Longitudinal analysis of psoriasis, rheumatoid arthritis, psoriatic arthritis, axial spondylo-arthritis and chronic intestinal diseases

A number of chronic, immune-mediated inflammatory diseases (IMID) have the potential to result in secondary diseases from the underlying pathophysiological form. They represent a group of clinically heterogeneous diseases, including psoriasis (Pso), psoriatic arthritis (PsA), axial spondyloarthritis (AxSpA) and chronic inflammatory bowel diseases (IBD, Crohn's disease, (CD); ulcerative colitis, (UC) ) include (El-Gabalawy et al. 2010; Zink et al. 2010). Rheumatoid arthritis (RA) is the most common inflammatory rheumatic disease; the “seronegative” form in particular is often initially coded non-specifically and often progresses into another defined musculoskeletal disease. In addition to genetic predisposition and type of immune reaction, the disease entities differ in terms of the organs primarily affected (e.g.
skin, joints, intestines) and within individual entities in disease progression and severity (Cho/Feldman 2015). Various drug therapies are available to treat IMID, which aim to reduce the inflammatory activity of the disease (Högenauer et al. 2019), which can achieve symptom relief, an improvement in quality of life or even remission (Moura et al. 2015; Schöffski et al. 2007; van Mens et al. 2019).

>> Those affected therefore have an increased risk of developing another, secondary IMID parallel to a primary IMID (El-Gabalawy et al. 2010; Högenauer et al. 2019; Aletaha et al. 2019). Different organ systems can be affected, which is why cooperation between different specialist disciplines makes sense to provide optimal care for these patients (Högenauer et al. 2019; Nast et al. 2017). The occurrence of IMID-related diseases therefore represents a challenge to the interdisciplinary diagnostic and therapeutic management of these patients (Högenauer et al. 2019). IMID can lead to a significant reduction in quality of life and sometimes irreversible organ damage, which is why early diagnosis and optimal therapy control are of great importance for the further progression of the disease (Zink et al. 2010; Moura et al. 2015; Russell et al. 2011; Schneider et al. 2020).
The present study addressed the following questions:
• How does the disease progress in new cases of IMID?
• How does the course of therapy follow the course of the disease?

Based on billing data from statutory health insurance (GKV), incident patients of various IMID were identified and their disease and drug therapy progression was observed over a period of four years. The course of the disease was examined based on subsequent IMID diagnoses. In order to investigate possible differences in the course of therapy depending on the course of the disease, a distinction was made between patients with and without a subsequent diagnosis.
methodology
Database and study design
The data basis for the study was the database of the Institute for Applied Health Research Berlin GmbH (InGef). The database includes aggregated, anonymized data sets from the German statutory health insurance (GKV) from around 60 health insurance companies with around eight million insured people nationwide and contains data on drug prescriptions as well as outpatient and inpatient diagnosis and service data. The analysis was carried out on the basis of a sample of around four million GKV insured people that is representative of the German population in terms of age and gender (Andersohn/Walker 2016). A retrospective longitudinal study was carried out to describe the course of the disease and drug therapy. Examples of patients were patients with an initial diagnosis (according to ICD-10-GM) of Pso (L40.- without L40.5), RA, differentiated into seropositive (M05.-) and seronegative (M06.-).
and IBD (MC or CU; K50 or K51.‑) (Table S1) were selected in 2014 and observed over a period of four years with regard to their disease progression, the occurrence of IMID subsequent diagnoses and the course of drug therapy.

Identification of the study cohort
The study cohort included patients who were insured continuously or continuously until death throughout the entire study period (calendar years 2014–2018) and who had a documented initial IMID diagnosis in the calendar year 2014 (“index year”). All patients with at least one confirmed diagnosis of the defined entities listed in Table S1 as an outpatient diagnosis in two different quarters (M2Q) or with at least one inpatient main or secondary diagnosis during the index year were included. For inclusion in the study cohort, a previous diagnosis-free period of 365 days (“washout” period) was also defined, in which none of the following diagnoses could be present: Pso (L40.- without L40.5), PsA (L40.5 and/or or M07), RA (M05.- and/or M06.-), IBD (K50 or K51.-), AxSpA (M45.-).

