Objective To evaluate the responsiveness of the EULAR Sjögren's Syndrome Patient Reported Index (ESSPRI) and EULAR Sjögren's Syndrome Disease Activity Index (ESSDAI) in patients with primary Sjögren's syndrome (pSS) treated with rituximab.
Methods Twenty-eight patients with pSS treated with rituximab (1000 mg) infusions on days 1 and 15 were included in the study. Data were collected prospectively at baseline and 16, 24, 36, 48 and 60 weeks after treatment. Internal responsiveness was assessed using standardised response means (SRM) and effect sizes (ES). SRM and ES <0.5, 0.5–0.8 and >0.8 were interpreted as small, moderate and large, respectively. External responsiveness was assessed using Spearman correlation coefficients.
Results Median (range) ESSPRI and ESSDAI scores at baseline were 6.7 (0.3–9.0) and 8 (2–18), respectively. Both indices improved significantly after treatment. SRM and ES values for ESSPRI and ESSDAI were ≥0.8 at week 16 and decreased afterwards, and were larger for ESSDAI than for ESSPRI. SRM and ES values for patient's and physician's global disease activity (GDA) and rheumatoid factor broadly followed the pattern of those of ESSPRI and ESSDAI. SRM and ES for stimulated whole salivary flow were small at all time points. At baseline and for most change scores, moderate to good correlations were found between ESSPRI and patient's GDA and between ESSDAI and physician's GDA. Poor association was found between ESSPRI and ESSDAI.
Conclusions ESSPRI and ESSDAI are sensitive measures of change in disease activity after therapeutic intervention, which supports the usefulness of these indices for future clinical trials in patients with pSS. The responsiveness of ESSDAI was greater than that of ESSPRI.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Sjögren's syndrome (SS) is a systemic autoimmune disease primarily characterised by chronic inflammation of the exocrine glands.1 The salivary and lacrimal glands are most commonly affected, resulting in dry mouth and eyes. These sicca symptoms may be accompanied by extraglandular manifestations that may involve almost any organ such as the joints, lungs or skin (vasculitis). Moreover, almost all patients with SS have restricting chronic fatigue.
Since many patients with SS have reduced health-related quality of life and are restricted in their social and work-related activities, there is a clear need for development of adequate treatment modalities to reduce SS-related symptoms and to halt the progression of the disease.2,–,4 To date, rituximab, a chimeric murine/human anti-CD20 monoclonal antibody, is the only effective non-symptomatic treatment known for primary SS (pSS). Data from open-label trials and randomised clinical trials provide evidence that rituximab is effective in reducing glandular and systemic disease manifestations of pSS.3,–,12
The heterogeneous nature of pSS, as well as its variable course, has made it difficult to quantify the extent and severity of the disease in individual patients. Two indices have recently been introduced that might meet this shortcoming, a patient-administered questionnaire to assess patient symptoms (European League Against Rheumatism (EULAR) Sjögren's Syndrome Patient Reported Index, ESSPRI) and a systemic activity index to evaluate systemic complications (EULAR Sjögren's Syndrome Disease Activity Index, ESSDAI).13 ,14 The usefulness of instruments designed to measure change over time or after therapeutic intervention is dependent on their validity and reliability and also on their potential to detect clinically relevant changes.15 ,16 Seror et al17 retrospectively investigated the sensitivity to change of ESSDAI over time in 96 patient profiles abstracted from the medical charts of patients with pSS. These authors used the assessment of whether the patient's condition changed or remained stable as an external anchor. Interestingly, the responsiveness of ESSDAI was found to be large, and it seemed to detect change over time more accurately than other known indices.17 Prospective data on the responsiveness of ESSDAI after therapeutic intervention in patients with pSS are currently lacking. Furthermore, no data on the responsiveness of ESSPRI are yet available. The aim of this study was therefore to evaluate the responsiveness of ESSPRI and ESSDAI in patients with pSS who were treated with rituximab.
A prospective single-centre study was performed.
All 28 patients were aged 18 years and fulfilled the revised European-US criteria for pSS.18 Patients were enrolled as part of a long-term follow-up study of (re)treatment with rituximab and were treated with rituximab (1000 mg) infusions at days 1 and 15 as first (n=8), second (n=15), third (n=3) or fourth (n=2) course of rituximab treatment. (Re)treatment was started when B cell levels increased, rheumatoid factor (IgM-Rf) levels increased, salivary flow decreased, subjective symptoms (sicca, fatigue) increased and/or extraglandular manifestations reappeared. Patients were evaluated at baseline and 16, 24, 36, 48 and 60 weeks after rituximab treatment.
Disease activity assessments
Disease activity was assessed using ESSPRI, ESSDAI, patient's and physician's global disease activity (GDA; on a 10 cm visual analogue scale), IgM-Rf level and stimulated whole salivary flow (SWS). ESSPRI is a patient-administered questionnaire to assess patient symptoms, whereas ESSDAI is a physician-administered systemic activity index to evaluate systemic complications.14 ,17 A description of ESSPRI and ESSDAI is given in table S1 in the online supplement. ESSDAI and the physician's GDA were assessed by two experienced rheumatologists. The IgM-Rf level was measured by nephelometry. SWS was collected in a standardised manner at a fixed time of the day.19 ,20 Flow rates were calculated according to the methods described by Burlage et al21 and Kalk et al.22 ,23
Generalised estimating equations were used to analyse disease activity assessments over time within subjects. Simple contrasts were used to compare follow-up visits with baseline.
