User login
CHICAGO – At least three phase 3 randomized scleroderma treatment trials presented at the annual meeting of the American College of Rheumatology failed to meet modified Rodnan Skin Score–based primary endpoints, but a different story emerged when data from the trials were analyzed using the novel ACR Composite Response Index in Systemic Sclerosis (CRISS).
The differences highlight the limitations of individual outcome measures like the modified Rodnan Skin Score (mRSS) and underscore the need for new measures that capture the complexity of systemic sclerosis (SSc) and are more sensitive to changes in disease severity, according to Robert Spiera, MD, director of the vasculitis and scleroderma program at the Hospital for Special Surgery in New York.
Such measures are needed to better assess the effects of treatment interventions in scleroderma trials, Dr. Spiera said in an interview.
The ACR CRISS
Development of the ACR CRISS was led by Dinesh Khanna, MBBS, a professor of medicine and director of the scleroderma program at the University of Michigan, Ann Arbor. He and his colleagues described the measure in 2016 (Arthritis Rheumatol. 2016 Feb;68[2]:299-311. doi: 10.1002/art.39501).
Its use involves a two-step process of identifying any significant disease worsening or new end-organ damage, and then calculating the probability of patient improvement after 1 year of treatment on a 0- to 1-point scale based on changes from baseline in five variables: the mRSS, percent predicted forced vital capacity (FVC), patient and physician global assessments, and the Health Assessment Questionnaire Disability Index (HAQ-DI).
A CRISS score of 0.6 or higher indicates likelihood that a patient improved on treatment. Of note, subjects with significant worsening of renal or cardiopulmonary involvement are classified as not improved (score of 0), regardless of improvements in other core items.
To devise the CRISS, the investigators compiled 150 patient profiles with standardized clinical outcome elements using patients with diffuse cutaneous systemic sclerosis (dcSSc). The profiles were assessed by 40 scleroderma experts who rated patient improvement or lack thereof over 12 months.
Using the 79% of profiles for which a consensus was reached, the investigators “fit logistic regression models in which the binary outcome referred to whether the patient was improved or not, and the changes in the core set items from baseline to follow-up were entered as covariates,” they explained.
This led to the selection of the five measures included in the final version, which was found to have sensitivity of 0.982 and specificity of 0.931. When evaluated in a previously completed 1-year randomized controlled trial, the index differentiated the effect of methotrexate from the effect of placebo (P = .02), they reported.
Based on these findings, the ACR board of directors granted “provisional” endorsement of the CRISS for use in SSc clinical studies, signifying that it had been quantitatively validated using patient data, but had not undergone validation using an external data set.
New data presented at the 2018 ACR meeting will likely lead to full approval of the measure once the studies validating the measure are published, according to Dr. Spiera.
The focuSSced study of tocilizumab
For example, in the phase 3 focuSSced study comparing the interleukin-6 (IL-6) receptor–alpha antibody tocilizumab (Actemra) with placebo in patients with SSc, the primary endpoint of mean change from baseline in mRSS was not met (–6.14 vs. –4.41 points with tocilizumab vs. placebo, respectively; P = .098), Dr. Khanna said during a report of findings from the double-blind portion of the study.
However, when the CRISS was applied and its performance prospectively assessed as an exploratory outcome at week 48 in the focuSSced trial, the differences between the groups were statistically significant. Dr. Khanna reported those findings in a separate presentation at the meeting, noting that they marked the first prospective evaluation of the ACR CRISS in a phase 3 trial in SSc.
The CRISS was applied – using a blinded review of adverse events and serious adverse events to complete step 1 – in all patients who received study treatment, stratified by baseline IL-6 levels.
Of 104 patients in the tocilizumab arm and 106 in the placebo arm, 6 and 13 patients, respectively, had cardio-pulmonary-renal involvement and therefore received a score of 0.
Using the ACR CRISS as a continuous measure, scores favored tocilizumab over placebo at week 48 with a median increase of 0.89 vs. 0.25 points (P = .023), and using the binary form of CRISS, 51% vs. 37% of patients in the treatment and placebo arms achieved the score cutoff of 0.60 or higher at week 48 (P = .035), he said.
“The focuSSced study validates the ACR CRISS endpoint for the first time in an independent prospective clinical trial and highlights the importance of step 1 as an indicator of reduced organ progression during 48 weeks of treatment,” Dr. Khanna concluded.
The ASSET trial of abatacept
Dr. Khanna also reported results from the 12-month phase 2 ASSET trial comparing abatacept (Orencia) and placebo in early dcSSc.
Again, the change from baseline in mRSS – the primary endpoint of the study – did not differ significantly with treatment vs. placebo (–6.24 vs. –4.49; P = .28).
