Disease-Modifying Antirheumatic Drug Therapy of Inflammatory Rheumatic Diseases in End-Stage Kidney Disease Open Access

Disease-modifying antirheumatic drugs (DMARDs) are essential in the therapeutic management of autoimmune, inflammatory rheumatic diseases. Various disease manifestations, particularly those affecting the musculoskeletal system, can be controlled with nonsteroidal anti-inflammatory drugs (NSAIDs) and systemic glucocorticoids to the extent that patients experience significantly fewer or sometimes no subjective symptoms. These drug classes may even offer control over disease progression (glucocorticoids in rheumatoid arthritis [1] or possible progression-slowing effect of diclofenac in patients with ankylosing spondylitis [2]). Nevertheless, the main problem associated with NSAIDs/glucocorticoids is adverse side effects. Glucocorticoid-induced side effects invariably manifest in all individuals treated with inadequately high doses over extended periods. On the other hand, DMARDs reduce disease progression with varying degrees of success and generally have better tolerability. However, specific risk factors associated with different substance groups may contraindicate their use (e.g., anti-TNF-alpha in cases of severe heart failure) [3]. Additionally, their onset of action is often somewhat delayed. Currently, three main DMARD groups are distinguished: conventional (cDMARDs), biologic (bDMARDs), including corresponding biosimilars (bsDMARDs), and targeted synthetic (tsDMARDs).

The assessment of excretory kidney function is a routine requirement in the context of pharmacological therapy whether for rheumatological or other indications. A potential decline in kidney function can lead to drug accumulation and, consequently, a higher risk of side effects, depending on the substance. From a nephrological perspective, two types of kidney disease are associated with reduced excretory function: acute kidney injury [4] and chronic kidney disease (CKD) [5]. In the terminal stage of the latter, referred to as CKD 5D, numerous types of drugs might accumulate and causes side effects. In CKD 5D patients, managing rheumatic disease is also becoming increasingly complex. Due to fear of side effects and possibly a lack of knowledge about the use of DMARDs in the terminal stage of CKD, affected individuals may occasionally or persistently be undertreated. This article aimed to discuss the applications of DMARDs in CKD stage 5D.

The current article is a narrative review. References were sourced from the following databases: PubMed, Web of Science, Cochrane Library, and Scopus. The search period spanned from 1975 to 2024.

The use of DMARDs has two primary objectives: to prevent structural changes in the musculoskeletal system and internal organs and to facilitate the rapid tapering of systemic glucocorticoid therapy. However, this objective is not universally applicable, as certain conditions, such as ankylosing spondylitis and psoriatic arthritis (PsA), do not consistently respond to glucocorticoids [6]. The topic of DMARDs in inflammatory rheumatic diseases is too broad to be thoroughly addressed in this context. On one hand, autoimmune-mediated diseases with potential DMARD indications are numerous and heterogeneous in terms of etiopathogenesis and clinical manifestations. On the other hand, the arsenal of therapeutic options, especially biological treatments, has fundamentally expanded over the past 20 years. Focusing solely on systemic lupus erythematosus, over 25 new therapeutics will be utilized in phase 2 and 3 clinical trials by the year 2029 (EULAR Congress presentation, June 2024, Vienna). For the most up-to-date recommendations on the use of DMARDs in various types of rheumatic diseases, please refer to the respective guidelines. Generally, the following groups of DMARDs can be identified: cDMARDs, bDMARDs, including their corresponding biosimilars (bsDMARDs), and tsDMARDs. Patients with end-stage kidney disease (ESKD) are noticeably less frequently included in prospective randomized or observational therapy studies. In fact, in many studies, CKD stage 5D is often considered an exclusion criterion.

CKD stage 5D defines ESKD [5]. Stage 5 is diagnosed when the eGFR falls below 15 mL/min, typically necessitating KRT. The most common cause worldwide is now type 2 diabetes mellitus [7], with diabetes prevalence rising across all regions for the past 30 years [8]. In Central Europe, approximately 15% of adults are affected by CKD [9]. The decline and, in many cases, the eventual loss of excretory kidney function in ESKD significantly impacts the pharmacokinetics of numerous medications. Recommendations for dosage adjustments of various drugs are available online and in corresponding professional information. However, for newer therapeutics, there is often no experience for advanced stages of CKD, specifically stage 5D, as this stage frequently serves as an exclusion criterion for study participation.

