Abstract
Objective:
To define the risk of laboratory abnormalities associated with commonly prescribed medications in the context of urinary stone disease.
Materials and Methods:
We used the Veterans Health Administration Corporate Warehouse to identify adults ≥ 18 years with an index episode of urinary stone disease. The development of laboratory abnormalities for those ever versus never on preventive pharmacological therapy were then compared using separate Cox proportional hazards regression models for each laboratory abnormality-drug class combination, adjusting for baseline patient characteristics and time-varying medication use over the study period. To provide a sense for absolute risk, we then used the models to estimate two-year predicted probabilities of developing a laboratory abnormality for each.
Results:
Thiazide use was associated with a risk of: hypokalemia (hazard ratio [HR], 2.85; 95% confidence interval [CI], 2.66–3.05), hyponatremia (HR 1.41, 95% CI: 1.24–1.60), and hypercalcemia (HR 2.04, 95% CI: 1.70–2.46). Potassium citrate was associated with a risk of hyperkalemia (HR 1.44, 95% CI: 1.23–1.69). Allopurinol was associated with a risk of cytopenia (HR 1.17, 95% CI: 1.02–1.33). No significant association was observed between thiazide use and hyperglycemia or allopurinol use and transaminitis. The greatest absolute risk was for hypokalemia with thiazide use: a two-year predicted probability of 17.3%.
Conclusions:
Preventive pharmacological therapy for urinary stone disease is associated with an increased risk of a variety of laboratory abnormalities, with the greatest absolute risk being hypokalemia with thiazides at over 1 in 6 patients. The study findings provide guidance for laboratory monitoring while on these therapies.
Keywords: urinary calculi, secondary prevention, Adverse Drug Reactions
Introduction
Urinary Stone Disease (USD) is a prevalent condition that affects 1 in 10 Americans.1 USD recurrence is also common, with up to 50% of patients developing a recurrence over 5 years.2 USD can have severe consequences including chronic kidney disease, worse quality of life, and increased costs of care.3–5 Therefore, preventing USD recurrence is imperative. Although preventive pharmacological therapy has been shown to reduce USD recurrence6,7, it remains underutilized by physicians8 and patient adherence to it is known to decrease over time.9 A recent systematic review of preventive pharmacological therapy highlighted knowledge gaps regarding its treatment harms.7 Therefore, poor characterization of preventive pharmacological therapy-related laboratory abnormalities and adverse effects might be one factor contributing to its underutilization.10 Indeed, studies have shown that provider specialty influences preventive pharmacological therapy use (nephrologists and primary care providers prescribed more preventive medications than urologists)8,11, supporting the hypothesis that providers unfamiliar with medication side effects might be reluctant to prescribe them.
The safety and tolerability of preventive pharmacological therapy are important considerations in understanding the burden of preventive treatment. Different classes of preventive pharmacological therapy exist, including: thiazide diuretics, alkali therapy, and uric acid-lowering drugs. Each of these drug classes are associated with specific laboratory abnormalities. For example, thiazides have been associated with hypokalemia, hyponatremia, hyperglycemia, and hypercalcemia; potassium alkali therapy with hyperkalemia; and uric acid-lowering drugs (allopurinol) with transaminitis and cytopenia. However, prior studies linking preventive pharmacological therapy with laboratory abnormalities have been limited by small sample size6 or using claims-based approaches rather than laboratory values, which is subject to coding errors.10 Therefore, our objective was to define the risk of laboratory abnormalities associated with preventive pharmacological therapy using data from a large sample of patients from the Veterans Health Administration.
Materials and Methods
Data Source and study population
We used the Veterans Health Affairs Corporate Data Warehouse, which includes inpatient and outpatient medical records and pharmacy data, from 23.5 million living and deceased U.S. Veterans who use Veterans Health Affairs facilities for health care. First, we identified adults ≥ 18 years with an index episode of USD between 2010–2020 using a definition of at least two outpatient evaluation and management encounters within a 12-month period or one inpatient or one surgical encounter for USD (based on pertinent procedure codes for the encounter, e.g. percutaneous nephrolithotomy, ureteroscopy, shockwave lithotripsy, etc. See Appendix for the full list). Patients with USD encounters in 2009 were excluded to ensure a minimum 12-month period without encounters preceding the index episode. For those with multiple outpatient/inpatient or surgical encounters for USD, the first encounter was considered the index encounter.
Exposure
The main exposure variable of interest was whether patients were prescribed preventive pharmacological therapy or not. We used relevant generic drug names to identify patients with a prescription fill for a thiazide diuretic, alkali, or a uric acid-lowering drug of at least 30 days’ supply after the index stone encounter (see Appendix). To minimize preventive pharmacological therapy prescription for non-USD indications in the cohort, patients on these agents in the 6 months prior to the index stone event were excluded. To study the effects of each drug class on laboratory abnormalities, patients who were on multiple preventive pharmacological therapy medications were censored.