Identification of subcohorts
Based on the study cohort, two subcohorts were formed within each entity, depending on whether a further IMID diagnosis occurred within the four-year follow-up period (“with subsequent diagnosis”) or not (“without subsequent diagnosis”). The following entities were taken into account: Pso, RA (seropositive or negative), PsA, IBD (MC or UC) and AxSpA (Table S2). The subcohorts were stratified by sex and age (PSO and IBD: <40 and ≥40 years, RA seropositive and seronegative: <45 and ≥45 years). The course of the disease was then described based on the number of patients with a subsequent diagnosis and the average duration from the documented initial diagnosis to the subsequent diagnosis.

Therapy classes
To describe the course of drug therapy, five therapy classes were defined: biologics (bDMARDs)/Janus kinase inhibitors (targeted synthetic (ts)DMARDs), conventional disease-modifying anti-rheumatic drugs (csDMARDs) and immunosuppressants, systemic glucocorticoids, non-steroidal anti-inflammatory drugs (NSAIDs). )/Coxibes and topical dermal therapies (Table 1, Table S3). The course of drug therapy was then described in the subcohorts based on the proportion of patients with at least one prescription of the defined therapy classes within the index year and within the four follow-up years. In addition, the average duration from the first documented diagnosis to the first prescription was determined.
Results
A total of 13,970 incident patients with one of the defined disease entities were identified in the index year (Fig. 1)
(0.33% of the insured). Of these, more than half had a documented PSO diagnosis (n = 7,109, 50.9%). Within the study cohort with a documented RA diagnosis, more patients had a seronegative than a seropositive RA diagnosis (n = 4,149, 29.7% vs. n = 598, 4.3%). Among patients with a documented IBD diagnosis (n = 2,114), UC diagnoses were more common than CD diagnoses (n = 1,194, 8.5% vs. n = 895, 6.4%). 25 patients had documentation of both CD and UC. 37.8% of the patients with an initial IBD diagnosis belonged to the younger age group (<40 years), in the other entities the proportion of patients who belonged to the younger age group was lower (Fig. 1, online appendix Table S4).

Course of the disease
During the observation period, at least one subsequent diagnosis was documented in 11.6% (n=1,623) of the incident patients - several subsequent diagnoses (multiple counts) were possible (Table S5). Of the n = 7,109 patients with an initial Pso diagnosis, a PsA diagnosis was most frequently diagnosed within the subcohort with subsequent illness (n = 471) during the four-year observation period with a proportion of 70.1%, on average after 1.2 years. documented. (Fig. 2, online appendix Table S5). Seronegative RA was documented in 35,2%. In the case of an initial seropositive RA diagnosis, the most common subsequent diagnosis observed was seronegative RA, while conversely, in the case of an initial seronegative RA diagnosis, seropositive RA was the most frequently documented subsequent diagnosis. In patients with an initial IBD diagnosis, a Pso diagnosis (43.4%) was most frequently documented within the subcohort with a subsequent diagnosis (n = 99) (Online Appendix Table S5).

Course of drug therapy
In the patients with Pso, within the subcohort with a subsequent diagnosis, there was a development over time from topical dermal therapy to therapy with systemic glucocorticoids, csDMARDs and immunosuppressants or bDMARDs/tsDMARDs (Fig. 3). Within the subcohort with a subsequent diagnosis, up to 14,2% (year 3) received a prescription for systemic glucocorticoids. In patients with RA (seropositive and seronegative), the proportion of those who received a prescription for csDMARDs or immunosuppressants reached 39.0% (patients with seropositive RA) and 57 respectively in year 1 in the subcohorts with a subsequent diagnosis. 5% (patients with seronegative RA).
Of the therapy classes considered, patients with IBD (with and without a subsequent diagnosis) most often received a prescription for csDMARDs or immunosuppressants. The proportion of these patients with a corresponding prescription reached up to 61.6% (subcohort with a subsequent diagnosis) or 50.7% (subcohort without a subsequent diagnosis) in year 4. In the subcohort with a subsequent diagnosis, between 10.1% (year 3) and 36.4% (year 4) of patients with IBD received a prescription for NSAIDs/Coxibes; in the subcohort without a subsequent diagnosis this was between 12.7% (year 1 ) and 35.6% (year 4) (Fig. 3).
In all entities considered, the proportions of patients who received a prescription for csDMARDs and immunosuppressants were higher in the subcohorts with a subsequent diagnosis than in the subcohorts without a subsequent diagnosis. bDMARDs/tsDMARDs were rarely prescribed in all entities. Over time, the proportion of patients with a corresponding prescription increased slightly, particularly within the subcohorts with a subsequent diagnosis, but was mostly below 10%. An exception here are patients with IBD with a subsequent diagnosis (year 4: 19.2%) (Fig. 3).
The average time until the drug therapy classes under consideration are prescribed as a result of the initial diagnosis is shown in Fig. 4 and Table S6.
discussion

