Publicação
Synovial pathobiology and patient stratification in rheumatoid arthritis
| Resumo: | Rheumatoid arthritis (RA) is the most common chronic systemic immune-mediated rheumatic disease. It causes considerable morbidity, reduces quality of life and increases mortality, resulting in high personal, social and economic costs. The last two decades have witnessed a major improvement in the prognosis of RA patients fueled by the better understanding of disease pathogenesis, the emergence of novel therapies and the adoption of structured tight control treatment strategies. However, many unmet needs remain in the management of RA. Biologic diseasemodifying antirheumatic drugs (bDMARDs) have undoubtedly improved clinical outcomes of patients refractory to conventional synthetic DMARDs (csDMARDs). Nevertheless, they represent a significant cost, 30-40% of patients do not respond at all to these drugs and only 20-25% achieve optimal response. In addition, the mechanisms of non-response remain unknown and, as such, bDMARDs are still used on a trial-and-error basis. Indeed, RA is a clinically heterogeneous condition with a highly variable natural course, several disease subsets and very inconsistent and unpredictable response to current treatments. Presently, the clinician is not able to base therapeutic decisions on reliable predictive markers, resulting in the delay of appropriate, effective treatment in poor prognostic cases. Thus, it is crucial to develop tools that will enable to stratify patients by prognosis and to tailor therapy accordingly. Synovial inflammation is the key drive in RA pathogenesis. It is highly heterogeneous and potentially related to the observed clinical variability. Recently, the emergence of minimally invasive ultrasound (US)-guided synovial biopsies has offered the possibility of accessing the focus of the disease process – the synovial tissue – in small joints, mostly affected by RA. If validated for clinical practice and research purposes, they could be adopted at a wider scale for the study of synovitis. We hypothesize that the synovial tissue pathobiology heterogeneity is associated with variable clinical phenotypes, disease course and treatment response patterns. If this proves true, synovial biopsy could be incorporated into routine diagnostic and therapeutic management of RA, informing prognosis and assisting treatment decisions. Eventually, this would enable a shift from the current trial-and-error approach, towards a rational pathobiology-based patient stratification and treatment personalization. In this thesis, we aimed to study the current status of rational treatment selection and patient stratification in RA, and the contribution of synovial tissue to achieve this goal. We first compared the effectiveness of two of the main classes of bDMARDs in RApatients treated in a real-world setting. Next, we reviewed the current standing of predictive biomarkers of response to bDMARD therapy in RA. In the third part, we validated the role of US-guided synovial biopsies as a novel technique to retrieve synovial tissue in clinical practice and research. Finally, we explored the potential of synovial tissue pathobiology in comparing RA subsets of disease and in enriching prediction of response-resistance to current bDMARD therapies. The efficacy of all bDMARDs is generally considered to be similar, based on indirect comparisons of clinical trials. However, the effectiveness of these drugs in clinical practice may be influenced by many other factors, including the greater heterogeneity of the real-life patient population. We therefore used nationwide data from Reuma.pt (Rheumatic Diseases Portuguese Register) to investigate the short-term effectiveness of tumor necrosis factor inhibitors (TNFis) and tocilizumab (a blocker of interleukin-6 receptor) in 524 RA patients. Treatment outcomes at 6 months tended to be better with tocilizumab, particularly in bDMARD-naïve patients. The results were influenced by the measures of disease activity used and statistical accommodation of confounders, but remained consistent even for clinical-based indices. These findings are relevant, considering the long-term benefit of early disease control. While not conclusively attesting relative drug superiority, they indicate variable bDMARD effectiveness in clinical practice. This implies that the common practice of electing TNFis as first-line agents is no longer valid and should be abandoned. Moreover, we confirmed the limited outcomes obtained with current therapeutic modalities in clinical practice, with only 20-30% of patients attaining the crucial target of early remission. To search for ways to improve treatment selection, we reviewed the literature for predictive biomarkers of response to bDMARDs. Overall, the many studies available have had limited impact in clinical practice. The association of rheumatoid factor and/or anti-citrullinated protein antibodies positivity with better response to rituximab or abatacept is one of