Outcomes
The primary outcome was the occurrence of a relevant laboratory abnormality for each class of preventive pharmacological therapy agent. Laboratory values were assessed based on Logical Observation Identifiers Names and Codes (LOINC, Appendix). For thiazide therapy, the outcomes examined were: hyponatremia, defined as ≤ 130 mEq/L, hypokalemia, defined as ≤ 3.2 mEq/L, hyperglycemia, defined as ≥ 200 mg/dL, and hypercalcemia, defined as ≥ 11 mg/dL. For alkali therapy (limited to potassium citrate), we examined hyperkalemia, defined as ≥ 5.5 mEq/L. For uric acid lowering agents (limited to allopurinol), the outcomes examined were: transaminitis, defined as an AST/ALT ≥ 100 U/L, and cytopenia, defined as a WBC of ≤ 3000/uL, hemoglobin of ≤ 8 mg/dL, or platelets ≤ 100,000/uL.
Study Variables
Variables examined and adjusted for in the statistical models were: Diabetes mellitus (based on relevant diagnosis codes within 12 months prior to index stone event), estimated glomerular filtration rate (eGFR, based on most recent serum creatinine values within 12 months prior to index stone event), sex, high-risk status for stone recurrence, race, ethnicity, and geographic region of patient residence. High-risk status for stone recurrence was defined as having at least one comorbid condition associated with USD risk within 12 months before index stone event, using a previously published approach (see Appendix for a list of conditions).12
Statistical Analysis
For each preventive pharmacological therapy drug class, we compared patient characteristics listed above stratified by whether patients ever filled a prescription over the study period versus never (Table 1). The modeling approach utilized Cox proportional hazard regression for each drug class and outcome combination (thiazides: hypokalemia, hyponatremia, hyperglycemia, hypercalcemia; alkali citrate: hyperkalemia; and allopurinol: transaminitis, cytopenia). For the thiazide-hyperglycemia model only, patients with an existing diagnosis of diabetes mellitus were excluded. Time zero for the model was set at time of index stone diagnosis, with censoring if the patient changed drug classes, died, or the study period ended. The outcome was the first occurrence of the relevant laboratory abnormality as noted above. To account for risk of laboratory abnormalities related to medication use over study follow-up more precisely, a time-varying covariate was used to identify periods during which patients were on preventive pharmacological therapy. In addition, as identification of a laboratory abnormality requires a blood draw, and patients did not obtain bloodwork at consistent intervals, a time-varying covariate representing the cumulative number of laboratory draws over the study period was also included. Other variables noted in the section above were additionally entered as baseline covariates. Hazard ratios with 95% confidence intervals were estimated. In addition, to provide a more intuitive sense of absolute risks of laboratory abnormalities, we estimated two-year probabilities of a laboratory abnormality as a function of medication use, at average values of other variables, from the models (Table 2).
Table 1.
Baseline characteristics of the PPT cohort. GFR = glomerular filtration rate (ml/min/1.73 m2).
| Variable | Thiazide (n=17050) | No Thiazides (n=187424) | Potassium Citrate (n=9431) | No Potassium Citrate (n=195043) | Allopurinol (n=7319) | No Allopurinol (n=197155) | Overall (n=204474) |
