Acute and chronic nephrotoxic effects from ambulatory medications

One in ten adults in Society suffer from chronic kidney disease (CKD), a disease characterized by progressive loss of kidney function. Altered kidney function can impact on the risk/benefit of many medications. In clinical practice, some medications are contraindicated to people with advanced CKD, others require dose adjustments or strict monitoring for safety. The extent by which potentially inappropriate medications or monitoring is given to people with CKD has not been sufficiently evaluated, particularly in Swedish healthcare.

While the side effects and risks of certain drugs are well known, others remain less understood, in part because of the under-representation of people with CKD in clinical trials and the rarity of adverse kidney outcomes, which would require unrealistically large trial populations with years of intervention. Pharmacoepidemiological research in real-world settings can bridge this gap, offering a larger, more diverse cohort and longer follow-up to evaluate medication safety and efficacy in this population.

The work presented in this thesis offers new findings into the potential nephrotoxicity of medications and the risk-benefit profiles of treatments in people with CKD through the epidemiological analysis of healthcare data.

Study I evaluates the use of nephrotoxic medications in patients with CKD at stages G3-5 within routine care settings. This analysis compared two distinct healthcare systems, one in Stockholm, Sweden, and the other in Pennsylvania, United States (US). Findings indicate that approximately one in five patients was prescribed potentially inappropriate nephrotoxic drugs, which could have been substituted with safer, non-nephrotoxic options. The association of a documented CKD diagnosis with reduced odds of inappropriate medication use suggests that increasing physicians' awareness of CKD in patients, along with enhancing knowledge about drug nephrotoxicity, may help reduce the prescription of nephrotoxic drugs and prevent medication-induced kidney damage.

Study II examines adherence to guideline recommendations for laboratory test monitoring of lithium therapy among adults with bipolar disorder in Stockholm. We found that lithium therapy was frequently initiated without prior assessment of kidney function, and only a small proportion of individuals received consistent annual testing for the recommended laboratory markers, including serum lithium levels. This study highlights a significant inconsistency between routine care practices and guideline recommendations, underscoring the need for improvements in monitoring practices.

Study III compares the risks of CKD progression and acute kidney injury (AKI) among patients initiating lithium versus valproate therapy. Within up to 10 years of therapy, we observed that patients receiving lithium experienced similar absolute risks of CKD progression or AKI than patients receiving valproate.

However, elevated serum lithium levels were associated with future risks, emphasizing the need for close monitoring and lithium dose adjustment.

Study IV investigates the potential of sodium-glucose cotransporter-2 inhibitors (SGLT-2i) to lower hyperkalemia risk and improve renin-angiotensin system inhibitor (RASi) persistence in patients with type 2 diabetes. We observed that SGLT-2i users had a lower rate of hyperkalemia compared to dipeptidyl peptidase 4 inhibitors (DPP-4i) users. However, the initiation of SGLT-2i did not significantly affect the rate of RASi discontinuation among RASi users, potentially explained by low treatment persistence.

Study V aims to identify through agnostic approaches medications associated with AKI occurrence. We conducted a drug-wide association study (DWAS) analyzing the discordance of treatments received prior to AKI hospitalizations and other moments before in the life trajectory of the same patient. We used two large cohorts from Sweden and Denmark. The study identified some medications with already known AKI risks, medications that may have initiated in response to conditions that lead to AKI and other candidates that warrant further evaluation in subsequent studies.

List of scientific papers

I. Bosi A, Yunwen A, Gasparini A, Wettermark B, Barany P, Bellocco R, Inker LA, Chang AR, McAdams-DeMarco M, Grams ME, Shin JI, Carrero JJ. Use of nephrotoxic medications in adults with chronic kidney disease in Swedish and US routine care. Clin Kidney J. 2021 Oct 29;15(3):442-451. https://doi.org/10.1093/ckj/sfab210

II. Bosi A, Ceriani L, Elinder CG, Bellocco R, Clase CM, Landen M, Carrero JJ, Runesson B. Quality of laboratory biomarker monitoring during treatment with lithium in patients with bipolar disorder. Bipolar Disord. 2023 Sep;25(6):499-506. https://doi.org/10.1111/bdi.13302

III. Bosi A, Clase CM, Ceriani L, Sjölander A, Fu EL, Runesson B, Chang Z, Landén M, Bellocco R, Elinder CG, Carrero JJ. Absolute and Relative Risks of Kidney Outcomes Associated With Lithium vs Valproate Use in Sweden. JAMA Netw Open. 2023 Jul 3;6(7):e2322056. https://doi.org/10.1001/jamanetworkopen.2023.22056

IV. Bosi A, Xu Y, Faucon AL, Huang T, Evans M, Shin JI, Fu EL, Carrero JJ, Hyperkalemia and maintenance of renin-angiotensin system inhibitor therapy after initiating SGLT-2 or DPP-4 inhibitors. Nephrol Dial Transplant. 2024 Oct 9:gfae227. https://doi.org/10.1093/ndt/gfae227

V. Bosi A, Lund LP, Mahalingasivam V, Mazhar F, Christiansen CF, Sjölander A, Pottegård A, Carrero JJ. Drug use and acute kidney injury: a Drug-Wide Association Study (DWAS) in Denmark and Sweden. https://doi.org/10.1093/ckj/sfae338

Throughout this thesis, these articles are referred to by their Roman numerals.