Description of studies

The literature search was conducted on 9 September 2015. There were 381 studies identified through database searching. After 111 duplicates were removed, the titles and abstracts of 270 reports were screened by two independent reviewers (KL and MM or CEP and JKM). Two hundred and thirty‐six reports were flagged as non‐applicable and 34 full‐text reports were assessed for eligibility (See Figure 1). Four reports were excluded (See: Characteristics of excluded studies), and 30 reports of 9 studies were included in the review as they met the pre‐defined inclusion criteria (Feagan 2013; Leiper 2011; Pena‐Rossi 2008; Probert 2003; Reinisch 2011; Rutgeerts 2005; Sandborn 2012; Sandborn 2014; Suzuki 2014). Four ongoing studies were identified (NCT00488631; NCT01551290; NCT01863771; NCT02039505).

Figure 1

---

---

Study flow diagram.

Three studies were excluded as they were not placebo controlled (Armuzzi 2005; Madsen 2001; Parikh 2013), and one study was excluded because it utilized the Work Productivity and Activity Impairment Questionnaire for ulcerative colitis (WPAI‐UC), which is not a validated HRQL instrument (Miner 2011).

Of the included studies, one trial investigated the efficacy of interferon‐ß‐1a (Pena‐Rossi 2008), one trial tested rituximab (an anti‐CD20 antibody) (Leiper 2011), and the remaining trials investigated tumor necrosis factor‐alpha (TNF‐α) antagonists (i.e. infliximab, adalimumab, and golimumab) and vedolizumab (a leukocyte trafficking inhibitor). Two studies investigated infliximab (Probert 2003; Rutgeerts 2005), three trials investigated adalimumab (Reinisch 2011; Sandborn 2012; Suzuki 2014), one trial studied golimumab (Sandborn 2014), and one study investigated vedolizumab (Feagan 2013).

Studies in Crohn's disease have shown that an increase in the IBDQ score of 16 to 32 points from baseline constitutes the lower and upper bounds of clinically meaningful improvement in HRQL (Feagan 2007). Based on these cut‐offs Rutgeerts 2005 defined an improvement in IBDQ as an increase of either > 16 or > 32 points from baseline. Sandborn 2012, Feagan 2013 and Suzuki 2014 defined an improvement in IBDQ as an increase of > 16 points from baseline. Pena‐Rossi 2008 defined an improvement in IBDQ as a > 15 point increase from baseline. Samsa 1999 determined that an increase of 3 to 5 points from baseline for the SF‐36 physical and mental component summary scores (PCS and MCS respectively) reflected a clinically meaningful improvement in HRQL. Rutgeerts 2005 defined an improvement in SF‐36 as an increase of either > 3 or > 5 points from baseline.

Risk of bias in included studies

The risk of bias assessment is summarized in Figure 2. The risk of bias was judged to be low in eight studies. Sequence generation was rated low risk for all studies. Eight of the nine trials utilized a centralized randomization technique and were rated as low risk of bias for allocation concealment. One study did not describe the method of allocation concealment and was rated as unclear risk of bias for this item (Probert 2003). Three studies were rated as having unclear risk of bias for blinding of participants and study personnel (Sandborn 2012; Sandborn 2014; Suzuki 2014). Eight of the nine included studies did not describe how outcome assessment was blinded and were therefore rated as having an unclear risk of bias. All of the included studies used adequate methods to deal with missing data except for Leiper 2011 which was rated as high risk for incomplete outcome data. Six out of sixteen patients in the rituximab group and two out of eight patients in the placebo group completed this twelve week study. All of the included studies were rated as low risk of bias for the selective reporting and other sources of bias domains.

Figure 2

---

---

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Effects of interventions

See: Summary of findings for the main comparison Interferon‐B‐1a versus placebo; Summary of findings 2 Rituximab versus placebo; Summary of findings 3 Infliximab versus placebo; Summary of findings 4 Adalimumab versus placebo; Summary of findings 5 Golimumab versus placebo; Summary of findings 6 Vedolizumab versus placebo; Summary of findings 7 TNF‐alpha antagonists versus placebo

Interferon‐ß‐1a versus placebo

Improved IBDQ at week 8

One study (N = 194) compared interferon‐ß‐1a to placebo (Pena‐Rossi 2008). There was no statistically significant difference in the proportion of patients who had improved IBDQ scores at week eight. Forty‐six percent (60/130) of interferon‐ß‐1a patients had improved IBDQ scores compared to 41% (26/64) of placebo patients (RR 1.14, 95% CI 0.80 to 1.61). A GRADE analysis indicated that the quality of evidence supporting this outcome was moderate due to sparse data (See summary of findings Table for the main comparison).