|---|---|---|---|---|---|---|---|
|
| |||||||
| Age (y) | |||||||
| Mean (SD) | 60.2 (12.2) | 60.6 (15.1) | 59.4 (13.4 | 60.6 (15.0) | 60.5 (15.0) | 62.4 (12.5) | 60.6 (14.9) |
| Gender | |||||||
| Female | 1191 (7.0%) | 13197 (7.0%) | 713 (7.6%) | 13675 (7.0%) | 14225 (7.2%) | 163 (2.2%) | 14388 (7.0) |
| Male | 15859 (93.0%) | 174227 (93.0%) | 8718 (92.4%) | 181368 (93.0%) | 182930 (92.8%) | 7156 (97.8%) | 190086 (93.0) |
| Race (n,%) | |||||||
| Missing | 884 | 11857 | 563 | 12178 | 12315 | 426 | 12741 |
| Black | 2417 (15.0%) | 19348 (11.0%) | 808 (9.1%) | 20957 (11.5%) | 20989 (11.4%) | 776 (11.3%) | 21765 (11.4) |
| Other | 447 (2.8%) | 4682 (2.7%) | 294 (3.3%) | 4835 (2.6%) | 4863 (2.6%) | 266 (3.9%) | 5129 (2.7) |
| White | 13302 (82.3%) | 151537 (86.3%) | 7766 (87.6%) | 157073 (85.9%) | 158988 (86.0%) | 5851 (84.9%) | 164839 (86.0) |
| Ethnicity (n,%) | |||||||
| Missing | 515 | 6677 | 322 | 6870 | 6969 | 223 | 7192 |
| Hispanic/L atino | 945 (5.7%) | 12351 (6.8%) | 540 (5.9%) | 12756 (6.8%) | 12931 (6.8%) | 365 (5.1%) | 13296 (6.7) |
| Non-Hispanic | 15590 (94.3%) | 168396 (93.2%) | 8569 (94.1%) | 175417 (93.2%) | 177255 (93.2%) | 6731 (94.9%) | 183986 (93.3) |
| Region (n,%) | |||||||
| Midwest | 3733 (21.9%) | 37867 (20.2%) | 1935 (20.5%) | 39665 (20.3%) | 40133 (20.4%) | 1467 (20.0%) | 41600 (20.3) |
| Northeast | 1804 (10.6%) | 23926 (12.8%) | 1207 (12.8%) | 24523 (12.6%) | 24930 (12.6%) | 800 (10.9%) | 25730 (12.6) |
| South | 8032 (47.1%) | 86318 (46.1%) | 3874 (41.1%) | 90476 (46.4%) | 90859 (46.1%) | 3491 (47.7%) | 94350 (46.1) |
| West | 3481 (20.4%) | 39313 (21.0%) | 2415 (25.6%) | 40379 (20.7%) | 41233 (20.9%) | 1561 (21.3%) | 42794 (20.9) |
| High risk (n, %) | 3951 (23.2%) | 48556 (25.9%) | 2388 (25.3%) | 50119 (25.7%) | 49539 (25.1%) | 2968 (40.6%) | 52507 (25.7) |
| Type 2 Diabetes (n,%) | 8669 (50.8%) | 76494 (40.8%) | 4511 (47.8%) | 80652 (41.4%) | 81008 (41.1%) | 4155 (56.8%) | 85163 (41.6) |
| GFR (n,%) | |||||||
| Missing | 1436 (8.4%) | 15176 (8.1%) | 665 (7.1%) | 15947 (8.2%) | 16020 (8.1%) | 592 (8.1%) | 16612 (8.1) |
| <30 | 167 (1.0%) | 4081 (2.2%) | 169 (1.8%) | 4079 (2.1%) | 3897 (2.0%) | 351 (4.8%) | 4248 (2.1) |
| 30–59 | 2528 (14.8%) | 28220 (15.1%) | 1876 (19.9%) | 28872 (14.8%) | 28848 (14.6%) | 1900 (26.0%) | 30748 (15.0) |
| ≥60 | 12919 (75.8%) | 139947 (74.7%) | 6721 (71.3%) | 146145 (74.9%) | 148390 (75.3%) | 4476 (61.2%) | 152866 (74.8) |
Table 2.
Two-year probability of developing a laboratory abnormalities while on PPT. T2DM = Diabetes Mellitus. GFR = glomerular filtration rate (ml/min/1.73 m2).
| Variable | Thiazides | Potassium Citrate | Allopurinol | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
| ||||||||||||||
| Hypokalemia | Hyponatremia | Hypergiycemia | Hypercalcemia | Hyperkalemia | Transaminitis | Cytopenia | ||||||||
|
| ||||||||||||||
| No Drug | Drug | No Drug | Drug | No Drug | Drug | No Drug | Drug | No Drug | Drug | No Drug | Drug | No Drug | Drug | |
|
| ||||||||||||||
| Overall | ||||||||||||||
| 8.1 | 17.3 | 3.3 | 3.8 | 1.6 | 1.5 | 1.2 | 2.1 | 3.8 | 5.4 | 5.6 | 5.2 | 7.8 | 7.6 | |
| Gender | ||||||||||||||
| Female | 12.3 | 29.2 | 2.5 | 2.9 | 1.2 | 0.5 | 1.6 | 2.8 | 2.5 | 2.5 | 3.6 | 1.8 | 6.7 | 2.1 |
| Male | 7.8 | 16.5 | 3.4 | 3.8 | 1.6 | 1.6 | 1.2 | 2.0 | 3.9 | 5.6 | 5.8 | 5.4 | 7.9 | 7.7 |
| Race | ||||||||||||||