the exceptions. Nonetheless, promising data exist for serum and, especially, synovial tissue-based biomarkers of differential response to bDMARDs. Emerging techniques such as US-guided synovial biopsies are likely to become fundamental in further investigating these findings in larger hypothesis-driven studies. Accordingly, in the third part of the thesis we validated the performance, utility and safety of US-guided synovial biopsies in clinical practice and research. In a multicenter study, we showed that US-guided synovial biopsies are as reliable as arthroscopy and superior to blind needle biopsies in retrieving good-quality samples for histological and molecular analysis. They are suitable for both large and small joints, can be used in sequential procedures of the same joint and are, thus, a valid option to adopt in translational research studies. The evidence on the outcomes of US-guided synovial biopsies in clinical practice is more limited. As such, we prospectively demonstrated the tolerability, usefulness, safety and efficacy of US-guided synovial needle biopsies performed in our department for clinical and research purposes. Synovial tissue with adequate quality could be successfully collected from all types of joints, bursae and tendon sheaths and frequently assisted in clinical diagnosis. Moreover, we were able to classify samples in terms of synovitis score, grade and pathotype, providing an objective quantification of the degree of synovitis. Together with sample quality, these features showed significant variation across patients, samples and sampling order, but not different cutting levels. Taking into account the probable expansion of US-guided synovial biopsies to several centers, there is a need to standardize procedures related to synovial tissue collection, processing and analysis. Hence, we joined a multicenter collaborative initiative to issue consensus-based points to consider for harmonization of synovial biopsy procedures in clinical practice and translational research. As a whole, the results obtained in this part of the thesis support the use of US-guided synovial biopsies as a novel means to study synovitis for clinical or research purposes. In this regard, it has the potential to foster the development of large studies aimed at improving prognostic and predictive algorithms in RA. In the last part of the thesis, we systematically applied US-guided synovial biopsies to the study of the relationship between synovial pathobiology and clinical, imaging, and treatment response outcomes of two RA subsets. Younger- (YORA) and elderly-onset RA (EORA) patients with early disease had an US-guided synovial biopsy before starting csDMARDs, thus prior to potential modification of disease tissue pathology. We demonstrated that EORA is associated with distinctive clinical features, higher comorbidity burden, increased US synovitis and more severe radiographic damage. Moreover, EORA was independently associated with worse clinical response and higher radiographic progression than YORA, even in spite of comparable therapy. The more severe phenotype of EORA was also reflected at the tissue level, with lesser improvement of histological synovitis and lower frequency of pauci-immune-fibroid pathotype after treatment. In addition, histological synovitis was informative of clinical and US activity, and this association differed according to RA subset and treatment status. Importantly, a lympho-myeloid pathotype predicted more active disease and radiographic progression, but was responsive to immunosuppressive treatment. The results of this study have practical clinical implications. The view of EORA as a benign subset of disease is not justified. Early aggressive intervention with a treat-to-target strategy, aiming at stringent control of disease, should be adopted for these vulnerable patients, particularly those with a lympho-myeloid synovial pathotype. As a closure to our efforts, we made a significant contribution for the first synovial tissue-driven controlled trial that randomized RA patients to two bDMARDs, based on synovial B cell infiltration status. The encouraging preliminary results indicate that rituximab is less effective than tocilizumab in patients with B cell-poor synovitis. If confirmed, they could be the first robust proof that synovial pathobiology may allow patient stratification and impactfully inform treatment decisions. Hopefully, this could mean that we are entering a new era in the management of RA, wherein individualized care and precision medicine can be routinely delivered to improve patient outcomes. In conclusion, this thesis provides new evidence of the current challenges in the management of RA and of the potential of US-guided synovial biopsies and synovial tissue pathobiology to improve RA care. |
|---|---|
| Autores principais: | Romão, Vasco Madeira Crispim |
| Assunto: | Artrite reumatoide Patobiologia sinovial Biopsia sinovial Estratificação de doentes Personalização terapêutica Teses de doutoramento - 2020 |