Internal responsiveness was defined as described by Husted et al,24 namely, the ability of a measure to change over a particular prespecified time frame. Internal responsiveness was assessed for all visits (compared with baseline) and for all disease activity assessments using standardised response mean (SRM) and effect size (ES). SRM was calculated as mean change in score between two visits divided by the SD of the change in score, whereas ES was calculated as mean change in score between two visits divided by the SD of the baseline score.24 SRM and ES <0.5 were interpreted as small, 0.5–0.8 as moderate and >0.8 as large.17 ,25 As indicated by Seror et al,17 the larger the SRM or ES for improved or worsened disease activity, the greater the responsiveness of the measure investigated. Furthermore, a SRM or an ES closer to zero when disease activity is unchanged indicates that the assessment of stability is more accurate.
The effect of rituximab is transient and treated patients usually experience relapse of pSS. This relapse parallels the return of B cells in the peripheral blood. Although the duration of the treatment effect differed between trials, it is usually seen from 12 weeks up to 24 or 36 weeks after treatment.12 We therefore expect large SRM and ES for ESSPRI and ESSDAI within this time frame and smaller values afterwards. SRM and ES were calculated between week 16 (supposed best effect of rituximab) and week 60 to evaluate the ability of flare detection of disease activity measures.
External responsiveness reflects the extent to which changes in a measure over a specified time frame relate to corresponding changes in a reference measure of health status.24 External responsiveness was assessed by relating scores of ESSPRI and ESSDAI to each other and to patient's GDA, physician's GDA, IgM-Rf level and SWS using Spearman correlation coefficients (ρ). Correlations <0.3 were interpreted as poor association, 0.3–0.6 as moderate association, 0.6–0.8 as good association and >0.8 as excellent association. It was hypothesised a priori that the correlations between ESSPRI and patient's GDA and between ESSDAI and physician's GDA would be moderate to good, indicating that they assess related but slightly different outcome constructs. Furthermore, it was hypothesised that IgM-Rf and SWS would have no more than moderate correlation with the ESSPRI and ESSDAI since these measures enfold only a small proportion of the clinical signs of pSS. Statistical analysis was performed with PASW Statistics 18 (SPSS, Chicago, Illinois, USA). p Values <0.05 were considered statistically significant.
Evaluation of change over time
ESSPRI revealed significant improvement (a reduction in patient symptoms) compared with baseline at 16, 24, 36 and 60 weeks after treatment with rituximab (table 2 and figure 1). ESSDAI showed significant improvement (a reduction in systemic complications) at weeks 16, 24 and 36. The results for each ESSDAI domain are given in table 3. All domains with some activity at baseline showed improvement, while most improvement was found in domains with the highest activity at baseline. For patient's and physician's GDA, significant improvements were found at all time points (table 2 and figure 1). The IgM-Rf level was significantly reduced at weeks 16, 24 and 36. SWS showed no improvements but also no decline in salivary flow at any time point.
Evaluation of internal responsiveness
As expected, SRM and ES for ESSPRI and ESSDAI were ≥0.8 at week 16 and decreased afterwards, with almost all scores being small at weeks 48 and 60 (table 4). SRM and ES were larger for ESSDAI than for ESSPRI, indicating better internal responsiveness for ESSDAI compared with ESSPRI. SRM and ES values for patient's and physician's GDA and IgM-Rf broadly followed the pattern of ESSPRI and ESSDAI values. For SWS, both SRM and ES were small at all time points.
Physician's GDA, ESSDAI and IgM-Rf level were able to detect flares; SRM and ES between week 16 and 60 were, respectively, 1.02 and 2.29 for physician's GDA, 0.82 and 1.67 for ESSDAI and 0.46 and 1.27 for IgM-Rf level. The other measures showed less ability to detect flares (SRM/ES: ESSPRI 0.69/0.49; patient's GDA: −0.21/−0.13; SWS: −0.21/−0.13).
Evaluation of external responsiveness
At baseline and for most change scores, significant moderate to good correlations were observed between ESSPRI and patient's GDA, and between ESSDAI and physician's GDA (table 5), as expected. On the contrary, no clear relation was found between ESSPRI and ESSDAI or between these indices and IgM-Rf and SWS.
To our knowledge, this is the first study to prospectively evaluate the responsiveness of ESSPRI and ESSDAI in patients with pSS after intervention therapy. Internal and external responsiveness of ESSPRI and ESSDAI were good, which supports the usefulness of these indices in clinical trials.
The responsiveness of the recently developed ESSPRI13 has not previously been reported, while retrospective validation of the ESSDAI has been performed in abstracted profiles of patients with pSS with systemic complications.17 The latter authors used the assessment of whether a patient's condition had changed or not as an external anchor to assess internal responsiveness. The most widely used method to assess internal responsiveness, which can be applied in prospective studies, is to evaluate the change in a measure after treatment that has been shown to be efficacious.24 However, studying responsiveness of disease activity indices in patients with pSS by evaluation after treatment is currently limited by the lack of standardised response criteria. Responsiveness can therefore only be evaluated at the group level and not at the individual patient level.
