doses. Whether this observation accurately reflects the true efficacy of this drug or is a result of bias cannot be determined.

Conclusions 

Amitriptyline compared to placebo, both provided in combination with a program of EBMP, did not overall demonstrate statistically significantly improved symptoms as determined by the GRA in a group of predominantly female and relatively treatment naïve patients with IC/PBS. The side effect profile of amitriptyline was acceptable, although adherence to higher doses was fairly low. Nevertheless, in a subgroup of subjects who achieved a dose of at least 50 mg per day, there was a suggestion that the drug was more effective than placebo.