Methotrexate is the first-line DMARD for confirmed rheumatoid arthritis [10]. Additional areas of application include PsA [11], uveitis [12], systemic lupus erythematosus [13] and Sjögren’s syndrome [14], systemic sclerosis and mixed connective tissue disease [15], idiopathic inflammatory myopathy [16], adult-onset Still’s disease [17], large vessel vasculitis and polymyalgia rheumatica [18], ANCA-associated vasculitides [19], and sarcoidosis [20]. Methotrexate is contraindicated below an eGFR of 30 mL/min due to the high risk of accumulation [21]. The elimination half-life of systemically administered MTX has been reported to range from 100 to 750 min [22]. However, Janknegt and colleagues [23] observed longer elimination periods in patients on peritoneal dialysis, with times extending up to 29 h. Under these conditions, there is naturally a significantly higher risk of undesirable complications, particularly bone marrow suppression. Cheung et al. [24] reported a severe pancytopenia in a dialysis-dependent patient undergoing low-dose MTX therapy for PsA. The issue of MTX myelotoxicity during hemodialysis therapy was also emphasized in a review by Al-Hasani and Rousseau [25]. Bone marrow suppression can occur rapidly and even at relatively low doses of MTX. Liu and colleagues [26] reported a severe case of leukopenia in a patient undergoing peritoneal dialysis who had received only three weekly doses of the medication at 7.5 mg. In 2002, Basile and colleagues [27] discussed whether low-dose MTX therapy is feasible under hemodialysis. The authors highlighted the prolonged persistence of MTX in the body (up to 20 days after a single 5 mg dose) and emphasized that MTX intoxication cannot be treated with hemodialysis, hemoperfusion, or plasmapheresis. Even in recent times (2023), severe pancytopenias have been reported in patients treated with MTX undergoing hemodialysis [28]. These case reports also highlight that MTX is evidently still being used in clinical practice for individuals with significantly impaired kidney function. In summary, the use of the medication in CKD stage 5D is strongly discouraged.

The dihydroorotate dehydrogenase inhibitor leflunomide is primarily used to treat rheumatoid arthritis as an alternative to methotrexate [29], with combinations of these substances also being possible. Another indication for its use is PsA [30]. The substance is already converted into its primarily active metabolite, teriflunomide, during intestinal absorption [31], and this metabolite can remain detectable in the body for up to 4 weeks. Its excretion occurs mainly through the stool, and elimination can be accelerated by administering cholestyramine [32]. However, leflunomide is not completely converted into teriflunomide. Approximately 90% of a single dose is eliminated, with nearly half excreted through urine (as leflunomide and oxolinic acid derivatives of teriflunomide) and the other half eliminated in feces (mostly as teriflunomide) [33]. The prescribing information for leflunomide does not list mild kidney impairment as a contraindication for its use. However, for moderate to severe kidney excretory dysfunction, the use of leflunomide is not recommended due to a lack of sufficient experience. Beaman and colleagues [34] reported the case of a 65-year-old dialysis-dependent woman undergoing treatment with leflunomide at a daily dose of 10 mg for rheumatoid arthritis. The concentrations of the active leflunomide metabolite A771726 (teriflunomide) were consistently within the low therapeutic range and did not differ between the times before and after hemodialysis. In 2013, Bergner et al. [35] published pharmacokinetic data on teriflunomide under hemodialysis. It was found that the procedure itself had little impact on the serum levels of the substance. Consequently, it was concluded that leflunomide can be used in patients with end-stage renal disease without dose adjustment. A similar conclusion was published by Russo et al. in 2013 [36] but in the context of peritoneal dialysis. In summary, it can be assumed that the use of leflunomide in patients undergoing KRT is medically acceptable and relatively safe. Therefore, the usual dosages should also be acceptable (between 10 and 20 mg once a day).

Although mycophenolic acid (MPA) is primarily used as an immunosuppressant in organ transplantation, there are also rheumatological indications. The substance can be used for the treatment of lupus nephritis classes III and IV, as confirmed in the latest guidelines from KDIGO (2024 [37]) and EULAR (2023 [13]). MPA also holds some significance in systemic sclerosis [38] and occasionally in rarer conditions such as eosinophilic fasciitis [39]. Naturally, only case reports exist for these instances. More than 90% of the orally absorbed amount is eliminated renally, primarily in the glucuronidated (conjugated) form [40]. Officially, doses of 2 × 1,000 mg should be avoided below an eGFR of 25 mL/min. Surprisingly, there are few references regarding the pharmacokinetics of mycophenolic acid (MPA) in the dialysis-dependent stage of acute or CKD, despite numerous studies on the use of MPA in kidney transplant patients. Zanker and colleagues [41] reported on 8 patients with post-transplant acute kidney injury, in whom serum levels of both MPA and its conjugated isoform, MPA glucuronide (MPAG), were measured. While MPA levels remained stable, MPAG levels increased as excretory kidney function decreased. Dialysis effectively lowered MPAG levels. No unwanted side effects of MPA were reported. Morgera et al. [42] presented data on MPA and MPAG levels in peritoneal dialysis patients after kidney transplantation. Similar to Zanker’s study, lower eGFR was associated with higher MPAG concentrations, and the MPA metabolite was also detectable in peritoneal ultrafiltrate. An interesting study published in 2017 [43] evaluated the capacity of MPA to preserve residual kidney function in 60 individuals undergoing peritoneal dialysis. The medication was administered over a 12-month period and was associated with better preservation of kidney function, demonstrating good tolerability. In summary, administering MPA to patients requiring KRT is likely safe, as the conversion from MPA to MPAG is maintained, and MPAG can be eliminated through both hemodialysis and peritoneal dialysis. The dosages of the preparations should be equivalent to those used for individuals with healthy kidneys, ranging from 1,000–1,500 mg daily, administered in two equal doses.