During the observation period, only a proportion of 6.63% of subsequent diagnoses were documented in incident patients with Pso. This is in contrast to cohort studies, which show that PsA is also documented in up to 42% of patients with Pso over a period of ten years (Gladman et al. 2005; Mease/Armstrong 2014). In the subcohort with a subsequent diagnosis, PsA was most frequently documented as a secondary disease. PsA often occurs in connection with Pso; in around 80% of cases, a disease with Pso precedes or occurs at the same time (Aletaha et al. 2019; Gladman et al. 2005; Mease/Armstrong 2014). In the present analysis, the observation period was limited to the first four years after an initial diagnosis. In 330, or 4.6%, of patients with an initial Pso diagnosis, PsA was documented within this period. The low percentage of initial pso diagnoses could be due to the type of data collection, the large number of treating and coding disciplines, and incorrect coding. However, the results of this study are comparable to an analysis also based on the InGef database, in which PsA was documented in 3.4% of patients with incident psoriasis vulgaris within four years (Rech et al. 2020). Due to the high prevalence of PsA in patients with Pso, the S3 guideline for the treatment of psoriasis vulgaris recommends annual screening for PsA using a questionnaire and cooperative rheumatological evaluation (Nast et al. 2017). The results of this study also underline that cooperation between the disciplines of dermatology and rheumatology in the treatment of patients with Pso is important right from the start, as PsA can occur within the first few years after an initial Pso diagnosis (70, 1%) was observed.
The second most frequently documented case in incident patients with Pso was seronegative RA (35.2%). Since Pso is rarely associated with RA - there are reports of a prevalence of RA in prevalent patients with Pso of less than 1% - possible incorrect or misunderstood PsA diagnoses can be suspected in the RA follow-up diagnoses observed. The latter could indicate that rheumatologists are not sufficiently involved in the diagnostic process due to a lack of interdisciplinary exchange. AxSpA or IBD was only documented in 6.6% and 8.9% of patients with an initial diagnosis of Pso (Table S5). Data from various cohort studies show that up to 42% of patients with extra-articular manifestations develop IBD, uveitis or psoriasis at the same time if AxSpA is the main focus (Vander Cruyssen et al. 2007).
In patients with Pso, there was a frequent use of systemic glucocorticoids within the subcohort with a subsequent diagnosis. Each of these patients over 10% received a corresponding prescription during the observation period, although systemic glucocorticoids are not part of the S3 guideline (Nast et al. 2017) and reducing the dose or discontinuing the medication leads to an aggravation of the symptoms (“rebound effect”) can lead (Augustin et al. 2011). A possible explanation could be separate treatment of the relevant organ systems, possibly by different (specialist) doctors, which means that joint problems may be treated with glucocorticoids without taking the overall condition of the patient into account. This indicates a need for improvement in cooperation
between different specialist groups in the care of primarily dermatologically affected patients with complaints in the musculoskeletal system.
Of the n = 2,114 (15.1%) initial IBD diagnoses, a UC and a MC in the index year were coded in 25 cases. Since a diagnostic distinction can be difficult due to similar symptoms (Wehkamp 2016), this indicates an erroneous different coding, which may have been created by different doctors. In cases in which differentiation between the two diseases is not possible, colitis indeterminata (ICD code K52.3) is present; however, this coding was not part of this analysis.
The first manifestation of IBD usually occurs in young adulthood, with the incidence being highest in the group of 15 to 29 year olds (Wehkamp 2016; Johnston/Logan 2008). However, a majority of the incident patients with IBD identified in this study were in the age group ≥40 years. Since individual disease flare-ups can occur several years apart (Stallmach et al. 2012; Kucharzik et al. 2020), it is possible that the chosen methodology (365-day diagnosis-free “washout” period) was not sufficient to completely exclude prevalent patients. Compared to the other entities, subsequent diagnoses occurred less frequently in patients with an initial diagnosis of IBD. This is consistent with epidemiological studies that showed that patients with IBD tend to have lower rates of secondary IMID diseases compared to other IMIDs (Aletaha et al. 2019). This may be due to the fact that the population of these patients tends to be younger. However, a current cohort study from Germany shows that in patients with IBD and a mean disease duration of 1.3 (+/- 2.4 years), 19.4% of patients have spondyloarthritis (Rios Rodriguez et al. 2022). Here too, the lack of interdisciplinary cooperation and the lack of awareness of extra-intestinal diseases are probably the reason for the low number of subsequent diagnoses.
