Results of the clinical trial

This chapter reports the results of the main RCT.

Recruitment

Recruitment took place over 39 months in 39 UK NHS hospitals from November 2015 to January 2019. The contribution from each site can been seen in Table 6.

Screening of participants commenced in November 2015, and the last participant was randomised in March 2019. The complete flow of participants through the GaPP2 trial is shown in the Consolidated Standards of Reporting Trials (CONSORT) diagram in Figure 1. Initially, 1348 individuals were approached for participation, of whom 414 were initially considered eligible based on clinical criteria. Of these women, 306 women were randomised, 76 did not return for randomisation for various reasons, 20 were ineligible because they did not return sufficient pain scores and 12 were found to be ineligible following collection of pain scores. A total of 153 participants were assigned to the gabapentin group and 153 to the placebo group. Ten women withdrew from the GaPP2 trial and one woman died. In addition, primary outcome data were unavailable for 51 and 49 women for the average and worst pain scores, respectively (across both groups) (see Figure 1). Reasons for trial withdrawal are provided in Table 7.

FIGURE 1 The CONSORT diagram of the flow of participants through the GaPP2 trial. a, Withdrawn consent for any further follow-up from the point of withdrawal. (Figure not reproduced here; see the original article.)

Pregnancy

Of the 306 randomised women, four became pregnant during follow-up in the GaPP2 trial. Of these women, two withdrew from treatment and any further follow-up once pregnancy was known, one withdrew only from treatment once pregnancy was known and one had withdrawn from treatment prior to knowing she was pregnant. All babies were delivered healthy, with no reported abnormalities.

Participant characteristics

The women had a mean age of 30 years and the majority were of white ethnicity. The minimised randomisation ensured balance between groups in terms of the proportion with a dysmenorrhoea pain score of ≥ 4 out of 10 (65% in both groups), current use of sex hormones (65%) and GHQ-12 questionnaire score (mean score of 4.7). The groups were also well balanced in all other baseline characteristics (Table 8).

Adherence to treatment

Of those participants with available adherence data, 101 out of 112 (90%) women in the gabapentin group were considered adherent, compared with 101 out of 109 (93%) women in the placebo group. A detailed breakdown of the extent of self-reported adherence to the study drug is shown in Table 9.

From the participants’ study drug diaries, the overall median number of capsules taken daily per week was calculated for each week by trial group and is presented in Figure 2. The median number of capsules taken during the dose-escalation phase was similar between the trial groups; thereafter, the gabapentin group generally took one capsule more than the placebo group throughout the treatment period. The median maximum-tolerated dose was 2100 mg (or placebo equivalent) for both groups at week 4; however, this reached 2700 mg (the maximum permitted dose) in later weeks for some women.

FIGURE 2 Median dose taken daily at each assessment week by trial group. Reproduced with permission from Horne et al.51 © 2020 The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC BY 4.0 license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: http://creativecommons.org/licenses/by/4.0/. (Figure not reproduced here; see the original article.)

Primary outcome

There were no significant between-group differences in both the worst and the average NRS pain scores. The mean worst NRS pain score was 7.1 (SD 2.6) in the gabapentin group and 7.4 (SD 2.2) in the placebo group (adjusted mean difference –0.20, 97.5% CI –0.81 to 0.42; p = 0.47). The mean average NRS pain score was 4.3 (SD 2.3) in the gabapentin group and 4.5 (SD 2.2) in the placebo group (adjusted mean difference –0.18, 97.5% CI –0.71 to 0.35; p = 0.45) (Table 10 and Figure 3).

FIGURE 3 Primary outcome plot of worst and average NRS scores. Reproduced with permission from Horne et al.51 © 2020 The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC BY 4.0 license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: http://creativecommons.org/licenses/by/4.0/. (Figure not reproduced here; see the original article.)

Sensitivity analyses

In the per-protocol analysis of the primary outcome comparison, including only the 101 women defined as adherent to taking the study drug in each trial group, the mean differences for worst and average pain changed only marginally. Similarly, when multiple imputation was used to estimate missing outcome data, the point estimate and CIs were almost identical to the intention-to-treat analysis of available data. Finally, when the interval between the end of screening and randomisation, when the study drug was commenced, was taken into account in the analysis model, there was no impact on the mean difference for either pain score. These sensitivity analyses are shown in Table 11.

A subgroup analysis was carried out for the three prespecified variables used in the minimisation algorithm, namely the presence of dysmenorrhoea, the baseline use of sex hormones and the GHQ-12 score defining depression or anxiety. There was no evidence of varying effect in the three prespecified subgroup analyses. The mean scores for worst and average pain in each subgroup are shown in Table 12.

Further analysis of pain scores, examined using a repeated-measures model, demonstrated a constant treatment effect across time points for both worst and average scores. The mean scores for worst and average pain at each time point are shown in Table 13.

Secondary outcome results

The proportion of women who reported a decrease in their pain score by at least 30% or 50% between baseline and 13–16 weeks post randomisation was calculated, and the proportions for each group compared. There were no differences in the proportion of participants achieving either percentage reduction (at least 30% or 50%) for both worst and average pain scores between the groups, as shown in Table 14.

The patient-reported outcomes were completed at the end of the treatment phase (weeks 16–17). No significant differences were noted in any patient-reported secondary outcomes. Summary scores and point estimates are provided in Table 15.

Women in the gabapentin group reported that they were taking fewer painkillers; however, these differences were marginal and not statistically examined (Table 16).

No differences were noted in the number of visits to health-care professionals for CPP between the gabapentin and the placebo groups (Table 17).

Adverse events

A higher proportion of women experienced a SAE in the gabapentin group (10/153, 7%) than in the placebo group (3/153, 2%) (p = 0.04) (Table 18). One participant, who was in the gabapentin group, died of a complication of pneumonia that was exacerbated by other comorbidities (type 2 diabetes and obesity), but this was not considered to be related to study participation. Dizziness and tiredness were the most frequently reported side effects of the study treatment, but a substantial proportion of women reported drowsiness and changes in mood and urinary patterns. Dizziness, drowsiness and visual disturbances were significantly more common in the gabapentin group than in the placebo group (see Table 18).