Factors Affecting Dissatisfaction to Treatments in Patients With Chronic Constipation and IBS With Constipation

Introduction: Chronic constipation (CC) and Irritable Bowel Syndrome with Constipation (IBS-C) are two overlapping conditions in which the patients are dissatisfied of their bowel habit. In a non negligible proportion of patients, the dissatisfaction persists after a first line treatment with effective laxatives such as macrogol and bisacodyl. Prucalopride and linaclotide are two relatively new treatments that are used in these conditions. Prucalopride is a prokinetic drug licensed for patients with CC not responsive to standard laxative treatments. Linaclotide is a guanylate cyclase-C agonist that has been shown to improve both abdominal pain and bowel habits in patients with IBS-C. Despite these new treatments, a consistent number of patients remains dissatisfied suggesting that they complain of a "refractory" or "intractable" form of constipation.

Both IBS-C and CC are diseases defined by patients symptoms, as well as the lack of response to treatments. The Patient Assessment of Constipation-Symptoms (PAC-SYM) is a validated self-report questionnaire able to assess over time both general constipation severity and cluster of symptoms specific of this condition and to indicate the patient perspective on the disorder. Some limitations on the performances of this questionnaire have however emerged and it has recently been demonstrated that a Modified version of the PAC-SYM (M:PAC-SYM) is a better tool to assess constipation severity in patients recruited in tertiary referral centres. A change > 0.24 in the M:PAC-SYM score has been shown to define whether the response to treatment is adequate or not, but how this variation persists in the long term has not yet been investigated.

Several factors may influence the lack of response to treatments. Stool characteristics, psychological alterations (somatisation), delayed colonic transit and increased resting anal pressure may influence constipation severity, but the relative role of these factors is still unknown. On the other hand the bowel frequency, that is commonly believed to reflect constipation, does not seem to play a major role.

Aims: The purpose of this study will be to perform a pooled analysis of multiple prospective treatment optimization N-of-1 trials in patients with IBS-C and CC aimed at: 1) define how many patients achieve a sustained improvement; 2) define which factors are associated with a sustained improvement; 3) evaluate the suitability of M:PAC-SYM for treatment optimization among patients with CC or IBS-C 4) assess in a longitudinal follow up the natural history of IBS-C and CC submitted to standardised treatments.

Study design: All patients aged 18 - 80 years old with IBS-C and CC will be included in this observational multicentric national study. Patients with secondary causes of CC will be excluded. The diagnosis of CC and IBS-C will be defined according to Rome III criteria. All patients should undergo a thorough clinical examination including a rectal examination searching for both anatomical and functional defects. Dietary advices and supplementation with soluble fibres in those patients reporting insufficient dietary fibre intake will be given. Basic principles of defecation will be given to all patients and in particular to those with dyssynergic behaviour during straining.

The following items will be recorded at enrolment: 1) demographic data; 2) clinical history including pregnancy and anorectal surgery; 3) causes of secondary constipation (TSH and calcium levels, ultrasonography and colonoscopy when necessary); 4) rectal examination (anal pressure at rest and puborectalis relaxation during simulated evacuation); 5) ongoing therapy; 6) "satisfaction" with the current therapy (recorded with two items: 1-7 point visual analogue scale (VAS) and -7 +7 VAS ; 7) M:PAC-SYM; 8) Somatisation (Short Form-12 (SF-12) including the 12 questions of the Symptom Checklist-90 Revised (SCL-90-R) somatisation score); 9) quality of life (SF36, a validated questionnaire evaluating patients' perceptions of functional impairment into eight subscales that give rise to two composite scores assessing perceived physical and mental health); 10) colonic transit time (recorded according to Metcalf et al, off of laxatives and maintaining the usual diet. Patients will ingest 20 radio-opaque markers at 9 on each of three consecutive days and an abdominal X ray will be obtained at 9 of the fourth day. Colonic transit time will be calculated as the number of retained markers x 1.2; 11) anal resting pressure (the anal pressure (mmHg) will be recorded in resting conditions for ten minutes and it will be measured as the average of the last 30 seconds pressure recording on the channel with the highest pressure).