Rituximab versus placebo

IBDQ at week 12

One study (N = 24) compared rituximab to placebo in patients with active, steroid‐refractory UC (Leiper 2011). There was no statistically significant difference in mean IBDQ change scores at week 12. The mean (SD) improvement in IBDQ score was 17 (45) points for rituximab patients compared to 2 (29) points for placebo patients (MD 15.00, 95% CI ‐14.83 to 44.83). A GRADE analysis indicated that the quality of evidence supporting this outcome was very low due to very sparse data (24 participants) and high risk of bias (See summary of findings Table 2).

Infliximab versus placebo

IBDQ at week 6 or 8

Two studies (Probert 2003; Rutgeerts 2005), reported mean IBDQ scores at week 6 or 8 among patients who received a 5 mg/kg infusion of infliximab (n = 265) or placebo (n = 264). There was a statistically significant improvement in the mean IBDQ score among infliximab patients compared to placebo (MD 18.58, 95% CI 13.19 to 23.97). A GRADE analysis indicated that the quality of evidence supporting this outcome was high (See summary of findings Table 3). One study (Rutgeerts 2005) reported mean IBDQ scores at week 6 among patients who received 10 mg/kg infliximab (n = 242) or placebo (n = 244). There was a statistically significant improvement in the mean IBDQ score among infiximab patients compared to placebo (MD 15.00, 95% CI 9.46 to 20.54). A GRADE analysis indicated that the quality of evidence supporting this outcome was high (See summary of findings Table 3).

Improved IBDQ (> 16 points or > 32 points from baseline) at week 8

One study reported the proportion of patients who had improved IBDQ scores at week eight (Rutgeerts 2005). There was a statistically significant difference in the proportion of patients who had improved IBDQ scores at week eight. At week 8, 69% (333/484) of infliximab patients had an improvement in IBDQ score of at least 16 points from baseline compared to 50% (121/244) of placebo patients (RR 1.39, 95% CI 1.21 to 1.60). A GRADE analysis indicated that the quality of evidence supporting this outcome was high (See summary of findings Table 3). Fifty‐five per cent (265/484) of infliximab patients had an improvement in IBDQ score of at least 32 points from baseline compared to 33% (80/244) of placebo patients (RR 1.67, 95% CI 1.37 to 2.03). A GRADE analysis indicated that the quality of evidence supporting this outcome was moderate due to sparse data (See summary of findings Table 3).

EQ‐5D at week 6

One study (N = 43) reported mean EQ‐5D scores at 6 weeks (Probert 2003). There was no statistically significant difference in mean EQ‐5D scores at week 6. The mean (SD) improvement in EQ‐5D score was 7 (17) points for infliximab patients compared to 4 (16) points for placebo patients (MD 3.00, 95% CI ‐6.87 to 12.87). A GRADE analysis indicated that the quality of evidence supporting this outcome was low due to sparse data and wide confidence interval (See summary of findings Table 3).

Improved SF‐36 PCS (> 3 or > 5 points from baseline)

One study (N = 728) reported the proportion of patients who had an improved SF‐36 PCS at week eight (Rutgeerts 2005). There was a statistically significant difference in the proportion of patients who had an improved SF‐36 PCS at week eight. Fifty‐nine per cent (286/484) of infliximab patients had an improvement in SF‐36 PCS of at least 3 points from baseline compared to 41% (99/244) of placebo patients (RR 1.46, 95% CI 1.23 to 1.72). Forty‐nine per cent (237/484) of infliximab patients had an improvement in SF‐36 PCS of at least 5 points from baseline compared to 33% (80/244) of placebo patients (RR 1.51, 95% CI 1.23 to 1.85). A GRADE analysis indicated that the quality of evidence supporting this outcome was moderate due to sparse data (See summary of findings Table 3).

Improved SF‐36 MCS (> 3 or > 5 points from baseline)

One study (n = 728) reported the proportion of patients who had an improved SF‐36 MCS at week eight (Rutgeerts 2005). There was a statistically significant difference in the proportion of patients who had an improved SF‐36 MCS at week eight. Fifty per cent (242/484) of infliximab patients had an improvement in SF‐36 MCS of at least 3 points from baseline compared to 34% (83/244) of placebo patients (RR 1.47, 95% CI 1.21 to 1.79). Forty‐three per cent (208/484) of infliximab patients had an improvement in SF‐36 MCS of at least 5 points from baseline compared to 30% (73/244) of placebo patients (RR 1.44, 95% CI 1.16 to 1.79). A GRADE analysis indicated that the quality of evidence supporting this outcome was moderate due to sparse data (See summary of findings Table 3).