| Black | 11.4 | 20.9 | 3.5 | 3.9 | 1.7 | 1.1 | 1.7 | 2.5 | 4.3 | 5.4 | 6.4 | 8.2 | 12.0 | 10.1 |
| Other | 10.0 | 17.4 | 3.7 | 4.7 | 1.7 | 1.2 | 1.1 | 1.9 | 3.4 | 7.8 | 5.8 | 8.2 | 7.7 | 1.6 |
| White | 7.7 | 17.2 | 3.3 | 3.7 | 1.6 | 1.6 | 1.2 | 2.1 | 3.7 | 5.3 | 5.5 | 4.7 | 7.4 | 7.6 |
| T2DM | ||||||||||||||
| Not T2DM | 7.7 | 20.2 | 2.4 | 3.3 | - | - | 1.1 | 1.6 | 2.9 | 4.9 | 4.8 | 6 | 7.2 | 8.2 |
| T2DM | 8.7 | 15.5 | 5.0 | 5.3 | - | - | 1.4 | 2.7 | 5.4 | 7.0 | 6.9 | 5.9 | 8.7 | 7.8 |
| eGFR | ||||||||||||||
| < 30 | 14.7 | 22.0 | 8.3 | 8.8 | 3.5 | 2.8 | 2.4 | 4.9 | 20.1 | 23.9 | 8.6 | 5.6 | 22.6 | 15.5 |
| 30 - 59 | 8.3 | 15.2 | 3.8 | 4.0 | 1.9 | 1.4 | 1.5 | 1.8 | 6.9 | 9.0 | 5.5 | 5.7 | 9.4 | 9.7 |
| >=60 | 8.4 | 18.8 | 3.3 | 3.9 | 1.6 | 1.8 | 1.2 | 2.1 | 3.2 | 5.1 | 5.8 | 6.0 | 7.6 | 7.8 |
| Missing | 4.8 | 13.9 | 1.9 | 2.2 | 0.9 | 0.0 | 0.9 | 2.1 | 3.7 | 4.6 | 3.9 | 4.7 | 5.0 | 6.1 |
| Age | ||||||||||||||
| 18 - 64 | 7.7 | 17.9 | 3.2 | 3.5 | 1.4 | 1.5 | 1.1 | 1.9 | 3.3 | 5.4 | 7.1 | 7.6 | 6.6 | 5.0 |
| 65 - 74 | 8.1 | 16.0 | 3.4 | 3.7 | 1.9 | 1.8 | 1.3 | 2.4 | 4.6 | 5.0 | 4.2 | 3.9 | 9.4 | 11.0 |
| 75 + | 10.2 | 18.6 | 3.9 | 5.0 | 2.7 | 3.0 | 1.4 | 2.1 | 5.1 | 8.2 | 4.1 | 3.4 | 12.5 | 11.2 |
| Lab Count | ||||||||||||||
| 0 - 2 | 3.5 | 12.8 | 1.7 | 2.2 | 0.6 | 0.5 | 0.9 | 2.1 | 2.3 | 3.4 | 4.0 | 4.4 | 5.1 | 4.3 |
| 3 - 5 | 16.5 | 24.7 | 5.3 | 4.3 | 26.3 | 1.8 | 1.4 | 1.3 | 5.7 | 6.0 | 9.7 | 10.1 | 17.3 | 17.0 |
| 6+ | 26.6 | 30.7 | 13.3 | 9.2 | 14.4 | 8.1 | 2.7 | 2.5 | 11.8 | 13.3 | 24.4 | 18.1 | 27.0 | 22.1 |
Post-hoc analysis
To further explore the thiazide-hypokalemia association, we conducted a dose-response analysis. Our main modeling approach used a complex, time-varying covariate approach, which makes it difficult to further incorporate differences in dosing. Thus, we used a cruder approach to examine the potential impact of dosing. We defined three different dosing groups for thiazides based on first prescription for thiazides within 6 months of the index stone event. Based on prior work, we defined these groups as: low (<12.5 mg/day chlorthalidone, <0.6125 mg/day indapamide, <25 mg/day hydrochlorothiazide), medium (12.5 to <25 mg/day chlorthalidone, 0.6125 to <1.25 mg/day indapamide, 25 to <50 mg/day hydrochlorothiazide), and high (≥25 mg/day chlorthalidone, ≥1.25 mg/day indapamide, ≥50 mg/day hydrochlorothiazide).13 We then looked for the first occurrence of hypokalemia within a 6–30 month window following the index stone event using a Cox model adjusted for baseline covariates. This simplified analysis gives a sense for potential contribution of dose.All analyses were conducted using R v4.0.5 with two-sided testing and α=0.05. The study was approved by the Institutional Review Board of the Veterans Administration Ann Arbor Healthcare System with a waiver of informed consent.
Results
The cohort consisted of 204,474 patients who met the inclusion criteria. Of these, 17,050 patients were on thiazide therapy; 9,431 were on alkali therapy; and 7,319 were on uric acid-lowering drugs (Table 1). Most of the cohort was male (93%), White (86.0%), non-Hispanic (93.3%), and located in the South (46.1%). Notable differences between preventive pharmacological therapy groups included a higher rate of diabetes mellitus (50.8% vs 40.8%) for those ever vs. never on thiazides (Table 1) and a lower prevalence of high risk for stone recurrence for those ever versus never on allopurinol (25.1% vs. 40.6%).