And again, as presented separately in a poster session at ACR, application of CRISS at 12 months – a secondary outcome measure in the trial – showed a significant difference between the groups.
The investigators prospectively performed step 1 of the CRISS using a case report form. Of 63 patients with complete data available for relevant outcomes at 12 months, 10 (5 in each group) had cardio-pulmonary-renal involvement and thus were given a score of 0.
Overall, there was evidence of significantly improved CRISS scores in the abatacept group vs. the placebo group (P = .03), he said.
Most of the individual CRISS variables, except HAQ-DI and patient global assessment, had statistically significant correlations with the CRISS, he noted, adding that “although the degree of correlation is high between the mRSS and CRISS, there is evidence that CRISS may be more sensitive to clinically meaningful treatment changes than the standard skin score endpoint.”
“This suggests further validation of CRISS as an independent primary endpoint for scleroderma clinical trials,” he concluded.
The phase 2b RISE-SSc study of riociguat
As in the focuSSced and ASSET studies, the primary efficacy endpoint of mean change from baseline in mRSS was not met in the randomized, double-blind, placebo-controlled RISE-SSc study evaluating the safety and efficacy of the soluble guanylate cyclase stimulator riociguat (Adempas) vs. placebo in 121 patients with early dcSSc, reported Oliver Distler, MD, a professor at University Hospital Zurich.
The mean change in mRSS at 52 weeks was 2.25 vs. 0.97 with riociguat vs. placebo, respectively (P = .08), although the difference in mRSS progression rate showed significant effects favoring riociguat (P = .02), he noted.
In this study, however, the proportion of patients with ACR CRISS probability of improvement (score of 0.60 or higher) at week 52 – a secondary study outcome – was the same at 18% in both arms, Dr. Distler said.
In a way, it’s helpful that the CRISS findings as well as the mRSS outcome in the RISE-SSc study were negative because it shows that “not everything comes up positive using the CRISS,” Dr. Spiera said.
An open-label extension trial of lenabasum
In the open-label extension of a phase 2 trial of lenabasum, Dr. Spiera and his colleagues also found the CRISS useful for assessing response in 36 patients. Lenabasum is a synthetic, nonimunosuppressive, selective cannabinoid receptor type 2 agonist that activates resolution of innate immune responses.
The agent continued to demonstrate acceptable safety and tolerability in dcSSc with no severe or serious adverse events or study discontinuations related to treatment during 12 months of open-label extension dosing, both from baseline and from the start of the extension, they reported in a poster at the ACR meeting. These assessments were based on ACR CRISS score, mRSS, physician global assessment, and multiple patient-reported outcomes.
The median CRISS score was 92% at week 52, and mRSS declined by a mean of 9.4 points (41.3% from baseline). More than a third of patients (35%) achieved a low mRSS of 10 or less.
The investigators noted, however, that definitive attribution of the findings to lenabasum is limited by the use of background therapy, the potential for spontaneous improvement in patients, and open-label dosing.
Evaluating immunosuppressive therapy in SSc
In another study presented at the ACR meeting, Boyang Zheng, MD, a second-year rheumatology fellow at McGill University in Montreal and his colleagues evaluated the effect of current immunosuppressive therapy on the ACR CRISS in 301 adult dcSSc patients without prior immunosuppression who were part of the Canadian Scleroderma Research Group (CSRG) registry.
Patients newly treated with methotrexate, azathioprine, mycophenolate and/or cyclophosphamide for at least 2 years (47 patients) were considered “exposed patients,” and untreated patients with at least the same follow-up duration were considered control subjects (254 patients).
Inverse probability of treatment weighting (IPTW) was performed to balance potential confounders in the two groups, including age, sex, disease duration, and CRISS variables, in an effort “to create a statistical cohort that would resemble a randomized, controlled trial,” Dr Zheng explained.
Prior to IPTW, treated patients trended towards more improvement after 1 year, but with “unimpressive absolute values.”
“But [after IPTW], when you look at overall CRISS at 1 year, more of the treated patients had actually improved – 23% vs. 11.8% in the untreated patients. After adjusting for age, sex, and disease duration, immunosuppression was associated with an almost twofold higher likelihood of improvement, although this was not statistically significant,” he said.
“Most importantly, after our balancing in the statistical cohort – so after balancing and adjusting for covariates ... immunosuppression use was still associated with a higher likelihood of improving [with an] odds ratio of 1.85 and P-value of 0.018,” he said.
The treated patients were sicker, and the CRISS was still able to capture individual patient improvement, he added.
