Lactic acidosis associated with metformin in diabetic patients with chronic kidney disease: a multicentre case-control study using electronic health records

Background: Several studies have assessed the risk of lactic acidosis with metformin use. However, data of this association in patients with renal impairment are still scarce and controversial. Our aim was therefore to assess the association between metformin and lactic acidosis in Spanish type 2 diabetic patients with chronic kidney disease. Methods: A case-control study (ALIMAR-C2) was performed using the electronic health records from hospitals linked to their corresponding primary healthcare regions. The cases were adult ( ≥ 18 years) diabetic patients with chronic kidney disease, admitted to seven Spanish hospitals from 2010 to 2016. Ten controls ( ≥ 18 years, diabetic patients with chronic kidney disease) per case were selected from the population within the same primary healthcare region of the hospital cases. The patients’ hospital health records were linked to their corresponding primary healthcare information. Analyses included multivariable logistic regression and adjustment for potential confounders. Results: Our study included 126 cases and 1,260 matched controls. The current use of metformin and administration at high doses (>2g) were associated with lactic acidosis (adjusted OR: 1.92, 95% CI: 1.21-3.03; OR: 3.13, 95% CI: 1.63-6.01, respectively). The estimated case fatality rate was 46.8% (95% CI: 38.3-55.5%). An increased risk of lactic acidosis was observed In patients with mild to moderate renal impairment (OR: 3.41, 95% CI: 1.48-7.85). As an unexpected nding, diuretic drugs use was also associated with lactic acidosis (OR: 2.73, 95% CI: 1.67 - 4.46). Conclusions: Metformin was associated with an increased risk of lactic acidosis in patients with type 2 diabetes mellitus and chronic kidney disease. New data is needed to conrm the association between diuretic drugs and lactic acidosis in this group of patients.


Background
Metformin is the initial drug treatment for adults with type 2 diabetes mellitus (DM2) but has been associated with lactic acidosis (LA), a very rare and life-threatening adverse event, especially in patients with renal disease [1,2]. The safety evidence supporting its use in these patients remains controversial and scarce. In 2016 the regulatory authorities of the Food and Drug Administration (FDA) and European Medicines Agency (EMA) carried out safety reviews, concluding that metformin can be used in patients with mild to moderate kidney impairment (stage 3a and stage 3b; estimated glomerular ltration rate [eGFR] 30-59 mL/min) but still warned against its use in cases of severe renal impairment (stage 4 and stage 5; eGFR<30mL/min) [3,4]. Additionally, a recent large cohort study showed that metformin does not appear to increase the risk of lactic acidosis in such patients while metformin should be used with caution in patients with severe renal impairment [5]. It is noteworthy that, prior to the aforementioned regulatory authorities' guidelines, observational studies had already shown an increased incidence of LA in patients exposed to metformin in parallel to the degree of impairment of renal function [6], as well as an increased risk of LA in patients with eGFR <60 mL/min, mainly due to the higher risk in patients with eGFR <45 mL/min [7].Prescribing metformin for people with renal impairment continues to be a matter of debate today, especially in patients with severely reduced kidney function. As a consequence, there is still considerable variation between the different international prescribing guidelines for metformin [8]. The aim of this case-control study (ALIMAR-C2 [Riesgo de Acidosis Láctica asociado al uso de MetforminA en pacientes diabéticos tipo 2 con enfermedad Renal crónica moderada-severa: estudio de Casos y Controles]), using electronic health records from hospitals linked to their corresponding primary healthcare regions, was therefore to assess the association between metformin and LA in Spanish patients with DM2 and reduced kidney function.

Study design and setting
We conducted a multicentre case-control study using data from 2010 to 2016, obtained from the electronic health records of seven hospitals and their primary healthcare institutions geographically located in two regions of Spain (Madrid and Catalonia). The study methodology has been described elsewhere [9]. The administrative, clinical and laboratory records from the data warehouses (DWs) of Catalan hospitals were linked with the primary healthcare data compiled in SIDIAP (Information System for Research in Primary Care) which contains pseudonymised clinical information from all the primary healthcare centres of the Institut Català de la Salut (ICS) [10]. Bellvitge University Hospital is the only hospital in which a DW integrates both clinical practice and primary healthcare data. In Madrid, the clinical data obtained from different sources and information systems integrated within the hospitals were linked to the Primary Healthcare Electronic Health Record (AP-Madrid) which contains data from all the primary healthcare centres within the Servicio Madrileño de Salud (SERMAS).