Enrolled patients will be directed to a one-month therapy with macrogol (17-34 mg per day) and if needed with bisacodyl (5-10 mg per day) ± bowel enema. Optimization of frequency of laxatives and dose titration will be encouraged. During the month, patients will fill a daily diary asking for bowel habits (stool frequency and consistency), abdominal pain, straining, incomplete evacuation and urgency. After one-month therapy (T1) patients will be admitted to the outpatient service for a follow up visit and "satisfaction" with the current therapy (recorded with two items: 1-7 point visual analogue scale (VAS) and -7 +7 VAS ) and M:PAC-SYM will be recorded.

Satisfied patients (satisfaction score>4) will be invited to continue the same therapy. Unsatisfied or partially satisfied patients (satisfaction score≤4) with a diagnosis of CC will be directed to one-month therapy with prucalopride 2 mg per day. Unsatisfied or partially satisfied patients (satisfaction score≤4) with a diagnosis of IBS-C will be directed to one-month therapy with linaclotide 290 mcg per day. When necessary, both patients with CC and IBS-C will be allowed to take bisacodyl (max 2 tablets a day, 5 mg each) and otilonium bromide (max 2 tablets a day, 40 mg each). As during the first month of therapy, patients will fill a daily diary asking for bowel habits (stool frequency and consistency according to 1-7 Bristol score ), abdominal pain (0-10) , straining, incomplete evacuation and urgency. After the one month therapy the patients will be evaluated as after the first follow-up visit (see T1). Patients will repeat for one month the previous treatment if satisfied or will be switched to the other treatment for one month, that will be repeated if satisfied.

Statistical analysis A pooled analysis will be performed on all patients achieving a sustained improvement or failing to achieve this improvement after all the three treatment options. Clinical and demographic information, along with self-reported outcomes will be abstracted from clinical charts and daily diary. The analytic dataset will include the following variables: Socio-demographics (age, gender, education, employment status); Clinical history (comorbidities, abdominal and extra-abdominal surgery, concurrent medications, previous pregnancy, life-style, dietary habits and ongoing therapy); Results of rectal examination; Psychological Profile (SF-12); Quality of life (SF36); Colonic transit time; Anal resting pressure; "satisfaction" with the current therapy (recorded with two items: 1-7 point visual analogue scale (VAS) and -7 +7 VAS ) and M:PAC-SYM; bowel habits (stool frequency and consistency), severity of abdominal pain, straining, incomplete evacuation and urgency/month.

Data will be reported as absolute and relative frequencies or means and standard deviations for discrete or continuous variables respectively. Agreement between responder status defined by treatment satisfaction scores or by M-PAC-SYM changes will be assessed with K coefficient for all assessment time-point. The investigators will calculate the variance partition coefficient from a random-intercept hierarchical logistic model (level 1: each study time-period; level 2: patient) in order to evaluate the proportion of variance explained by treatment-related vs patient-related factors in the likelihood of treatment response (change in satisfaction>2). Such model would allow to account for the auto-correlation of each observation within patients. Treatment modality will be accommodated with time-varying covariates indicating active/inactive treatments at each study periods. If the variance partition coefficient suggest that a significant share of variance in the likelihood of treatment response is explained by patient-related factors, socio-demographic, clinical, psychological and physiological patient-related factors will be included in the model. In order to assess the role of constipation-related symptoms in driving patients global satisfaction with treatment, the investigators will specify two additional models. In the first additional model, the M:ABD and the M:STO subscale scores of the M:PAC-SYM questionnaire will be included in the model of treatment response likelihood as time-varying covariate. In the second specification all M:PAC-SYM items will be included simultaneously as time-varying covariates.

Power calculations relative to key analysis planned for the study:

1. Impact of somatization and quality of life on treatment response likelihood (change in satisfaction>2): A sample of 82 patients completing at least 4 assessments would achieve 80% power of detecting a f=0,25 effect size difference in treatment response likelihood for patients diagnosed with somatization disorder, based on established SCL-90-R cut-off.
2. Impact of constipation-related symptoms on treatment response likelihood (satisfaction>2): A sample of 38 patients completing at least 4 assessments would achieve 80% power of detecting a f=0,25 effect size difference in treatment response likelihood for patients reporting each constipation-related symptoms (absent-mild symptoms vs moderate-severe PAC-SYM items) after Bonferroni correction for multiple testing (corrected alpha= 0.0045455).

According to these power calculations it is planned to reach the sample size of 90 patients achieving a sustained response or dissatisfied of the proposed treatments.