Though limited by the observational design and the fact that “IPTW balancing cannot correct for all possible confounders,” the findings “provide novel evidence to support the use of immunosuppression in diffuse systemic sclerosis, and this reassures us in our current practice; it provides evidence that the CRISS seems to have better sensitivity to change than the mean of individual disease measures,” he said.
However, it also shows that “we have a long way to go to have better treatment for our patients,” he added, noting that only a minority (23%) of the patients in this study improved on the CRISS.
Moving forward in systemic sclerosis
Indeed, in order to move forward in developing better treatment for SSc, it is important to have the best possible means for assessing the effects of the potential new treatments, Dr. Spiera said.
“The mRSS is a good, reliable outcome measure in terms of intra- and inter-rater reliability, and we know it has meaning in the real world; in a patient with rapidly progressive skin thickening and a skin score that is getting higher and higher, we know they have worse prognosis in terms of function and even survival,” he said. “On the other hand, it’s just one piece of this very complicated story when you’re dealing with patients with systemic sclerosis, some of whom can have important internal organ involvement and not even have skin involvement.”
The main challenge with the skin score is how it performs in clinical trials when it comes to demonstrating whether a drug works, he added.
“This is particularly relevant in an era where our better understanding of the disease has led to candidate therapeutic agents that we have reason to think might work, but where clinical trials haven’t shown a significant difference between the groups using the skin score.”
Conversely, there have been studies in which skin scores improve, but the patient is doing terribly, he noted.
“That patient would not be [shown to be] doing well using the CRISS,” he said, explaining that the complex formula used to determine the CRISS score, which is “heavily weighted toward skin score,” allows for an overall score that is “greater than the sum of its parts.” That is, if a patient is improving in multiple domains indicating that the patient is responding to therapy, more credit is given for each of those parts.
“So my sense, and I think it’s the consensus in the community of clinical investigators, is that the skin score is not adequately sensitive to change in the context of a trial and doesn’t capture the disease holistically enough to be the optimal measure for scleroderma clinical trials,” he said. “We think the CRISS score works better in terms of capturing improvement in the course of a trial, and also in capturing other outcomes that are really, really important to patients and to their physicians, like disability or lung function.”
The hope is that, given the evidence, regulatory agencies will look favorably on the ACR CRISS as an outcome measure in clinical trials moving forward, and that understanding of its use and value will increase across the rheumatology community, he said.
Dr. Spiera has received research grants, consulting fees, and/or other payments from many companies involved in SSc treatments, including Roche/Genentech, which markets tocilizumab, and Corbus Pharmaceuticals, which is developing lenabasum.
Dr. Khanna is a consultant to Roche/Genentech and Bayer, which markets riociguat, and other companies. He has received research grants from Bayer, Bristol-Myers Squibb (which markets abatacept), and Pfizer. The ASSET trial he presented was sponsored by an NIH/NIAID Clinical ACE grant and an investigator-initiated grant by Bristol-Myers Squibb.
Dr. Distler has a consultancy relationship and/or has received research funding from Bayer, Roche/Genentech, and other companies. He also has a patent for a treatment for systemic sclerosis.
Dr. Zheng reported having no disclosures.
sworcester@mdedge.com
CHICAGO – At least three phase 3 randomized scleroderma treatment trials presented at the annual meeting of the American College of Rheumatology failed to meet modified Rodnan Skin Score–based primary endpoints, but a different story emerged when data from the trials were analyzed using the novel ACR Composite Response Index in Systemic Sclerosis (CRISS).
The differences highlight the limitations of individual outcome measures like the modified Rodnan Skin Score (mRSS) and underscore the need for new measures that capture the complexity of systemic sclerosis (SSc) and are more sensitive to changes in disease severity, according to Robert Spiera, MD, director of the vasculitis and scleroderma program at the Hospital for Special Surgery in New York.
Such measures are needed to better assess the effects of treatment interventions in scleroderma trials, Dr. Spiera said in an interview.
The ACR CRISS
Development of the ACR CRISS was led by Dinesh Khanna, MBBS, a professor of medicine and director of the scleroderma program at the University of Michigan, Ann Arbor. He and his colleagues described the measure in 2016 (Arthritis Rheumatol. 2016 Feb;68[2]:299-311. doi: 10.1002/art.39501).
Its use involves a two-step process of identifying any significant disease worsening or new end-organ damage, and then calculating the probability of patient improvement after 1 year of treatment on a 0- to 1-point scale based on changes from baseline in five variables: the mRSS, percent predicted forced vital capacity (FVC), patient and physician global assessments, and the Health Assessment Questionnaire Disability Index (HAQ-DI).
A CRISS score of 0.6 or higher indicates likelihood that a patient improved on treatment. Of note, subjects with significant worsening of renal or cardiopulmonary involvement are classified as not improved (score of 0), regardless of improvements in other core items.