Sulfasalazine (SSZ) is used to treat rheumatoid arthritis [44], either as monotherapy or in combination with methotrexate [45]. Another indication is ankylosing spondylitis when peripheral joints are affected [46]. However, the nearly always present sacroiliitis in AS does not benefit from the use of this substance [47]. SSZ is poorly absorbed, with an absorption rate of only 3–12%. Its elimination half-life ranges from 5 to 10 h. In the colon, most of the SSZ is broken down into 5-aminosalicylic acid and sulphapyridine. The latter is almost completely absorbed and is eliminated through hydroxylation, glucuronidation, and polymorphic acetylation [48]. At this point, a quote from the professional information on SSZ is provided (https://labeling.pfizer.com/ShowLabeling.aspx?id=2088): “For patients with impaired liver and/or kidney function (especially elderly patients), the maximum and duration dose should not exceed 2–3 film-coated tablets per day.” Inami and colleagues [49] conducted a pharmacokinetic study on a dialysis patient undergoing SSZ therapy. The study revealed no significant elimination of the substance through dialysis. However, the AUC0>infinity for the metabolites 5-aminosalicylic acid and sulfapyridine were higher than in healthy individuals. This parameter typically reflects the overall exposure to a drug. A dose reduction of SSZ in dialysis patients was concluded. Akiyama et al. [50] published pharmacokinetic data on 8 hemodialysis patients with rheumatoid arthritis. At a dose of 500 mg, the AUC for SSZ was comparable to data from phase I studies of the substance. The concentration of sulfapyridine was higher than in non-dialysis individuals, but not in the toxic range.

The substance is an essential therapeutic for systemic lupus erythematosus in all stages of the disease [51]. Additionally, it can be used for mild forms of rheumatoid arthritis [52] and in the therapy of mixed connective tissue disease [53]. The terminal elimination half-life of hydroxychloroquine (HCQ) reaches 40 days, as the medication gradually transitions from deeper tissue compartments into the bloodstream [54]. Even if dialysis therapy could eliminate 100% of the substance available in the blood during a 5-h treatment period, it would have only marginal effects on redistribution from the tissue. The professional information (https://www.heumann.de/fileadmin/user_upload/produkte) on HCQ states: “Caution is advised for patients with impaired kidney or liver function. The dosage may need to be reduced accordingly.” More specific information is not provided at all. First, a case report from 1987 should be mentioned [55]. The authors treated a hemodialysis patient with sarcoidosis-associated hypercalcemia with HCQ for 24 weeks, resulting in the resolution of the hypercalcemia. No significant side effects of the therapy were observed. In contrast, El-Solia and colleagues [56] reported on a 25-year-old SLE patient undergoing peritoneal dialysis. A severe hypoglycemic episode only resolved after discontinuing HCQ therapy. Broder and colleagues [57] addressed a crucial aspect of SLE therapy in CKD stage 5, specifically the nature of long-term treatment for the underlying rheumatic disease. They evaluated a cohort of 2,654 patients. Among this group, 220 individuals received HCQ as monotherapy, and 509 patients received HCQ in combination with glucocorticoids. For the majority of patients, HCQ was either never used or discontinued upon reaching CKD 5D. This, however, resulted in an average disproportionately high glucocorticoid dose. The rare use of HCQ in CKD 5D patients with SLE highlights the unclear data and explains the title of a 2022 review article: “Shall We Use Hydroxychloroquine in Hemodialysis Patients?” [58]. It would currently be presumptuous to recommend the use of HCQ in stage 5D of CKD.