| Ano: | 2020 |
| País: | Portugal |
| Tipo de documento: | tese de doutoramento |
| Tipo de acesso: | acesso restrito |
| Instituição associada: | Universidade de Lisboa |
| Idioma: | inglês |
| Origem: | Repositório da Universidade de Lisboa |
| Resumo: | Rheumatoid arthritis (RA) is the most common chronic systemic immune-mediated rheumatic disease. It causes considerable morbidity, reduces quality of life and increases mortality, resulting in high personal, social and economic costs. The last two decades have witnessed a major improvement in the prognosis of RA patients fueled by the better understanding of disease pathogenesis, the emergence of novel therapies and the adoption of structured tight control treatment strategies. However, many unmet needs remain in the management of RA. Biologic diseasemodifying antirheumatic drugs (bDMARDs) have undoubtedly improved clinical outcomes of patients refractory to conventional synthetic DMARDs (csDMARDs). Nevertheless, they represent a significant cost, 30-40% of patients do not respond at all to these drugs and only 20-25% achieve optimal response. In addition, the mechanisms of non-response remain unknown and, as such, bDMARDs are still used on a trial-and-error basis. Indeed, RA is a clinically heterogeneous condition with a highly variable natural course, several disease subsets and very inconsistent and unpredictable response to current treatments. Presently, the clinician is not able to base therapeutic decisions on reliable predictive markers, resulting in the delay of appropriate, effective treatment in poor prognostic cases. Thus, it is crucial to develop tools that will enable to stratify patients by prognosis and to tailor therapy accordingly. Synovial inflammation is the key drive in RA pathogenesis. It is highly heterogeneous and potentially related to the observed clinical variability. Recently, the emergence of minimally invasive ultrasound (US)-guided synovial biopsies has offered the possibility of accessing the focus of the disease process – the synovial tissue – in small joints, mostly affected by RA. If validated for clinical practice and research purposes, they could be adopted at a wider scale for the study of synovitis. We hypothesize that the synovial tissue pathobiology heterogeneity is associated with variable clinical phenotypes, disease course and treatment response patterns. If this proves true, synovial biopsy could be incorporated into routine diagnostic and therapeutic management of RA, informing prognosis and assisting treatment decisions. Eventually, this would enable a shift from the current trial-and-error approach, towards a rational pathobiology-based patient stratification and treatment personalization. In this thesis, we aimed to study the current status of rational treatment selection and patient stratification in RA, and the contribution of synovial tissue to achieve this goal. We first compared the effectiveness of two of the main classes of bDMARDs in RApatients treated in a real-world setting. Next, we reviewed the current standing of predictive biomarkers of response to bDMARD therapy in RA. In the third part, we validated the role of US-guided synovial biopsies as a novel technique to retrieve synovial tissue in clinical practice and research. Finally, we explored the potential of synovial tissue pathobiology in comparing RA subsets of disease and in enriching prediction of response-resistance to current bDMARD therapies. The efficacy of all bDMARDs is generally considered to be similar, based on indirect comparisons of clinical trials. However, the effectiveness of these drugs in clinical practice may be influenced by many other factors, including the greater heterogeneity of the real-life patient population. We therefore used nationwide data from Reuma.pt (Rheumatic Diseases Portuguese Register) to investigate the short-term effectiveness of tumor necrosis factor inhibitors (TNFis) and tocilizumab (a blocker of interleukin-6 receptor) in 524 RA patients. Treatment outcomes at 6 months tended to be better with tocilizumab, particularly in bDMARD-naïve patients. The results were influenced by the measures of disease activity used and statistical accommodation of confounders, but remained consistent even for clinical-based indices. These findings are relevant, considering the long-term benefit of early disease control. While not conclusively attesting relative drug superiority, they indicate variable bDMARD effectiveness in clinical practice. This implies that the common practice of electing TNFis as first-line agents is no longer valid and should be abandoned. Moreover, we confirmed the limited outcomes obtained with current therapeutic modalities in clinical practice, with only 20-30% of patients attaining the crucial target of early remission. To search for ways to improve treatment selection, we reviewed the literature for predictive biomarkers of response to bDMARDs. Overall, the many studies available have