Adalimumab versus placebo

Three trials evaluated the change in HRQL with adalimumab administration. Reinisch 2011 investigated adalimumab over an eight week period as an induction agent for UC. Patients (N = 576) were randomized (1:1:1) to adalimumab 160/80 mg (160 mg at week 0, 80 mg at week 2 and 40 mg at weeks 5 and 6), adalimumab 80/40 mg (80 mg at week 0 and 40 mg at weeks 2, 4 and 6) or placebo. The authors found that patients in the 160/80 mg group had significantly improved IBDQ and SF‐36 PCS scores at week 8. Patients in the 80/40 mg group only showed an improvement in SF‐36 score at week 4. There was no improvement in the MCS dimension of the SF‐36 score across any of the groups. While the mean HRQOL scores were reported at baseline, week 4 and week 8 for all treatment groups this data could not be included in analyses because standard deviations were not reported.

IBDQ at week 8 or 52

Sandborn 2012 (n = 494) studied the effect of adalimumab for induction and maintenance treatment of UC. Patients received adalimumab 160 mg at week 0, 80 mg at week 2 and 40 mg every other week or placebo for 52 weeks. At week 8, there was a statistically significant improvement in the mean IBDQ score among adalimumab patients compared to placebo (MD 9.00, 95% CI 2.65 to 15.35). A GRADE analysis indicated that the quality of evidence supporting this outcome was moderate due to a wide confidence interval (See summary of findings Table 4). At week 52, there was a statistically significant improvement in the mean IBDQ score among adalimumab patients compared to placebo (MD 8.00, 95% CI 0.68 to 15.32). A GRADE analysis indicated that the quality of evidence supporting this outcome was moderate due to a wide confidence interval (See summary of findings Table 4).

Improved IBDQ (> 16 points from baseline) at week 8 or 52

Two studies (n = 768 patients) reported the proportion of patients who had improved IBDQ scores at week eight or 52 (Sandborn 2012; Suzuki 2014). There was a statistically significant difference in the proportion of patients who had improved IBDQ scores at week eight. At week 8, 53% (224/425) of adalimumab patients had an improvement in IBDQ score of at least 16 points from baseline compared to 44% (150/342) of placebo patients (RR 1.23, 95% CI 1.06 to 1.43). A GRADE analysis indicated that the quality of evidence supporting this outcome was moderate due to sparse data (See summary of findings Table 4). Although IBDQ scores tended to decrease over time there was a statistically significant difference in the proportion of patients who had improved IBDQ scores at week 52. Twenty‐six per cent (110/425) of adalimumab patients had an IBDQ score of at least 16 points greater than baseline compared to 15% (52/342) of placebo patients (RR 1.73, 95% CI 1.28 to 2.34). A GRADE analysis indicated that the quality of evidence supporting this outcome was moderate due to sparse data (See summary of findings Table 4).

Golimumab versus placebo

IBDQ at week 6

Sandborn 2014 reported on the effect of golimumab administration on IBDQ scores in patients receiving induction therapy. There was a statistically significant difference in the proportion of patients who had improved IBDQ scores at week 6 in patients who received 200mg/100 mg (MD 12.20, 95% CI 6.52, 17.88; 504 patients) and 400 mg/200 mg (MD 12.10, 95% CI 6.40 to 17.80; 508 patients) dosing regimens compared to placebo. GRADE analyses indicated that the quality of evidence supporting this outcome for both dose groups was high (See summary of findings Table 5).

Vedolizumab versus placebo

Vedolizumab was investigated as an induction (week 6) and maintenance agent (week 52) in a large multi‐center, randomized, double‐blind, placebo‐controlled trial (GEMINI1) that integrated two study cohorts and involved 1406 patients with moderate to severe UC (Feagan 2013). In the induction phase , 374 patients (cohort 1) were assigned to two intravenous doses of 300 mg of vedolizumab, at weeks 0 and 2 with an additional 521 patients (cohort 2) receiving open‐label vedolizumab at weeks 0 and 2. In the maintenance phase of the trial, patients from either cohort who responded to vedolizumab at week 6 (defined as a change in Mayo Clinic score of > 3 points and a decrease of at least 30% from baseline, with an decrease in the rectal bleeding subscore of >1 point or an absolute rectal bleeding subscore of 0 or 1) were randomly assigned to either continued vedolizumab therapy every 4 or 8 weeks or placebo for up to 52 weeks of treatment.