To devise the CRISS, the investigators compiled 150 patient profiles with standardized clinical outcome elements using patients with diffuse cutaneous systemic sclerosis (dcSSc). The profiles were assessed by 40 scleroderma experts who rated patient improvement or lack thereof over 12 months.
Using the 79% of profiles for which a consensus was reached, the investigators “fit logistic regression models in which the binary outcome referred to whether the patient was improved or not, and the changes in the core set items from baseline to follow-up were entered as covariates,” they explained.
This led to the selection of the five measures included in the final version, which was found to have sensitivity of 0.982 and specificity of 0.931. When evaluated in a previously completed 1-year randomized controlled trial, the index differentiated the effect of methotrexate from the effect of placebo (P = .02), they reported.
Based on these findings, the ACR board of directors granted “provisional” endorsement of the CRISS for use in SSc clinical studies, signifying that it had been quantitatively validated using patient data, but had not undergone validation using an external data set.
New data presented at the 2018 ACR meeting will likely lead to full approval of the measure once the studies validating the measure are published, according to Dr. Spiera.
The focuSSced study of tocilizumab
For example, in the phase 3 focuSSced study comparing the interleukin-6 (IL-6) receptor–alpha antibody tocilizumab (Actemra) with placebo in patients with SSc, the primary endpoint of mean change from baseline in mRSS was not met (–6.14 vs. –4.41 points with tocilizumab vs. placebo, respectively; P = .098), Dr. Khanna said during a report of findings from the double-blind portion of the study.
However, when the CRISS was applied and its performance prospectively assessed as an exploratory outcome at week 48 in the focuSSced trial, the differences between the groups were statistically significant. Dr. Khanna reported those findings in a separate presentation at the meeting, noting that they marked the first prospective evaluation of the ACR CRISS in a phase 3 trial in SSc.
The CRISS was applied – using a blinded review of adverse events and serious adverse events to complete step 1 – in all patients who received study treatment, stratified by baseline IL-6 levels.
Of 104 patients in the tocilizumab arm and 106 in the placebo arm, 6 and 13 patients, respectively, had cardio-pulmonary-renal involvement and therefore received a score of 0.
Using the ACR CRISS as a continuous measure, scores favored tocilizumab over placebo at week 48 with a median increase of 0.89 vs. 0.25 points (P = .023), and using the binary form of CRISS, 51% vs. 37% of patients in the treatment and placebo arms achieved the score cutoff of 0.60 or higher at week 48 (P = .035), he said.
“The focuSSced study validates the ACR CRISS endpoint for the first time in an independent prospective clinical trial and highlights the importance of step 1 as an indicator of reduced organ progression during 48 weeks of treatment,” Dr. Khanna concluded.
The ASSET trial of abatacept
Dr. Khanna also reported results from the 12-month phase 2 ASSET trial comparing abatacept (Orencia) and placebo in early dcSSc.
Again, the change from baseline in mRSS – the primary endpoint of the study – did not differ significantly with treatment vs. placebo (–6.24 vs. –4.49; P = .28).
And again, as presented separately in a poster session at ACR, application of CRISS at 12 months – a secondary outcome measure in the trial – showed a significant difference between the groups.
The investigators prospectively performed step 1 of the CRISS using a case report form. Of 63 patients with complete data available for relevant outcomes at 12 months, 10 (5 in each group) had cardio-pulmonary-renal involvement and thus were given a score of 0.
Overall, there was evidence of significantly improved CRISS scores in the abatacept group vs. the placebo group (P = .03), he said.
Most of the individual CRISS variables, except HAQ-DI and patient global assessment, had statistically significant correlations with the CRISS, he noted, adding that “although the degree of correlation is high between the mRSS and CRISS, there is evidence that CRISS may be more sensitive to clinically meaningful treatment changes than the standard skin score endpoint.”
“This suggests further validation of CRISS as an independent primary endpoint for scleroderma clinical trials,” he concluded.
The phase 2b RISE-SSc study of riociguat
As in the focuSSced and ASSET studies, the primary efficacy endpoint of mean change from baseline in mRSS was not met in the randomized, double-blind, placebo-controlled RISE-SSc study evaluating the safety and efficacy of the soluble guanylate cyclase stimulator riociguat (Adempas) vs. placebo in 121 patients with early dcSSc, reported Oliver Distler, MD, a professor at University Hospital Zurich.
The mean change in mRSS at 52 weeks was 2.25 vs. 0.97 with riociguat vs. placebo, respectively (P = .08), although the difference in mRSS progression rate showed significant effects favoring riociguat (P = .02), he noted.