As shown in Figure 1, in the adjusted Cox models thiazide treatment was associated with an increased risk of hypokalemia (HR 2.85, 95% CI: 2.7–3.1 p<0.001). Thiazides were also associated with hyponatremia (HR 1.41, 95% CI: 1.2–1.6, p<0.001) and hypercalcemia (HR 2.04, 95% CI: 1.7–2.5, p<0.001). No significant association was identified between thiazides and hyperglycemia (HR 1.22, 95% CI: 0.9–1.6, p=0.2). Potassium citrate was associated with hyperkalemia (HR 1.44, 95% CI: 1.2–1.7, p<0.001). Uric Acid-lowering drugs were not associated with transaminitis (HR 0.85, 95% CI: 0.7–1.0, p=0.1), but were associated with cytopenia (HR 1.17, 95% CI: 1.0–1.3, p=0.022.
Figure 1.
Forest plot showing the various hazard ratios for developing a side effect on preventative pharmacologic therapy
In Table 2 we display two-year predicted probabilities of laboratory abnormalities for each drug class and by patient subgroups, based on the Cox models. The greatest absolute risk was for hypokalemia with thiazide use, with a two-year predicted probability of 17.3% for patients on thiazide therapy versus 8.1% for those not. When analyzing by subgroup, the two-year predicted probability for developing hypokalemia on thiazide therapy was particularly high for women at 29.2%, compared to 16.5% for men. In addition, the probability of hypokalemia for patients of Black race on thiazides was 20.9% compared to White patients at 17.2%. The probabilities of hyponatremia and hypercalcemia on thiazide therapy were more modest, at 3.8% (versus 3.3% not on therapy) and 2.1% (versus 1.2% not on therapy), respectively. On potassium citrate therapy, the probability of hyperkalemia was 5.4%, versus. 3.8% not on therapy. In a post-hoc analysis designed to examine for a potential dose-response effect of thiazide therapy on the risk of hypokalemia, we found that the occurrence of hypokalemia in the low, medium, and high dose thiazide groups was 6.6%, 11.5%, and 18.1%, respectively, in the 6–30 month window following the index stone event. In a Cox model adjusting for baseline covariates only, relative to the low dose group, the HR for hypokalemia was increased for the medium (HR 1.90, 95% CI: 1.5–2.4, p<0.001) and high thiazide dose groups (HR 2.97, 95% CI 2.3–3.9, p<0.001).
Discussion
In a large national cohort of patients on preventive pharmacological therapy for USD, we demonstrated significant associations with an increased risk of various laboratory abnormalities. The greatest absolute risk was hypokalemia for patients on thiazide therapy, with over 1 in 6 patients predicted to develop it over a two-year period. Additionally, there were significant but more modest risks of hyponatremia and hypercalcemia with thiazides, and hyperkalemia with potassium citrate therapy. No significant associations were identified between thiazides and hyperglycemia, or allopurinol and transaminitis, and although allopurinol had a statistically significant association with cytopenia, the effect was small and likely of limited clinical importance.
The risk of hypokalemia with thiazides has been reported in large studies in the context of hypertension.14 However, studies examining the risk of hypokalemia and thiazides in the context of stone disease have been smaller in size or claims-based.6,10 Indeed, our group had previously reported the occurrence of laboratory abnormalities with different preventive pharmacological therapy classes using a claims based approach and identified that thiazides were associated with higher odds of hypokalemia (OR: 2.01, 95% CI: 1.44–2.81).10 Our findings are notable, as to our knowledge, this is the largest study to examine the risk of hypokalemia for USD patients on thiazides using confirmatory laboratory values.
There may be important differences modifying the risk of hypokalemia with thiazides in the context of USD. For example, it is possible that USD patients could have higher rates of baseline hypokalemia; indeed, patients with a Type I distal renal tubular acidosis can have hypokalemia, nephrocalcinosis, and calcium-USD.15 Similarly, patients with Bartter Syndrome, a salt-wasting tubulopathy can also develop nephrocalcinosis, USD, and hypokalemia.16 Subgroups of the thiazide cohort at high risk for hypokalemia included women and Black Americans. Sex-differences for thiazide-induced hypokalemia (higher risk in women) have been previously reported.17 Prior evidence also supports racial-differences in baseline hypokalemia and medication-induced hypokalemia, with African Americans at greater risk.18
There were several reassuring negative findings in the study. For example, we found no significant association between thiazide use and hyperglycemia. Although there are several proposed mechanisms for thiazide-induced hyperglycemia19, evidence from observational trials have been mixed20 and post-hoc analyses of large clinical trials did not find associations between new-onset diabetes and thiazides.21 Collectively, these results suggest the risk of new development of hyperglycemia in patients without diabetes mellitus is low. Furthermore, we observed no significant association between allopurinol and transaminitis, and only a very modest association with cytopenia. This may be related to the rarity of these effects. Reactions to allopurinol such as allopurinol hypersensitivity syndrome, which can cause hepatotoxicity, only occur in 1 out of 1000 patients.22 Allopurinol is thought to mainly function as an inhibitor of xanthine oxidase, but there is evidence suggesting it can also affect purine metabolism, which can ultimately contribute to cytopenia.23 However, serious hematologic laboratory abnormalities of allopurinol including agranulocytosis24 and bone marrow suppression25 have primarily been noted in case reports.