Azathioprine (AZA) is a medication that has been available for decades and has a wide range of applications beyond inflammatory rheumatic diseases. It is used for conditions such as rheumatoid arthritis ([44] – though it is rarely used for this indication anymore), ANCA-associated vasculitis (AAV) [19], systemic lupus erythematosus [13], sarcoidosis [59], Behçet’s disease [60], and others. By using the search terms “azathioprine” and “dialysis” or “azathioprine” and “dialysis” and “dosing”, no specific references were found as of June 2024. Once again, we must refer to the prescribing information (https://www.ratiopharm.de/assets/products): “Since impaired renal function can lead to slower elimination of AZA and its metabolites, it may be advisable to reduce the initial dose in patients with renal impairment. Patients should be monitored for dose-dependent side effects.” The only binding requirement is the recommendation to lower the initial dose. Monitoring every AZA therapy involves quantifying transaminases, cholestasis parameters, and conducting blood count analyses. If these results are unremarkable, which would generally argue against accumulation, the use of AZA in confirmed indications at CKD stage 5D appears acceptable (1.5–2 mg/kg daily, two equivalent doses).

At this point, it is important to provide some additional observations regarding the therapy control examinations for patients with CKD 5D who are undergoing cDMARD therapy. As indicated in the text, leflunomide, SSZ, MPA, and AZA are considered relatively safe medications. The recommended control examinations are analogous to those performed on individuals with healthy kidneys. For instance, AZA necessitates blood count checks and the assessment of transaminases and cholestasis parameters. The prescribing information for AZA specifies that monitoring should occur every 2–3 months during treatment, provided the medication is well tolerated. In the context of cDMARDs, it is reasonable to consider slightly more frequent checks for CKD 5D patients, as this approach appears to be safe and practical. For instance, monthly measurements could be performed in conjunction with the nephrological parameters (calcium, phosphate, etc.) that are already indicated. Although there are no binding guidelines in this regard, it is recommended that therapeutic attention be increased.

The first class of biologic DMARDs available in rheumatology was TNF-alpha inhibitors. This group includes adalimumab, certolizumab pegol, etanercept, golimumab, and infliximab. The most important indications for these drugs are rheumatoid arthritis [10], PsA [11], and ankylosing spondylitis [47]. They have also been used in less common conditions such as giant cell arteritis [61]. Biosimilars (bsDMARDs) are now available as alternatives to each individual TNF-alpha inhibitor. The available data on anti-TNF-alpha therapy in patients with stage 5D CKD are very limited. The technical information on adalimumab does not include any dosing recommendations for patients with CKD 5D, as the medication has not been tested in this population. The same applies to certolizumab, golimumab, and infliximab. In contrast, etanercept therapy does not require dosage adjustment for individuals with ESKD (25 or 50 mg weekly). A 2012 report summarized the experience of administering adalimumab to ankylosing spondylitis patients requiring peritoneal dialysis [62]. Additional systematic data are unavailable as of June 2024.

Systematic data on anti-IL-6, abatacept, and anti-IL-17 are lacking. The decision to use either any of these substances in CKD stage 5D must be made on an individual basis. Patients must be informed about the insufficient experience with bDMARDs in those requiring dialysis, and written consent must be obtained.

The available substance is ustekinumab, which is used in treating patients with PsA and plaque psoriasis [63]. Compared to most anti-TNF-alpha preparations, ustekinumab has not been tested in patients with ESKD, and therefore, official dosing recommendations are unavailable. Spathakis and colleagues [64] reported on a 47-year-old patient with refractory Crohn’s disease who was successfully treated with anti-IL12/23 while permanently requiring dialysis therapy. The drug was well tolerated. Ustekinumab is therefore most likely a therapeutic option for CKD stage 5D patients with established PsA (45 mg administered subcutaneously, a subsequent 45 mg dose 4 weeks later, thereafter, additional doses every 12 weeks).

The substance inhibits BLyS-induced activation of both autoreactive and non-autoreactive B cells. Its most important clinical application is systemic lupus erythematosus. Both the latest version of the “KDIGO 2024 Clinical Practice Guideline for the Management of Lupus Nephritis” [37] and the “EULAR Recommendations for the Management of Systemic Lupus Erythematosus: 2023 Update” [13] identify belimumab as one of four options for remission induction in severe lupus nephritis. Officially, belimumab does not require dosage adjustment, even in severe cases of excretory kidney dysfunction, although experience in CKD stage 5 is limited. In 2020, Binda et al. [65] reported on a peritoneal dialysis patient who continued belimumab therapy, initiated before transplantation, even after receiving a kidney transplant. Two years later, Liu and colleagues [66] published their experience with belimumab administration in 7 SLE patients on dialysis. The antibody treatment improved immunologic parameters in all subjects, and only one pulmonary infection was diagnosed. Although no large groups of dialysis-dependent patients have been treated with belimumab, its use in CKD stage 5D still appears to be justifiable (10 mg/kg body weight, initially twice at 2-week infusion intervals, then every 4 weeks).