Study population
We identi ed all patients admitted to the hospital with a diagnosis of LA (pH <7.35 and plasmatic lactic acid concentration >5 mM/L within the rst 24 and 72 hours after admission, respectively) from 2010 to 2016. The date of admission was used as the index date. The inclusion criteria were as follows: (1) at least 18 years of age, (2) hospital or primary healthcare diagnosis of DM2 before the index date, (3) chronic kidney disease (CKD) stage 3a (mild-moderate), 3b (moderate-severe) or 4 (severe) of the Kidney Disease Improving Global Outcomes (KDIGO) classi cation [11] during the 2-year period before the index date (excluding the previous 2 weeks), taking into account data from the primary healthcare database, and (4) availability of any information recorded on the primary healthcare database within a 1-year period before the index date. We excluded cases with: (1) diabetic ketoacidosis during the current in-hospital stay; (2) hospital or primary healthcare diagnosis of type 1 diabetes mellitus, human immunode ciency virus disease or solid organ transplant before the index date; (3) hospital or primary healthcare diagnosis of malignant neoplasm (except skin cancer other than melanoma; including pheochromocytoma) within a 5-year period before the index date [9]. In Catalonia, patients not registered in the hospital referral area were also excluded.
Furthermore, people assigned to the hospital's primary healthcare region were chosen for the control group and matched in a ratio of 10:1 by age, gender, CKD stage and year of admission. They were at least 18 years old, had DM2 diagnosed before the index date and a CKD stage as de ned for the cases.
Additionally, they had information recorded on the primary healthcare database within a 1-year period before the index date. We excluded controls with: (1) type 1 diabetes mellitus, human immunode ciency virus disease or a solid organ transplant before the index date; (2) malignant neoplasm (except skin cancer other than melanoma; including pheochromocytoma) within a 5-year period before the index date, and (3) patients not resident in the area of study.

Measurements
The hospital databases provided information on the patients' characteristics, including age and gender, hospital course data (admission date, in-hospital death, admission to critical care unit), laboratory test data (values and dates for lactic acid and haemoglobin concentration, and pH). The primary healthcare databases provided laboratory test data (values and dates for serum creatinine and haemoglobin concentration), and information on the drugs prescribed (anatomical therapeutic chemical [ATC] codes; prescription dates for metformin, other non-insulin antidiabetic drugs [NIADs], insulin, diuretics [high ceiling: furosemide, torasemide; low ceiling: hydrochlorothiazide, chlortalidone, xipamide, indapamide; potassium-sparing diuretics: spironolactone, eplerenone], renin-angiotensin system [RAS] inhibitors, and non-steroidal anti-in ammatory drugs [NSAIDs]; National Drug Code [NDC]; and prescribed posology for metformin). In the case of drug combinations, each drug was classi ed in its corresponding ATC group with a record of the drug's dosage in the combination. Additionally, diagnosis dates and codes (International Classi cation of Diseases, 9 th and 10 th revisions [ICD-9, ICD-10], and International Classi cation of Primary Care, 2 nd revision [ICPC-2]) were obtained from both the hospital and primary healthcare databases. Renal function was estimated using serum creatinine data between 2 years and 2 weeks before the index date. The eGFR was calculated using the CKD-Epidemiology equation and stage according to the KDIGO classi cation (≥90, 60-89 45-59, 30-44, <30 mL/min/1.73m 2 ) [12]. In the case of eGFR estimates resulting in different CKD stages for an individual patient across the 2-year period, the worst CKD stage closest to the index date was assigned to this patient when this was not followed by a better stage. Detailed information regarding the variables has been described previously [9].

Exposure de nition
Exposure to metformin, other NIADs and insulin was de ned as current use (prescription within a 30 dayperiod before the index date) or global use (prescription within a 365 day-period before the index date). We would like to point out that global use was de ned in the statistical analysis plan.
The length of the exposure was de ned as the time between the start and end prescription dates. A gap in drug prescription of ≤30 consecutive days was not considered as discontinued exposure. The prescribed daily dose of metformin was calculated according to the posology recorded by the prescriber and the strength corresponding to the NDC, and was categorised into <1g, 1-2g, and >2g. Exposure to diuretics, RAS inhibitors and NSAIDs were de ned as prescriptions during the 30 day-period before the index date.