In this study, however, the proportion of patients with ACR CRISS probability of improvement (score of 0.60 or higher) at week 52 – a secondary study outcome – was the same at 18% in both arms, Dr. Distler said.
In a way, it’s helpful that the CRISS findings as well as the mRSS outcome in the RISE-SSc study were negative because it shows that “not everything comes up positive using the CRISS,” Dr. Spiera said.
An open-label extension trial of lenabasum
In the open-label extension of a phase 2 trial of lenabasum, Dr. Spiera and his colleagues also found the CRISS useful for assessing response in 36 patients. Lenabasum is a synthetic, nonimunosuppressive, selective cannabinoid receptor type 2 agonist that activates resolution of innate immune responses.
The agent continued to demonstrate acceptable safety and tolerability in dcSSc with no severe or serious adverse events or study discontinuations related to treatment during 12 months of open-label extension dosing, both from baseline and from the start of the extension, they reported in a poster at the ACR meeting. These assessments were based on ACR CRISS score, mRSS, physician global assessment, and multiple patient-reported outcomes.
The median CRISS score was 92% at week 52, and mRSS declined by a mean of 9.4 points (41.3% from baseline). More than a third of patients (35%) achieved a low mRSS of 10 or less.
The investigators noted, however, that definitive attribution of the findings to lenabasum is limited by the use of background therapy, the potential for spontaneous improvement in patients, and open-label dosing.
Evaluating immunosuppressive therapy in SSc
In another study presented at the ACR meeting, Boyang Zheng, MD, a second-year rheumatology fellow at McGill University in Montreal and his colleagues evaluated the effect of current immunosuppressive therapy on the ACR CRISS in 301 adult dcSSc patients without prior immunosuppression who were part of the Canadian Scleroderma Research Group (CSRG) registry.
Patients newly treated with methotrexate, azathioprine, mycophenolate and/or cyclophosphamide for at least 2 years (47 patients) were considered “exposed patients,” and untreated patients with at least the same follow-up duration were considered control subjects (254 patients).
Inverse probability of treatment weighting (IPTW) was performed to balance potential confounders in the two groups, including age, sex, disease duration, and CRISS variables, in an effort “to create a statistical cohort that would resemble a randomized, controlled trial,” Dr Zheng explained.
Prior to IPTW, treated patients trended towards more improvement after 1 year, but with “unimpressive absolute values.”
“But [after IPTW], when you look at overall CRISS at 1 year, more of the treated patients had actually improved – 23% vs. 11.8% in the untreated patients. After adjusting for age, sex, and disease duration, immunosuppression was associated with an almost twofold higher likelihood of improvement, although this was not statistically significant,” he said.
“Most importantly, after our balancing in the statistical cohort – so after balancing and adjusting for covariates ... immunosuppression use was still associated with a higher likelihood of improving [with an] odds ratio of 1.85 and P-value of 0.018,” he said.
The treated patients were sicker, and the CRISS was still able to capture individual patient improvement, he added.
Though limited by the observational design and the fact that “IPTW balancing cannot correct for all possible confounders,” the findings “provide novel evidence to support the use of immunosuppression in diffuse systemic sclerosis, and this reassures us in our current practice; it provides evidence that the CRISS seems to have better sensitivity to change than the mean of individual disease measures,” he said.
However, it also shows that “we have a long way to go to have better treatment for our patients,” he added, noting that only a minority (23%) of the patients in this study improved on the CRISS.
Moving forward in systemic sclerosis
Indeed, in order to move forward in developing better treatment for SSc, it is important to have the best possible means for assessing the effects of the potential new treatments, Dr. Spiera said.
“The mRSS is a good, reliable outcome measure in terms of intra- and inter-rater reliability, and we know it has meaning in the real world; in a patient with rapidly progressive skin thickening and a skin score that is getting higher and higher, we know they have worse prognosis in terms of function and even survival,” he said. “On the other hand, it’s just one piece of this very complicated story when you’re dealing with patients with systemic sclerosis, some of whom can have important internal organ involvement and not even have skin involvement.”
The main challenge with the skin score is how it performs in clinical trials when it comes to demonstrating whether a drug works, he added.
“This is particularly relevant in an era where our better understanding of the disease has led to candidate therapeutic agents that we have reason to think might work, but where clinical trials haven’t shown a significant difference between the groups using the skin score.”
Conversely, there have been studies in which skin scores improve, but the patient is doing terribly, he noted.
“That patient would not be [shown to be] doing well using the CRISS,” he said, explaining that the complex formula used to determine the CRISS score, which is “heavily weighted toward skin score,” allows for an overall score that is “greater than the sum of its parts.” That is, if a patient is improving in multiple domains indicating that the patient is responding to therapy, more credit is given for each of those parts.