Anti-CD20 therapy is well established for severe cases of rheumatoid arthritis [10], and it is also quite effective in AAV. In AAV, rituximab can be used for both remission induction and maintenance therapy [19]. According to the product information, no pharmacokinetic data are available for patients with liver or kidney insufficiency. In clinical practice, the use of anti-CD20 therapy in patients with impaired renal function, including those requiring dialysis, is common worldwide. This practice is supported by recommendations for managing AAV, which advocate the use of Rituximab for remission induction [19], particularly in cases with generalized disease involving pulmonary-renal syndrome, including those requiring dialysis. For this reason alone, an extensive search for references that have systematically examined the use of anti-CD20 in dialysis-dependent patients is almost unnecessary (different dosing regimens, e.g., rheumatoid arthritis 1,000 mg i.v., two doses, 2 weeks apart, every 6 months).

bDMARDs have significantly expanded the range of effective therapeutic options for inflammatory rheumatic diseases, marking a transformative advancement in rheumatology since the early 2000s. However, it is important to recognize that most of these treatments, primarily consisting of specific antibodies, come with substantial financial costs. The advent of biosimilars, marked by the expiration of patent protection for a growing array of active substances, has led to a notable decline in therapy costs. Nevertheless, the cost of adalimumab remains approximately USD 3,000 for a 3-month regimen. Given the absence of official approval for numerous bDMARDs in CKD stage 5D, patients may need to seek coverage from their respective health insurance companies. It is crucial to note that chronic dialysis is inherently associated with exorbitant financial burdens, stemming from not only the cost of dialysis therapy itself but also the necessity of a multitude of medications for nephrological and cardiological indications, along with the costs of medical examinations and treatments resulting from the complications of CKD in the long term. Consequently, a meticulous evaluation must be conducted in each individual case to ascertain the indispensability of bDMARDs.

JAK kinase inhibitors are a relatively new class of substances represented by the following drugs (in alphabetical order): baricitinib, filgotinib, tofacitinib, and upadacitinib. These substances inhibit the autophosphorylation of cytokine receptors, thereby preventing adequate signal transduction following cytokine binding to the receptor. Autophosphorylation is normally catalyzed by JAK kinases. The primary rheumatological indications for these drugs are rheumatoid arthritis [67] and PsA [68]. In 2014, Krishnaswami et al. [69] published a study on 12 dialysis patients who were treated with 10 mg of tofacitinib daily. The maximum plasma concentration did not differ from those with normal kidney function, but the elimination half-life was significantly reduced. Palmer and colleagues [70] reported on a COVID-19 patient who received baricitinib at 2 mg daily while undergoing continuous venovenous hemodialysis. The maximum plasma concentration of the drug was comparable to that in subjects without kidney disease, but the total clearance was significantly reduced. As of June 2024, specific references on filgotinib and upadacitinib are not available. Considering that JAK kinase inhibitors are relatively new substances and given the delayed elimination under dialysis conditions, their use in patients with CKD 5D is likely not advisable. Despite this, the product information for tofacitinib advises continuing the medication even if dialysis has been initiated, recommending only a dose reduction to 5 mg twice daily. Table 1 summarizes main studies discussed in the text. Table 2 summarizes all DMARDs discussed and their potential application in CKD stage 5D.

The systematic data on the use of DMARDs in the terminal stage of CKD are limited. This applies not only to newer agents, specifically biologic DMARDs, but also to older, conventional drugs. Among the conventional DMARDs, the following substances may likely be used with relatively low risk in dialysis patients: leflunomide, MPA, SSZ, and AZA. From the group of biologics, belimumab and rituximab may be applicable, as indicated by their approved uses. Other biologics have not been systematically studied in dialysis-dependent patients, and there are currently (almost) no renal indications for anti-TNF-alpha or other biologics beside belimumab and anti-CD20. Given that JAK kinase inhibitors are relatively new substances, caution should generally be exercised in cases of reduced kidney function. In our view, there are no significant differences between hemodialysis and peritoneal dialysis in terms of the safety or applicability of DMARDs.

The authors declare that they have no conflicts of interest.

No funding was provided for the study.

Daniel Patschan wrote the article. Igor Matyukhin, Friedrich Stasche, and Oliver Ritter searched for references. Susann Patschan designed the article and assisted in reference collection and writing.