Statistical analysis
Baseline characteristics were summarised for cases and controls using standard descriptive statistics and a descriptive comparative analysis was carried out.
Conditional logistic regression was used to control for matches on age, gender, renal stage and year of index date. Crude and adjusted odds ratios (OR) with 95% con dence intervals (CI) were estimated to assess the risk of LA associated with metformin. The following drugs competed with metformin: alpha glucosidase inhibitors, dipeptidyl peptidase 4 (DPP-4) inhibitors, insulins and analogues, other blood glucose lowering drugs, sulfonylureas, thiazolidinediones, glucagon-like peptide 1 (GLP1) receptor agonists, NSAIDs, potassium-sparing diuretics, high ceiling diuretics, low ceiling diuretics and RAS inhibitors.
A stepwise procedure was used to select the baseline covariates to be included in the adjusted models.
The variables selected were: alcohol use, liver disease, acute myocardial infarction, arterial peripheric arteriopathy, heart failure, chronic respiratory disease, dementia, seizures, gastroenteritis and dehydration.
Subgroup analyses were performed according to the daily dose of metformin, current and global use, and also the CKD stage.
Additionally, the overall case fatality rate of LA as well as the case fatality rate strati ed by CKD stage were calculated from the number of deaths among cases and the total number of cases.
The possibility of detection bias was studied by analysing the frequency of determination of plasmatic lactate levels in patients with metabolic acidosis according to the status of metformin exposure. This analysis was performed with data from two of the participating hospitals in a sample of episodes of urgent hospital admission with pH <7.35 during the rst 24 hours. An OR with its 95% CI for each hospital was provided.
All statistical analyses were performed with R statistical package version 3.6.0.

Results
Among the 1,619,481 patients admitted to hospital between January 2010 and December 2016, 126 cases were identi ed ( Figure 1). LA cases showed statistically signi cant higher morbidity than controls in acute myocardial infarction (23.8% vs 12.6%), peripheral arterial disease (24.6% vs 15.4%), heart failure (38.1% vs 26.7%), seizures (4% vs 1.3%), dehydration (6.3% vs 1.8%), chronic alcoholism (13.5% vs 6.2%) and acute alcohol intoxication (2.4% vs 0%). Table 2 shows the OR for association between LA and different drugs. Compared with no use, current use of metformin was associated with a higher risk of LA (OR: 1.92, 95% CI: 1.21-3.03). A dose-effect trend was observed, although the increase in risk was statistically signi cant only for the higher dosage category (>2g) ( Table 3). Similar results were found for global use of metformin (see Additional File 2: Table 2 and Additional File 3: Table 3).
Subgroup analysis according to the CKD stage showed that current use of metformin was associated with LA in patients with mild-moderate renal impairment (stage 3a; OR 3.41, 95% CI 1.48-7.85; Table 4). However, this association was not found in patients with severe renal impairment (stage 4).
In the analysis by different drugs (Table 2 and Additional File 2: Table 2), neither the current use nor global use of other hypoglycaemic drugs, nor current use of NSAIDs or RAS inhibitors was associated with LA. However, the current use of diuretics was associated with LA (OR: 2.73, 95% CI: 1.67 -4.46). The risk of LA by a classi cation of the diuretics is shown in Table 2. The subgroup analysis according to CKD stage showed an association with a higher risk of LA for high ceiling diuretics in CKD stage 3b, for low ceiling diuretics in CKD stage 3a, and for potassium sparing diuretics in CKD stages 3b and 4 ( Table  5).
Acidosis lactic detection bias was not found in any of the two hospitals studied (Hospital 1 (Catalonia): OR 0.95, 95% CI 0.70-1.28; Hospital 2 (Madrid): OR 1.07, 95% CI 0.69-1.67). Patients with metabolic acidosis admitted via the emergency department who were metformin users did not show a higher risk of their plasmatic lactate levels being determined than non-metformin users.