“So my sense, and I think it’s the consensus in the community of clinical investigators, is that the skin score is not adequately sensitive to change in the context of a trial and doesn’t capture the disease holistically enough to be the optimal measure for scleroderma clinical trials,” he said. “We think the CRISS score works better in terms of capturing improvement in the course of a trial, and also in capturing other outcomes that are really, really important to patients and to their physicians, like disability or lung function.”
The hope is that, given the evidence, regulatory agencies will look favorably on the ACR CRISS as an outcome measure in clinical trials moving forward, and that understanding of its use and value will increase across the rheumatology community, he said.
Dr. Spiera has received research grants, consulting fees, and/or other payments from many companies involved in SSc treatments, including Roche/Genentech, which markets tocilizumab, and Corbus Pharmaceuticals, which is developing lenabasum.
Dr. Khanna is a consultant to Roche/Genentech and Bayer, which markets riociguat, and other companies. He has received research grants from Bayer, Bristol-Myers Squibb (which markets abatacept), and Pfizer. The ASSET trial he presented was sponsored by an NIH/NIAID Clinical ACE grant and an investigator-initiated grant by Bristol-Myers Squibb.
Dr. Distler has a consultancy relationship and/or has received research funding from Bayer, Roche/Genentech, and other companies. He also has a patent for a treatment for systemic sclerosis.
Dr. Zheng reported having no disclosures.
sworcester@mdedge.com
CHICAGO – At least three phase 3 randomized scleroderma treatment trials presented at the annual meeting of the American College of Rheumatology failed to meet modified Rodnan Skin Score–based primary endpoints, but a different story emerged when data from the trials were analyzed using the novel ACR Composite Response Index in Systemic Sclerosis (CRISS).
The differences highlight the limitations of individual outcome measures like the modified Rodnan Skin Score (mRSS) and underscore the need for new measures that capture the complexity of systemic sclerosis (SSc) and are more sensitive to changes in disease severity, according to Robert Spiera, MD, director of the vasculitis and scleroderma program at the Hospital for Special Surgery in New York.
Such measures are needed to better assess the effects of treatment interventions in scleroderma trials, Dr. Spiera said in an interview.
The ACR CRISS
Development of the ACR CRISS was led by Dinesh Khanna, MBBS, a professor of medicine and director of the scleroderma program at the University of Michigan, Ann Arbor. He and his colleagues described the measure in 2016 (Arthritis Rheumatol. 2016 Feb;68[2]:299-311. doi: 10.1002/art.39501).
Its use involves a two-step process of identifying any significant disease worsening or new end-organ damage, and then calculating the probability of patient improvement after 1 year of treatment on a 0- to 1-point scale based on changes from baseline in five variables: the mRSS, percent predicted forced vital capacity (FVC), patient and physician global assessments, and the Health Assessment Questionnaire Disability Index (HAQ-DI).
A CRISS score of 0.6 or higher indicates likelihood that a patient improved on treatment. Of note, subjects with significant worsening of renal or cardiopulmonary involvement are classified as not improved (score of 0), regardless of improvements in other core items.
To devise the CRISS, the investigators compiled 150 patient profiles with standardized clinical outcome elements using patients with diffuse cutaneous systemic sclerosis (dcSSc). The profiles were assessed by 40 scleroderma experts who rated patient improvement or lack thereof over 12 months.
Using the 79% of profiles for which a consensus was reached, the investigators “fit logistic regression models in which the binary outcome referred to whether the patient was improved or not, and the changes in the core set items from baseline to follow-up were entered as covariates,” they explained.
This led to the selection of the five measures included in the final version, which was found to have sensitivity of 0.982 and specificity of 0.931. When evaluated in a previously completed 1-year randomized controlled trial, the index differentiated the effect of methotrexate from the effect of placebo (P = .02), they reported.
Based on these findings, the ACR board of directors granted “provisional” endorsement of the CRISS for use in SSc clinical studies, signifying that it had been quantitatively validated using patient data, but had not undergone validation using an external data set.
New data presented at the 2018 ACR meeting will likely lead to full approval of the measure once the studies validating the measure are published, according to Dr. Spiera.
The focuSSced study of tocilizumab
For example, in the phase 3 focuSSced study comparing the interleukin-6 (IL-6) receptor–alpha antibody tocilizumab (Actemra) with placebo in patients with SSc, the primary endpoint of mean change from baseline in mRSS was not met (–6.14 vs. –4.41 points with tocilizumab vs. placebo, respectively; P = .098), Dr. Khanna said during a report of findings from the double-blind portion of the study.
