Two Biomarkers Predict Response to Rituximab in RA

CHICAGO – Early B-cell repopulation and rheumatoid factor positivity predict early relapse in patients with rheumatoid arthritis who are being treated with rituximab, according to an observational study of 104 patients.

"We’re interested in predicting duration of response to rituximab," said lead author Dr. Edward M. Vital of the University of Leeds (England). "[W]hilst the majority of patients do respond, virtually all will relapse at some point. ... [T]he time to relapse is very variable, so that means that we don’t know what the best retreatment regime is. Some advocate retreatment every 6 months, but this may be excessive for some patients, with implications for cost and possibly safety."

With the sensitive B-cell assay, 88% of patients had detectable B cells by 6 months. However, a small group of patients have persistent B-cell depletion at 6 months, and this group had a longer clinical response. In fact, their responses continued to improve from 6 to 9 months, said Dr. Vital. Maintaining a low B-cell number is likely to maintain a good clinical response, he said, although not all patients with B-cell return relapsed immediately. Dr. Vital presented his data at the annual meeting of the American College of Rheumatology.

The study’s objective was to examine whether known predictors of initial response to rituximab could be used after treatment to predict maintenance of response to 12 months.

All 104 patients were positive for either rheumatoid factor or anti–cyclic citrullinated peptide at baseline, prior to treatment with rituximab in a regimen involving infused doses of 1,000 mg with concomitant methotrexate or leflunomide. Of the original study population, 78 (75%) had a known EULAR (European League Against Rheumatism) response at 6 months, and these were investigated for time to relapse. All use of corticosteroids and NSAIDs was similarly controlled in this group, and outcomes were measured by EULAR response and DAS28 (Disease Activity Score including a 28-joint count) at 0, 6, 9, and 12 months by one of two joint-count assessors.

Highly sensitive flow cytometry with a limit of detection of 0.0001 cells x 10⁹/L was used to measure naive, memory B, and plasmablast subsets. Conventional flow cytometry has a limit of detection of 0.01 cells x 10⁹/L.

At 6 months, 44 (57%) of the 78 patients had a EULAR moderate response, and 34 (43%) had a EULAR good response. In patients with detectable B cells at 6 months, DAS28 worsened by a mean 0.45, whereas in patients with undetectable B cells, it improved by a mean 0.44.

At 12 months, 40 patients (51%) had no response, 22 patients (28%) had a EULAR moderate response, and 16 patients (21%) had a EULAR good response.

A binary logistic regression model was created to predict continued EULAR response at 12 months. It included B-cell detection, plasmablast count, rheumatoid factor, and DAS28, and produced a significant (P = .001) model with an overall accuracy of 72%.

"Since rheumatoid factor and DAS28 were the strongest predictors of these, this suggested that it might actually be possible to make a very simple prediction model that doesn’t require specialized techniques," said Dr. Vital.

The authors subsequently created a simplified, two-variable categorical model using a rheumatoid factor threshold of 108 IU/mL, and a DAS28 threshold of 3.7, to predict relapse. This simplified model’s overall accuracy was 69% (P = .001).

"As you can see, we still have a significant model with just a slight reduction in the overall accuracy. ... If at 6 months the DAS28 was less than 3.7 and the rheumatoid factor was less than 108, then 73% of those patients maintained their clinical response out to 12 months ... and if both [DAS28 and rheumatoid factor] were higher, then only 8% of patients responded out to 12 months," said Dr. Vital.

The study concluded that these findings – in particular, the levels used in the two-variable model – need to be validated in another population. When validated, they may allow for selection of retreatment regime based on B-cell biomarkers.

Dr. Vital is on the Roche speakers bureau, and Roche provided rituximab for the patients in this study. The study was funded by the U.K. National Institute for Health Research Doctoral Research Fellowship and sponsored by the NIHR and the University of Leeds.

CHICAGO – Early B-cell repopulation and rheumatoid factor positivity predict early relapse in patients with rheumatoid arthritis who are being treated with rituximab, according to an observational study of 104 patients.

"We’re interested in predicting duration of response to rituximab," said lead author Dr. Edward M. Vital of the University of Leeds (England). "[W]hilst the majority of patients do respond, virtually all will relapse at some point. ... [T]he time to relapse is very variable, so that means that we don’t know what the best retreatment regime is. Some advocate retreatment every 6 months, but this may be excessive for some patients, with implications for cost and possibly safety."

With the sensitive B-cell assay, 88% of patients had detectable B cells by 6 months. However, a small group of patients have persistent B-cell depletion at 6 months, and this group had a longer clinical response. In fact, their responses continued to improve from 6 to 9 months, said Dr. Vital. Maintaining a low B-cell number is likely to maintain a good clinical response, he said, although not all patients with B-cell return relapsed immediately. Dr. Vital presented his data at the annual meeting of the American College of Rheumatology.

The study’s objective was to examine whether known predictors of initial response to rituximab could be used after treatment to predict maintenance of response to 12 months.

All 104 patients were positive for either rheumatoid factor or anti–cyclic citrullinated peptide at baseline, prior to treatment with rituximab in a regimen involving infused doses of 1,000 mg with concomitant methotrexate or leflunomide. Of the original study population, 78 (75%) had a known EULAR (European League Against Rheumatism) response at 6 months, and these were investigated for time to relapse. All use of corticosteroids and NSAIDs was similarly controlled in this group, and outcomes were measured by EULAR response and DAS28 (Disease Activity Score including a 28-joint count) at 0, 6, 9, and 12 months by one of two joint-count assessors.

Highly sensitive flow cytometry with a limit of detection of 0.0001 cells x 10⁹/L was used to measure naive, memory B, and plasmablast subsets. Conventional flow cytometry has a limit of detection of 0.01 cells x 10⁹/L.