There are limitations to our study. First, the cohort we analyzed is from the VHA, which is predominantly male with multiple co-morbidities. Thus, our results may not be as generalizable to a female, non-VHA population. Nevertheless, our study analyzed 14,388 women, which is much larger than other studies to examine this issue and our model adjusted for comorbidities and sex. Given its observational nature, ascertainment bias may have also been an issue. In particular, blood draws would have only been performed if the ordering physician felt there was a clinical indication and we could only detect if a patient had a laboratory abnormality if bloodwork was checked. However, to address this issue, we used a time-varying analysis including a covariate which accounted for blood work done over the course of the study period. Furthermore, because the general Veteran population has multiple medical comorbidities, routine blood counts and serum chemistries are often obtained, reducing the effect of the ascertainment bias. Third, there may be selection bias in who is prescribed preventive pharmacological therapy, which could tend to underestimate risks of laboratory abnormalities relative to the control group. For example, the relative risk of hyperkalemia in patients on potassium citrate with an eGFR < 30 mL/min/1.73 m2 versus the control group was fairly small (23.9% versus 20.1%, respectively). This likely represents selective prescribing of potassium to patients with reduced kidney function who may have started out with lower levels of serum potassium. Nevertheless, our multivariable models adjusted for several clinical characteristics to mitigate these issues. Lastly, this study is reflective of patterns of laboratory abnormalities as observed in the setting of real-world prescribing preventive pharmacological therapy.
Despite guideline recommendations26 and evidence supporting their use6,7, the use of preventive pharmacological therapy remains low.8 One factor contributing to this could be prescriber concerns about the associated risk of laboratory abnormalities. The American Urological Association Guidelines on Medical Management of Kidney Stones recommend that clinicians should obtain periodic blood testing to assess for adverse effects in patients on pharmacological therapy.26 However, the type and timing of testing is recommended to be tailored to the patient’s comorbidities and medications. Patient concerns likely also contribute to known issues with adherence to preventive pharmacological therapy.9 Indeed, a systematic review of preventive pharmacological therapy from the Agency of Healthcare Research and Quality identified that most trials did not report treatment harms.7 Our study shows that the risk of hypokalemia is notable, suggesting that prescribers should monitor patients on thiazides closely for this. There also seems to be a dose dependent risk for thiazides. The risks of hyponatremia, hypercalcemia, and hyperkalemia are more modest, but still warrant monitoring. Finally, risks of hyperglycemia with thiazides and transaminitis or cytopenia with allopurinol are very low, suggesting prescribers may only to need to monitor for these laboratory abnormalities in selected patients at high risk (e.g., patients on allopurinol with concomitant bone marrow suppressing agents). Although our study focuses on the use of thiazides, potassium citrate, and allopurinol in the context of preventive pharmacological therapy for USD, given the widespread use of these medications for other indications, our findings have important clinical implications for the general medical community.
Conclusions
This study identified that preventive pharmacological therapy is associated with various laboratory abnormalities. Thiazides have a notable risk of hypokalemia, which occurs in more than 1 out of 6 patients and warrants close laboratory monitoring, with more modest risks noted for hyponatremia and hypercalcemia with thiazides, and hyperkalemia with potassium citrate. Medical professionals who prescribe these medications should be aware of these laboratory abnormalities and screen patients periodically to diagnose and treat any potential abnormalities that arise.
Supplementary Material
Supplementary Table 1A. Adjusted cox proportional hazard regression for developing side effects on Thiazide therapy. GFR = glomerular filtration rate (ml/min/1.73 m2). Hyponatremia was defined as ≤ 130 mEq/L, hypokalemia was defined as ≤ 3.2 mEq/L, hyperglycemia was defined as ≥ 200 mg/dL, hypercalcemia was defined as ≥ 11 mg/dL.