However, when the CRISS was applied and its performance prospectively assessed as an exploratory outcome at week 48 in the focuSSced trial, the differences between the groups were statistically significant. Dr. Khanna reported those findings in a separate presentation at the meeting, noting that they marked the first prospective evaluation of the ACR CRISS in a phase 3 trial in SSc.
The CRISS was applied – using a blinded review of adverse events and serious adverse events to complete step 1 – in all patients who received study treatment, stratified by baseline IL-6 levels.
Of 104 patients in the tocilizumab arm and 106 in the placebo arm, 6 and 13 patients, respectively, had cardio-pulmonary-renal involvement and therefore received a score of 0.
Using the ACR CRISS as a continuous measure, scores favored tocilizumab over placebo at week 48 with a median increase of 0.89 vs. 0.25 points (P = .023), and using the binary form of CRISS, 51% vs. 37% of patients in the treatment and placebo arms achieved the score cutoff of 0.60 or higher at week 48 (P = .035), he said.
“The focuSSced study validates the ACR CRISS endpoint for the first time in an independent prospective clinical trial and highlights the importance of step 1 as an indicator of reduced organ progression during 48 weeks of treatment,” Dr. Khanna concluded.
The ASSET trial of abatacept
Dr. Khanna also reported results from the 12-month phase 2 ASSET trial comparing abatacept (Orencia) and placebo in early dcSSc.
Again, the change from baseline in mRSS – the primary endpoint of the study – did not differ significantly with treatment vs. placebo (–6.24 vs. –4.49; P = .28).
And again, as presented separately in a poster session at ACR, application of CRISS at 12 months – a secondary outcome measure in the trial – showed a significant difference between the groups.
The investigators prospectively performed step 1 of the CRISS using a case report form. Of 63 patients with complete data available for relevant outcomes at 12 months, 10 (5 in each group) had cardio-pulmonary-renal involvement and thus were given a score of 0.
Overall, there was evidence of significantly improved CRISS scores in the abatacept group vs. the placebo group (P = .03), he said.
Most of the individual CRISS variables, except HAQ-DI and patient global assessment, had statistically significant correlations with the CRISS, he noted, adding that “although the degree of correlation is high between the mRSS and CRISS, there is evidence that CRISS may be more sensitive to clinically meaningful treatment changes than the standard skin score endpoint.”
“This suggests further validation of CRISS as an independent primary endpoint for scleroderma clinical trials,” he concluded.
The phase 2b RISE-SSc study of riociguat
As in the focuSSced and ASSET studies, the primary efficacy endpoint of mean change from baseline in mRSS was not met in the randomized, double-blind, placebo-controlled RISE-SSc study evaluating the safety and efficacy of the soluble guanylate cyclase stimulator riociguat (Adempas) vs. placebo in 121 patients with early dcSSc, reported Oliver Distler, MD, a professor at University Hospital Zurich.
The mean change in mRSS at 52 weeks was 2.25 vs. 0.97 with riociguat vs. placebo, respectively (P = .08), although the difference in mRSS progression rate showed significant effects favoring riociguat (P = .02), he noted.
In this study, however, the proportion of patients with ACR CRISS probability of improvement (score of 0.60 or higher) at week 52 – a secondary study outcome – was the same at 18% in both arms, Dr. Distler said.
In a way, it’s helpful that the CRISS findings as well as the mRSS outcome in the RISE-SSc study were negative because it shows that “not everything comes up positive using the CRISS,” Dr. Spiera said.
An open-label extension trial of lenabasum
In the open-label extension of a phase 2 trial of lenabasum, Dr. Spiera and his colleagues also found the CRISS useful for assessing response in 36 patients. Lenabasum is a synthetic, nonimunosuppressive, selective cannabinoid receptor type 2 agonist that activates resolution of innate immune responses.
The agent continued to demonstrate acceptable safety and tolerability in dcSSc with no severe or serious adverse events or study discontinuations related to treatment during 12 months of open-label extension dosing, both from baseline and from the start of the extension, they reported in a poster at the ACR meeting. These assessments were based on ACR CRISS score, mRSS, physician global assessment, and multiple patient-reported outcomes.
The median CRISS score was 92% at week 52, and mRSS declined by a mean of 9.4 points (41.3% from baseline). More than a third of patients (35%) achieved a low mRSS of 10 or less.
The investigators noted, however, that definitive attribution of the findings to lenabasum is limited by the use of background therapy, the potential for spontaneous improvement in patients, and open-label dosing.