Improved IBDQ (> 16 points from baseline) at week 6 or 52

There was a statistically significant difference in the proportion of patients who had improved IBDQ scores at week six. Thirty‐seven per cent (83/225) of vedolizumab patients had an improvement in IBDQ score of at least 16 points from baseline compared to 23% (34/149) of placebo patients (RR 1.62, 95% CI 1.15 to 2.27). A GRADE analysis indicated that the quality of evidence supporting this outcome was moderate due to sparse data (See summary of findings Table 6). There was a statistically significant difference in the proportion of patients who had improved IBDQ scores at 52 weeks. Sixty‐four percent of patients receiving maintenance vedolizumab had an improvement in IBDQ score of at least 16 points from baseline compared to 38% (48/126) of placebo patients (RR 1.67, 95% CI 1.31 to 2.12). A GRADE analysis indicated that the quality of evidence supporting this outcome was moderate due to sparse data (See summary of findings Table 6).

SF‐36 PCS at week 6 or 52

Feagan 2013 reported mean SF‐36 PCS at week six and 52. At week 6, there was a statistically significant improvement in the mean SF‐36 PCS among vedolizumab patients compared to placebo (MD 2.60, 95% CI 1.22 to 3.98). A GRADE analysis indicated that the quality of evidence supporting this outcome was moderate due to sparse data (See summary of findings Table 6). At week 52 there was a statistically significant difference in mean SF‐36 PCS among patients receiving maintenance vedolizumab every four weeks (MD 2.80, 95% CI 0.96 to 4.64) and every eight weeks compared to placebo (MD 3.40, 95% CI 1.56 to 5.24). A GRADE analysis indicated that the quality of evidence supporting this outcome was moderate due to sparse data (See summary of findings Table 6).

Feagan 2013 reported mean SF‐36 MCS at week six and 52. At week 6, there was a statistically significant improvement in the mean SF‐36 MCS among vedolizumab patients compared to placebo (MD 4.60, 95% CI 2.69 to 6.51). A GRADE analysis indicated that the quality of evidence supporting this outcome was moderate due to sparse data (See summary of findings Table 6). At week 52 there was a statistically significant difference in mean SF‐36 MCS among patients receiving maintenance vedolizumab every four weeks (MD 4.80, 95% CI 2.33 to 7.27) and every eight weeks compared to placebo (MD 4.80, 95% CI 2.29 to 7.31). A GRADE analysis indicated that the quality of evidence supporting this outcome was moderate due to sparse data (See summary of findings Table 6).

TNF‐alpha antagonists versus placebo

IBDQ at week 6 or 8

The results of theTNF‐alpha antagonist trials were pooled to estimate the overall effect of this kind of therapy on HRQL in UC. The 10 mg/kg infliximab group and the 400 mg/200 mg golimumab group were omitted from this analysis since 5mg/kg of infliximab and 200 mg/100 mg of golimumab are more commonly used doses in clinical practice. Four studies (Probert 2003; Rutgeerts 2005; Sandborn 2012; Sandborn 2014), reported mean IBDQ scores at week 6 or 8 among patients who received TNF‐alpha antagonists (n = 784) or placebo (n = 781). The pooled analysis revealed a statistically significant improvement in the mean IBDQ scores favouring TNF‐alpha antagonist treatment (MD 13.71, 95% CI 10.40 to 17.01). A GRADE analysis indicated that the quality of evidence supporting this outcome was moderate due to unexplained heterogeneity (See summary of findings Table 7).

Improved IBDQ (> 16 points from baseline) at week 6 or 8

Three studies (n = 1495) reported the proportion of patients who had improved IBDQ scores at week six or eight (Rutgeerts 2005; Sandborn 2012; Suzuki 2014). There was a statistically significant difference in the proportion of patients who had improved IBDQ scores. Sixty‐two per cent (557/909) of patients who received TNF‐alpha antagonists had an improvement in IBDQ score of at least 16 points from baseline compared to 46% (271/586) of placebo patients (RR 1.32, 95% CI 1.19 to 1.46). A GRADE analysis indicated that the quality of evidence supporting this outcome was high (See summary of findings Table 7).