Supplementary Table 1B. Adjusted cox proportional hazard regression for developing side effects on Potassium citrate therapy. hyperkalemia was defined as ≥ 5.5 mEq/L
Supplementary Table 1C. Adjusted cox proportional hazard regression for developing side effects on Allopurinol therapy. Transaminitis was defined as an AST/ALT ≥ 100 U/L and cytopenia was defined as a (WBC of ≤ 3,000/uL, Hemoglobin of ≤ 8 mg/dL, or Platelets ≤ 100,000/uL)
Funding:
Research reported in this publication was supported by the National Institute of Diabetes and Digestive and Kidney Diseases of the National Institutes of Health under award number R01DK121709. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Matthew Lee is supported by The Ohio State University College of Medicine Research Innovation and Career Development Award
Footnotes
Declaration of Competing Interest
Dr. John Doe no conflict
Dr. Jane Doe paid consultant to company “x”
Ethics approval: This study was deemed exempt by the Institutional Review Board of the Veterans Administration Ann Arbor Healthcare system.
AI: The authors did not use generative AI or AI-assisted technologies in the development of this manuscript
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Hill AJ, Basourakos SP, Lewicki P, et al. Incidence of Kidney Stones in the United States: The Continuous National Health and Nutrition Examination Survey. J Urol. Apr 2022;207(4):851–856. doi: 10.1097/ju.0000000000002331 [DOI] [PubMed] [Google Scholar]
- 2.Ljunghall S Incidence of upper urinary tract stones. Miner Electrolyte Metab. 1987;13(4):220–7. [PubMed] [Google Scholar]
- 3.Alexander RT, Hemmelgarn BR, Wiebe N, et al. Kidney stones and kidney function loss: a cohort study. Bmj. Aug 29 2012;345:e5287. doi: 10.1136/bmj.e5287 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Penniston KL, Sninsky BC, Nakada SY. Preliminary Evidence of Decreased Disease-Specific Health-Related Quality of Life in Asymptomatic Stone Patients. J Endourol. May 2016;30 Suppl 1:S42–5. doi: 10.1089/end.2016.0074 [DOI] [PubMed] [Google Scholar]
- 5.Litwin MS SC, editors. Urologic Diseases in America. vol NIH Publication No. 12–7865. US Department of Health and Human Services, Public Health Service, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases. US Government Printing Office; 2012. [Google Scholar]
- 6.Ettinger B, Citron JT, Livermore B, Dolman LI. Chlorthalidone reduces calcium oxalate calculous recurrence but magnesium hydroxide does not. J Urol. Apr 1988;139(4):679–84. doi: 10.1016/s0022-5347(17)42599-7 [DOI] [PubMed] [Google Scholar]
- 7.Fink HA, Wilt TJ, Eidman KE, et al. Medical management to prevent recurrent nephrolithiasis in adults: a systematic review for an American College of Physicians Clinical Guideline. Ann Intern Med. Apr 2 2013;158(7):535–43. doi: 10.7326/0003-4819-158-7-201304020-00005 [DOI] [PubMed] [Google Scholar]
- 8.Krampe N, Oerline MK, Hsi RS, et al. Understanding the Barriers to Preventive Pharmacological Therapy Use in Older Patients With Urinary Stone Disease. Urol Pract. Jan 2024;11(1):218–225. doi: 10.1097/upj.0000000000000480 [DOI] [PubMed] [Google Scholar]
- 9.Parks JH, Asplin JR, Coe FL. Patient adherence to long-term medical treatment of kidney stones. J Urol. Dec 2001;166(6):2057–60. [PubMed] [Google Scholar]
- 10.Hsi RS, Crivelli JJ, Yan PL, Shahinian V, Hollingsworth JM. Comparison of Class-Specific Side Effects Across Preventative Pharmacologic Therapies for Kidney Stone Disease. Urol Pract. Jan 2024;11(1):172–178. doi: 10.1097/upj.0000000000000470 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Dauw CA, Yi Y, Bierlein MJ, et al. Factors Associated With Preventive Pharmacological Therapy Adherence Among Patients With Kidney Stones. Urology. Jul 2016;93:45–9. doi: 10.1016/j.urology.2016.03.030 [DOI] [PubMed] [Google Scholar]