Evaluating immunosuppressive therapy in SSc
In another study presented at the ACR meeting, Boyang Zheng, MD, a second-year rheumatology fellow at McGill University in Montreal and his colleagues evaluated the effect of current immunosuppressive therapy on the ACR CRISS in 301 adult dcSSc patients without prior immunosuppression who were part of the Canadian Scleroderma Research Group (CSRG) registry.
Patients newly treated with methotrexate, azathioprine, mycophenolate and/or cyclophosphamide for at least 2 years (47 patients) were considered “exposed patients,” and untreated patients with at least the same follow-up duration were considered control subjects (254 patients).
Inverse probability of treatment weighting (IPTW) was performed to balance potential confounders in the two groups, including age, sex, disease duration, and CRISS variables, in an effort “to create a statistical cohort that would resemble a randomized, controlled trial,” Dr Zheng explained.
Prior to IPTW, treated patients trended towards more improvement after 1 year, but with “unimpressive absolute values.”
“But [after IPTW], when you look at overall CRISS at 1 year, more of the treated patients had actually improved – 23% vs. 11.8% in the untreated patients. After adjusting for age, sex, and disease duration, immunosuppression was associated with an almost twofold higher likelihood of improvement, although this was not statistically significant,” he said.
“Most importantly, after our balancing in the statistical cohort – so after balancing and adjusting for covariates ... immunosuppression use was still associated with a higher likelihood of improving [with an] odds ratio of 1.85 and P-value of 0.018,” he said.
The treated patients were sicker, and the CRISS was still able to capture individual patient improvement, he added.
Though limited by the observational design and the fact that “IPTW balancing cannot correct for all possible confounders,” the findings “provide novel evidence to support the use of immunosuppression in diffuse systemic sclerosis, and this reassures us in our current practice; it provides evidence that the CRISS seems to have better sensitivity to change than the mean of individual disease measures,” he said.
However, it also shows that “we have a long way to go to have better treatment for our patients,” he added, noting that only a minority (23%) of the patients in this study improved on the CRISS.
Moving forward in systemic sclerosis
Indeed, in order to move forward in developing better treatment for SSc, it is important to have the best possible means for assessing the effects of the potential new treatments, Dr. Spiera said.
“The mRSS is a good, reliable outcome measure in terms of intra- and inter-rater reliability, and we know it has meaning in the real world; in a patient with rapidly progressive skin thickening and a skin score that is getting higher and higher, we know they have worse prognosis in terms of function and even survival,” he said. “On the other hand, it’s just one piece of this very complicated story when you’re dealing with patients with systemic sclerosis, some of whom can have important internal organ involvement and not even have skin involvement.”
The main challenge with the skin score is how it performs in clinical trials when it comes to demonstrating whether a drug works, he added.
“This is particularly relevant in an era where our better understanding of the disease has led to candidate therapeutic agents that we have reason to think might work, but where clinical trials haven’t shown a significant difference between the groups using the skin score.”
Conversely, there have been studies in which skin scores improve, but the patient is doing terribly, he noted.
“That patient would not be [shown to be] doing well using the CRISS,” he said, explaining that the complex formula used to determine the CRISS score, which is “heavily weighted toward skin score,” allows for an overall score that is “greater than the sum of its parts.” That is, if a patient is improving in multiple domains indicating that the patient is responding to therapy, more credit is given for each of those parts.
“So my sense, and I think it’s the consensus in the community of clinical investigators, is that the skin score is not adequately sensitive to change in the context of a trial and doesn’t capture the disease holistically enough to be the optimal measure for scleroderma clinical trials,” he said. “We think the CRISS score works better in terms of capturing improvement in the course of a trial, and also in capturing other outcomes that are really, really important to patients and to their physicians, like disability or lung function.”
The hope is that, given the evidence, regulatory agencies will look favorably on the ACR CRISS as an outcome measure in clinical trials moving forward, and that understanding of its use and value will increase across the rheumatology community, he said.
Dr. Spiera has received research grants, consulting fees, and/or other payments from many companies involved in SSc treatments, including Roche/Genentech, which markets tocilizumab, and Corbus Pharmaceuticals, which is developing lenabasum.
Dr. Khanna is a consultant to Roche/Genentech and Bayer, which markets riociguat, and other companies. He has received research grants from Bayer, Bristol-Myers Squibb (which markets abatacept), and Pfizer. The ASSET trial he presented was sponsored by an NIH/NIAID Clinical ACE grant and an investigator-initiated grant by Bristol-Myers Squibb.
Dr. Distler has a consultancy relationship and/or has received research funding from Bayer, Roche/Genentech, and other companies. He also has a patent for a treatment for systemic sclerosis.
Dr. Zheng reported having no disclosures.
sworcester@mdedge.com
REPORTING FROM THE ACR ANNUAL MEETING