In our study, ESSPRI, ESSDAI (improvement in all domains with some activity at baseline; most improvement in domains with highest activity at baseline), patient's and physician's GDA and IgM-Rf improved significantly after treatment with rituximab. The duration of the beneficial effect differed for the various activity measures, but all measures showed improvements up to 36 weeks after treatment. These results are in line with the course of the subjective and objective improvements reported by Meijer et al.9 ,26 No improvement in SWS was observed, which might be due to the long disease duration (median >5 years) of the subjects included in the trial. A longer disease duration is associated with less salivary secretory potential,27 and the beneficial effect of rituximab treatment on salivary flow is only observed in patients with a reasonable residual salivary flow.5 Furthermore, there might still be a beneficial effect of rituximab on salivary gland function as no further decline in salivary flow was observed during follow-up, which commonly occurs in placebo-treated patients with pSS followed for a similar period of time.9
Responsiveness is an important measure for evaluating whether an instrument is able to detect changes as a result of treatment. Although the sample size in our study was relatively small for assessing responsiveness, data were available for each patient at six different time points and showed consistent results.
Internal responsiveness was large for ESSDAI at weeks 16, 24 and 36 and decreased at weeks 48 and 60. This reduction in responsiveness is expected because B cells are returning and the effect of rituximab is fading.9 Furthermore, ESSDAI was able to detect flares. SRM and ES values for ESSPRI broadly followed the pattern of ESSDAI, but ESSDAI detected changes in disease activity more accurately. In addition, ESSDAI is superior to ESSPRI in flare detection. This might be explained by the fact that ESSDAI is a composite measure of clinical signs and symptoms and laboratory results, while ESSPRI is composed of three questions on dryness, pain and fatigue. Nevertheless, since the sensitivity to change of ESSPRI is reasonable, dryness, pain and fatigue are apparently important distinguishing features of the disease.
Physician's GDA showed large internal responsiveness and was able to detect flares. One might argue that this makes the physician's GDA a better instrument to measure disease activity than ESSDAI, especially since it is less time-consuming to complete. However, Seror et al14 showed that physician's GDA has limited reliability because the influence of patient symptoms and signs on the physicians' evaluation of disease activity is extremely variable. Furthermore, in our study, physician's GDA was assessed by only two physicians, both of whom have extensive experience in this field. Finally, these physicians were not blinded for time points, which may also have influenced their assessment. Therefore, it is thought that ESSDAI results in a more objective, reliable and predictable outcome than physician's GDA in clinical trials.
The internal responsiveness of patient's GDA was also large, although flares were difficult to detect. The larger internal responsiveness observed for patient's GDA compared with ESSPRI might be related to the fact that patient's GDA measures more than patient's symptoms; for example, this measure is likely to be influenced by the effect of treatment on systemic signs.
As mentioned above, a limitation of studying external responsiveness of any disease activity index in patients with pSS is the lack of standardised response criteria. Therefore, it is not possible to use the receiver operating characteristic method to assess the ability of an index to reflect both change (in terms of sensitivity) and no change (in terms of specificity) on external response criteria.24 In the present study we therefore used correlation coefficients to assess external responsiveness. The correlations obtained for ESSPRI and ESSDAI were promising. As expected, moderate correlations between ESSPRI and patient's GDA and between ESSDAI and physician's GDA were found, indicating that these indices assess related but slightly different outcome constructs. A poor association was found between ESSPRI and ESSDAI. A possible explanation for this poor association is that, although both ESSPRI and ESSDAI improved after treatment, the weight given to a certain change by the physician and the patient is different. For example, a physician can rate an increase in salivary flow as highly contributory while the patient rates this increase in salivary flow only as satisfactory as his sensation of a dry mouth has only slightly improved. Another possible explanation for the poor association between ESSPRI and ESSDAI might be that ESSPRI and ESSDAI measure different domains of pSS. In our opinion, combining both indices would be preferable for use in clinical trials since both subjective (ESSPRI) and more objective (ESSDAI) measures are important for evaluating response to treatment.
In conclusion, this study shows that ESSPRI and ESSDAI are sensitive indices to measure change in disease activity after therapeutic intervention, which supports their usefulness for future clinical trials in patients with pSS. The responsiveness of ESSDAI was more prominent than that of ESSPRI.
The authors thank J Bulthuis-Kuiper for her assistance in collecting the data and Dr M Jalving for reading the manuscript and providing constructive criticism.
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Web Only Data - This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Ethics approval This study was conducted with the approval of the local ethics committee at the University Medical Center Groningen, The Netherlands (METc2008.179) and all patients provided written informed consent to participate in the study.
Funding This investigator-driven study was financially supported by Roche, Woerden, The Netherlands, which supplied the study medication. There was no involvement of the funding source in study design, patient recruitment, data collection, analysis and interpretation and writing of the report.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.