had limited impact in clinical practice. The association of rheumatoid factor and/or anti-citrullinated protein antibodies positivity with better response to rituximab or abatacept is one of the exceptions. Nonetheless, promising data exist for serum and, especially, synovial tissue-based biomarkers of differential response to bDMARDs. Emerging techniques such as US-guided synovial biopsies are likely to become fundamental in further investigating these findings in larger hypothesis-driven studies. Accordingly, in the third part of the thesis we validated the performance, utility and safety of US-guided synovial biopsies in clinical practice and research. In a multicenter study, we showed that US-guided synovial biopsies are as reliable as arthroscopy and superior to blind needle biopsies in retrieving good-quality samples for histological and molecular analysis. They are suitable for both large and small joints, can be used in sequential procedures of the same joint and are, thus, a valid option to adopt in translational research studies. The evidence on the outcomes of US-guided synovial biopsies in clinical practice is more limited. As such, we prospectively demonstrated the tolerability, usefulness, safety and efficacy of US-guided synovial needle biopsies performed in our department for clinical and research purposes. Synovial tissue with adequate quality could be successfully collected from all types of joints, bursae and tendon sheaths and frequently assisted in clinical diagnosis. Moreover, we were able to classify samples in terms of synovitis score, grade and pathotype, providing an objective quantification of the degree of synovitis. Together with sample quality, these features showed significant variation across patients, samples and sampling order, but not different cutting levels. Taking into account the probable expansion of US-guided synovial biopsies to several centers, there is a need to standardize procedures related to synovial tissue collection, processing and analysis. Hence, we joined a multicenter collaborative initiative to issue consensus-based points to consider for harmonization of synovial biopsy procedures in clinical practice and translational research. As a whole, the results obtained in this part of the thesis support the use of US-guided synovial biopsies as a novel means to study synovitis for clinical or research purposes. In this regard, it has the potential to foster the development of large studies aimed at improving prognostic and predictive algorithms in RA. In the last part of the thesis, we systematically applied US-guided synovial biopsies to the study of the relationship between synovial pathobiology and clinical, imaging, and treatment response outcomes of two RA subsets. Younger- (YORA) and elderly-onset RA (EORA) patients with early disease had an US-guided synovial biopsy before starting csDMARDs, thus prior to potential modification of disease tissue pathology. We demonstrated that EORA is associated with distinctive clinical features, higher comorbidity burden, increased US synovitis and more severe radiographic damage. Moreover, EORA was independently associated with worse clinical response and higher radiographic progression than YORA, even in spite of comparable therapy. The more severe phenotype of EORA was also reflected at the tissue level, with lesser improvement of histological synovitis and lower frequency of pauci-immune-fibroid pathotype after treatment. In addition, histological synovitis was informative of clinical and US activity, and this association differed according to RA subset and treatment status. Importantly, a lympho-myeloid pathotype predicted more active disease and radiographic progression, but was responsive to immunosuppressive treatment. The results of this study have practical clinical implications. The view of EORA as a benign subset of disease is not justified. Early aggressive intervention with a treat-to-target strategy, aiming at stringent control of disease, should be adopted for these vulnerable patients, particularly those with a lympho-myeloid synovial pathotype. As a closure to our efforts, we made a significant contribution for the first synovial tissue-driven controlled trial that randomized RA patients to two bDMARDs, based on synovial B cell infiltration status. The encouraging preliminary results indicate that rituximab is less effective than tocilizumab in patients with B cell-poor synovitis. If confirmed, they could be the first robust proof that synovial pathobiology may allow patient stratification and impactfully inform treatment decisions. Hopefully, this could mean that we are entering a new era in the management of RA, wherein individualized care and precision medicine can be routinely delivered to improve patient outcomes. In conclusion, this thesis provides new evidence of the current challenges in the management of RA and of the potential of US-guided synovial biopsies and synovial tissue pathobiology to improve RA care. |
|---|