At 6 months, 44 (57%) of the 78 patients had a EULAR moderate response, and 34 (43%) had a EULAR good response. In patients with detectable B cells at 6 months, DAS28 worsened by a mean 0.45, whereas in patients with undetectable B cells, it improved by a mean 0.44.

At 12 months, 40 patients (51%) had no response, 22 patients (28%) had a EULAR moderate response, and 16 patients (21%) had a EULAR good response.

A binary logistic regression model was created to predict continued EULAR response at 12 months. It included B-cell detection, plasmablast count, rheumatoid factor, and DAS28, and produced a significant (P = .001) model with an overall accuracy of 72%.

"Since rheumatoid factor and DAS28 were the strongest predictors of these, this suggested that it might actually be possible to make a very simple prediction model that doesn’t require specialized techniques," said Dr. Vital.

The authors subsequently created a simplified, two-variable categorical model using a rheumatoid factor threshold of 108 IU/mL, and a DAS28 threshold of 3.7, to predict relapse. This simplified model’s overall accuracy was 69% (P = .001).

"As you can see, we still have a significant model with just a slight reduction in the overall accuracy. ... If at 6 months the DAS28 was less than 3.7 and the rheumatoid factor was less than 108, then 73% of those patients maintained their clinical response out to 12 months ... and if both [DAS28 and rheumatoid factor] were higher, then only 8% of patients responded out to 12 months," said Dr. Vital.

The study concluded that these findings – in particular, the levels used in the two-variable model – need to be validated in another population. When validated, they may allow for selection of retreatment regime based on B-cell biomarkers.

Dr. Vital is on the Roche speakers bureau, and Roche provided rituximab for the patients in this study. The study was funded by the U.K. National Institute for Health Research Doctoral Research Fellowship and sponsored by the NIHR and the University of Leeds.

CHICAGO – Early B-cell repopulation and rheumatoid factor positivity predict early relapse in patients with rheumatoid arthritis who are being treated with rituximab, according to an observational study of 104 patients.

"We’re interested in predicting duration of response to rituximab," said lead author Dr. Edward M. Vital of the University of Leeds (England). "[W]hilst the majority of patients do respond, virtually all will relapse at some point. ... [T]he time to relapse is very variable, so that means that we don’t know what the best retreatment regime is. Some advocate retreatment every 6 months, but this may be excessive for some patients, with implications for cost and possibly safety."

With the sensitive B-cell assay, 88% of patients had detectable B cells by 6 months. However, a small group of patients have persistent B-cell depletion at 6 months, and this group had a longer clinical response. In fact, their responses continued to improve from 6 to 9 months, said Dr. Vital. Maintaining a low B-cell number is likely to maintain a good clinical response, he said, although not all patients with B-cell return relapsed immediately. Dr. Vital presented his data at the annual meeting of the American College of Rheumatology.

The study’s objective was to examine whether known predictors of initial response to rituximab could be used after treatment to predict maintenance of response to 12 months.

All 104 patients were positive for either rheumatoid factor or anti–cyclic citrullinated peptide at baseline, prior to treatment with rituximab in a regimen involving infused doses of 1,000 mg with concomitant methotrexate or leflunomide. Of the original study population, 78 (75%) had a known EULAR (European League Against Rheumatism) response at 6 months, and these were investigated for time to relapse. All use of corticosteroids and NSAIDs was similarly controlled in this group, and outcomes were measured by EULAR response and DAS28 (Disease Activity Score including a 28-joint count) at 0, 6, 9, and 12 months by one of two joint-count assessors.

Highly sensitive flow cytometry with a limit of detection of 0.0001 cells x 10⁹/L was used to measure naive, memory B, and plasmablast subsets. Conventional flow cytometry has a limit of detection of 0.01 cells x 10⁹/L.

At 6 months, 44 (57%) of the 78 patients had a EULAR moderate response, and 34 (43%) had a EULAR good response. In patients with detectable B cells at 6 months, DAS28 worsened by a mean 0.45, whereas in patients with undetectable B cells, it improved by a mean 0.44.

At 12 months, 40 patients (51%) had no response, 22 patients (28%) had a EULAR moderate response, and 16 patients (21%) had a EULAR good response.

A binary logistic regression model was created to predict continued EULAR response at 12 months. It included B-cell detection, plasmablast count, rheumatoid factor, and DAS28, and produced a significant (P = .001) model with an overall accuracy of 72%.

"Since rheumatoid factor and DAS28 were the strongest predictors of these, this suggested that it might actually be possible to make a very simple prediction model that doesn’t require specialized techniques," said Dr. Vital.

The authors subsequently created a simplified, two-variable categorical model using a rheumatoid factor threshold of 108 IU/mL, and a DAS28 threshold of 3.7, to predict relapse. This simplified model’s overall accuracy was 69% (P = .001).

"As you can see, we still have a significant model with just a slight reduction in the overall accuracy. ... If at 6 months the DAS28 was less than 3.7 and the rheumatoid factor was less than 108, then 73% of those patients maintained their clinical response out to 12 months ... and if both [DAS28 and rheumatoid factor] were higher, then only 8% of patients responded out to 12 months," said Dr. Vital.

The study concluded that these findings – in particular, the levels used in the two-variable model – need to be validated in another population. When validated, they may allow for selection of retreatment regime based on B-cell biomarkers.

Dr. Vital is on the Roche speakers bureau, and Roche provided rituximab for the patients in this study. The study was funded by the U.K. National Institute for Health Research Doctoral Research Fellowship and sponsored by the NIHR and the University of Leeds.