Discussion
To our knowledge, this multicentre case-control study on LA is one of the largest studies published to date and it is noteworthy that LA was de ned based on objective parameters (laboratory tests: pH value and lactic acid level). The use of metformin was associated with a higher risk of LA in diabetic patients with chronic kidney disease, especially in patients on a higher dose of metformin. Unexpectedly, LA was also associated with diuretics use.
The use of metformin for DM2 in patients with reduced kidney function remains controversial and data on safety in this population is scarce. The prescription of metformin for people with renal impairment continues to be a matter of debate, especially with this type of patient [3,4]. Our ndings have shown that a current use of metformin increases the risk of lactic acidosis (OR: 3.41, 95% CI: 1.48-7.85) in patients with mild-moderate CKD (stage 3a; eGFR: 45-59 mL/min). These ndings are partly in line with the results described in a UK population cohort study on the same group of patients (adjusted HR: 6.06, 95% CI: 1.37-27.10) [7]. Conversely, metformin use has not been associated with an increased risk of LA in other studies [5] [13] [14]. It should be noted that recommendations on the use of metformin in patients with CKD are based on observational studies, which could represent a limitation for its endorsement [3,4]. This nding could possibly be explained by a channeling bias. In consequence, patients with moderatesevere renal impairment were not candidates to receive metformin. In our setting, patients with severe CKD could be less exposed to this drug. [14]. This might explain the mixed opinions regarding metformin use in clinical guidelines [8]. Our ndings suggest that a higher dose of metformin is associated with a higher risk of LA, in line with other ndings that have shown an increased risk with high daily exposure to metformin in patients with CKD (adjusted HR: 13.0, 95% CI: 2.36-72.0) [7]. In brief, based on our results, the use of metformin in DM2 and CKD patients should be carried out with caution.
Current diuretics use was unexpectedly found to be associated with a higher risk of LA and this association was also observed in the analysis by diuretic type. To our knowledge, this is the rst time such a nding has been reported. In order to nd an explanation, we performed a post hoc analysis excluding patients with heart failure (38% of our analysis population), given that these patients are usually treated with diuretics. The risk association between diuretics and LA was still present (see Additional File 4: Table 4). Another factor studied in order to explain this association was the effect of dehydration as diuretics may negatively affect renal performance, especially in dehydrated patients. However, only 8 patients suffered from dehydration. Such a low number of patients with dehydration cannot contribute to an explanation of the association found between diuretics and LA. In fact, although we believe this nding may be misleading, future studies are required to study this association.
Half our cases died. According to the evidence available [15], this high morbidity rate is to be expected. The low admission rate to critical care units is consistent with the admission criteria; the patients who were not admitted were older and had a worst diagnosis than those who were admitted (see Additional File 1: Table 1). This study has some limitations. First, there were differences regarding the data stored in the electronic databases. Likewise, different diagnosis coding systems were used, and we cannot guarantee that all diagnoses were suitablygathered (such as heart failure). However these are inherent limitations to studies based on databases. Second, we had not taken into account changes in eGFR over time. However, only one recent study has taken this methodological consideration into account [13] and, moreover, earlier evidence has demonstrated consistency without its use. Third, exposure could be overestimated as the analysis was carried out according to drug prescriptions and not to the dispensation or actual intake of the drug. Nevertheless, this situation would be the same for both cases and controls and, consequently, the potential overestimation is not expected to have a signi cant effect on the results. Fourth, although the channeling bias was not taken into account, a detection bias analysis was planned in the protocol.
Over-diagnosis of LA in metformin users was not observed despite the longstanding assumption that metformin may be associated with lactic acidosis.
In brief, in this study metformin was associated with an increased risk of lactic acidosis in patients with type 2 diabetes mellitus with chronic kidney disease. New data is needed to con rm the association between diuretic drugs and lactic acidosis in this group of patients. The ALIMAR-C2 study was entered in the European Union electronic Register of Post-Authorisation Studies (EU PAS Register Number: EUPAS13969, June 2016) and received approval from the European Network of Centres for Pharmacoepidemiology and Pharmacovigilance (ENCePP) (http://www.encepp.eu/encepp/viewResource.htm?id=14215).
Regarding the data contained in the databases, and according to Spanish legislation on the con dentiality and protection of data (Ley Orgánica 3/2018, de 5 de diciembre, de Protección de Datos Personales y garantía de los derechos digitales), data included in this study consist of the identi cation of incident cases of LA in hospitals and some speci c information from primary healthcare. In order to link this information, the same identity code is needed for each patient. A coding system was used to link this information while keeping the identity of the patients anonymous.

Consent for publication
Not applicable.

Availability of data and materials
The datasets used and analysed during the current study are available from the corresponding author on reasonable request.

Con ict of interest
The authors declare that they have no con ict of interest.

Funding
The project received a research grant from the Carlos III Institute of Health, Ministry of the Economy and Competitiveness (Spain), in 2015 (reference PI15/00764) under the Health Strategy Action 2015, within the Technical, Scienti c and Innovation Research National Plan 2013-2016 (Co-funded by the European Regional Development Fund or ERDF). The funders had no role in the study design, data collection or analysis, the decision to publish or the preparation of the manuscript. Our study protocol has also undergone a peer-review through the Carlos III Institute of Health.
Author contributions SV is the principal investigator. CP was actively involved in the conception and design of the study. MA, MM, RM, and AG are members of the coordinating group and were also involved in the conception and design of the study.
MA, SV, CP, AG, OP and RM drafted the manuscript. All authors made contributions to the design of the study, critically reviewed the manuscript and approved the nal version.