- 12.Hsi RS, Yan PL, Crivelli JJ, Goldfarb DS, Shahinian V, Hollingsworth JM. Comparison of Selective vs Empiric Pharmacologic Preventive Therapy of Kidney Stone Recurrence With High-Risk Features. Urology. Jun 2022;164:74–79. doi: 10.1016/j.urology.2021.12.037 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Hsi RS, Yan PL, Maalouf NM, et al. Thiazide Dose, Urine Calcium, and Symptomatic Kidney Stone Events. JAMA Netw Open. Aug 1 2024;7(8):e2428953. doi: 10.1001/jamanetworkopen.2024.28953 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Group. AOaCftACR. Major outcomes in high-risk hypertensive patients randomized to angiotensin-converting enzyme inhibitor or calcium channel blocker vs diuretic: The Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT). Jama. Dec 18 2002;288(23):2981–97. doi: 10.1001/jama.288.23.2981 [DOI] [PubMed] [Google Scholar]
- 15.Fuster DG, Moe OW. Incomplete Distal Renal Tubular Acidosis and Kidney Stones. Adv Chronic Kidney Dis. Jul 2018;25(4):366–374. doi: 10.1053/j.ackd.2018.05.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Rodriges JHP, Menezes Silva LAW, Soares SBM, Cruz RRO, Mrad FCC, Simoes ESAC. Clinical Course of Patients with Bartter Syndrome. Iran J Kidney Dis. May 2022;16(3):162–170. [PubMed] [Google Scholar]
- 17.Toner JM, Ramsay LE. Thiazide-induced hypokalaemia; prevalence higher in women. Br J Clin Pharmacol. Sep 1984;18(3):449–52. doi: 10.1111/j.1365-2125.1984.tb02488.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Chen Y, Sang Y, Ballew SH, et al. Race, Serum Potassium, and Associations With ESRD and Mortality. Am J Kidney Dis. Aug 2017;70(2):244–251. doi: 10.1053/j.ajkd.2017.01.044 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Dronavalli S, Bakris GL. Mechanistic Insights into Diuretic-Induced Insulin Resistance. Hypertension. 2008/December/01 2008;52(6):1009–1011. doi: 10.1161/HYPERTENSIONAHA.108.120923 [DOI] [PubMed] [Google Scholar]
- 20.Carter BL, Einhorn PT, Brands M, et al. Thiazide-induced dysglycemia: call for research from a working group from the national heart, lung, and blood institute. Hypertension. Jul 2008;52(1):30–6. doi: 10.1161/hypertensionaha.108.114389 [DOI] [PubMed] [Google Scholar]
- 21.Savage PJ, Pressel SL, Curb JD, et al. Influence of long-term, low-dose, diuretic-based, antihypertensive therapy on glucose, lipid, uric acid, and potassium levels in older men and women with isolated systolic hypertension: The Systolic Hypertension in the Elderly Program. SHEP Cooperative Research Group. Arch Intern Med. Apr 13 1998;158(7):741–51. doi: 10.1001/archinte.158.7.741 [DOI] [PubMed] [Google Scholar]
- 22.Alotaibi T, Bjazevic J, Kim R, et al. Allopurinol hypersensitivity syndrome Raising awareness of an uncommon but potentially serious adverse event among kidney stone patients. Can Urol Assoc J. May 2024;18(5):E167–e172. doi: 10.5489/cuaj.8685 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Fox IH, Wyngaarden JB, Kelley WN. Depletion of erythrocyte phosphoribosylpyrophosphate in man. N Engl J Med. Nov 26 1970;283(22):1177–82. doi: 10.1056/nejm197011262832201 [DOI] [PubMed] [Google Scholar]
- 24.Wilkinson DG. Allopurinol and agranulocytosis. Lancet. Dec 17 1977;2(8051):1282–3. doi: 10.1016/s0140-6736(77)92683-6 [DOI] [Google Scholar]
- 25.Alhubaishi AA. Pancytopenia and Septic Infection Caused by Concurrent Use of Allopurinol and Mercaptopurine: A Case Report Illustrating the Importance of Clinical Pharmacist Consultation. Am J Case Rep. Aug 23 2019;20:1245–1247. doi: 10.12659/ajcr.914166 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Pearle MS, Goldfarb DS, Assimos DG, et al. Medical management of kidney stones: AUA guideline. J Urol. Aug 2014;192(2):316–24. doi: 10.1016/j.juro.2014.05.006 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Table 1A. Adjusted cox proportional hazard regression for developing side effects on Thiazide therapy. GFR = glomerular filtration rate (ml/min/1.73 m2). Hyponatremia was defined as ≤ 130 mEq/L, hypokalemia was defined as ≤ 3.2 mEq/L, hyperglycemia was defined as ≥ 200 mg/dL, hypercalcemia was defined as ≥ 11 mg/dL.
Supplementary Table 1B. Adjusted cox proportional hazard regression for developing side effects on Potassium citrate therapy. hyperkalemia was defined as ≥ 5.5 mEq/L
Supplementary Table 1C. Adjusted cox proportional hazard regression for developing side effects on Allopurinol therapy. Transaminitis was defined as an AST/ALT ≥ 100 U/L and cytopenia was defined as a (WBC of ≤ 3,000/uL, Hemoglobin of ≤ 8 mg/dL, or Platelets ≤ 100,000/uL)