A significant proportion of patients with IBD received a prescription for NSAIDs/coxibes. Within the subcohorts with and without a subsequent diagnosis, the proportion of patients with a corresponding prescription rose to over a third each in the fourth year of observation (36.4% and 35.6%, respectively). The reason could be the treatment of joint pain, which can occur as extra-intestinal symptoms in patients with IBD (Arvikar/Fisher 2011). While coxibs can be used for joint pain in IBD (Preiß et al. 2014), NSAIDs are considered contraindicated due to their relapse-inducing effect; In particular, long-term treatment with NSAIDs is not recommended (Tränkner et al. 2003; Harbord et al. 2016). In this analysis, NSAIDs/coxibes were considered as an aggregate therapy class, but the results of a previous cross-sectional analysis (Leipe et al. 2023) indicate that the proportion of this therapy class was mainly driven by prescriptions of NSAIDs. This suggests that an interdisciplinary clarification with the involvement of rheumatologists may not have taken place and highlights the need for interdisciplinary therapy management for patients with IMID.
A clear majority of patients with an initial diagnosis of RA had a documented seronegative RA diagnosis. However, in the literature it is assumed that around half of patients have seropositive RA at the beginning of the disease (Myasoedova et al. 2020). Therefore, possible incorrect coding or diagnoses can be suspected here, which is why interpretation of the results is only possible to a limited extent. Therapy with csDMARDs is the recommended first-line therapy for patients with RA (Schneider et al. 2020; Fiehn et al. 2018; Smolen et al. 2019), but the proportion of patients with a corresponding prescription in the year of diagnosis fell rather low out of. Since a lack of csDMARD therapy in the diagnosis of RA would not be compatible with the valid guidelines, this discrepancy could be due to incorrect documentation.
bDMARDs/tsDMARDs were rarely used in the entities considered compared to the other defined therapy classes. However, over time, an increase in the proportion of patients who received a prescription for bDMARDs/tsDMARDs was observed, particularly within the subcohorts with a subsequent diagnosis. A possible explanation for this tendency is that, on the one hand, JAK inhibitors were not (yet) approved for all entities during the entire observation period (Walter 2019; Pfizer 2019). On the other hand, bDMARDs/tsDMARDs are generally only recommended as second-line therapy in moderate and severe disease courses and in cases of inadequate response to other conventional therapy classes (Nast et al. 2017; Kucharzik et al. 2020; Preiß et al. 2014; Smolen et al. 2019; Gossec et al. 2020). This indicates that patients with a subsequent diagnosis may have tended to have a more severe course than patients without a subsequent diagnosis.
The strength of this study is the underlying representative database, which offered the opportunity to use a large
adjusted and nationwide sample of legally insured people to authentically depict the supply situation. GKV routine data are collected under routine conditions and are therefore not subject to the distorting restrictions that can arise with primary data collection or register data, for example they are not limited to certain specialist disciplines. However, billing data depends on how coding is carried out in everyday clinical practice; for example, incorrect coding can occur. The longitudinal design provided insights into the course of the disease and therapy. However, due to data protection guidelines, the observation period was limited to four years after initial diagnosis. The descriptive, exploratory study design also did not allow for a causal interpretation of the results. In addition, the underlying database could not reflect the severity of the disease and the use of over-the-counter preparations. An individual examination of the course of the disease and therapy for the patients was not planned. In addition, the specialties of the doctors responsible for the initial and subsequent diagnoses and for drug prescriptions were not analyzed.
The results indicate a need for improvement in the cooperation between different disciplines, especially between derma- and rheumatology, in the diagnosis and therapy control of IMID-related diseases. <<