To our knowledge, this study is the largest case series of SMX-induced crystalline nephropathies to date. This comprehensive investigation involved the collection of clinical data associated with SMX-induced crystalluria, enabling us to identify risk factors and characterize kidney outcomes associated with SMX-induced crystalline nephropathy. Our aim was also to study the kinetics of impaired kidney function upon SMX introduction and after drug discontinuation, to assess the prognosis and kidney survival of affected patients, and to analyze the characteristics of crystalluria (including crystal morphology, urinary pH and density).
Previous studies have examined the population and risk factors for acute kidney injury on CMX12,13,34.
The cohort consisted mainly of elderly patients, frequently hypertensive, which aligns with findings from previous studies12,13,34. All patients in our study were immunocompromised, a condition leading to confirmed or suspected opportunistic infections that required SMX therapy. Of the immunocompromised patients, 9 of 11 were on long-term corticosteroid therapy, and only one was affected by HIV which is a notable finding in comparison with other studies including more HIV-positive patients.
The majority of patients in our study had preexisting chronic kidney disease (63.6%), including 4 patients with stage 4 CKD. Previous studies suggested that preexisting chronic kidney disease is a risk factor for SMX-induced acute kidney injury12,35.
Based on Le Gall et al., a median IGS-2 score of 37 corresponds to an expected hospital mortality rate of approximately 21.7%, a severe condition but an intermediate risk for patients admitted to intensive care units36.
Four patients exhibited hyperkalemia, which was secondary to acute kidney injury. Trimethoprim, which has an inhibitory effect on tubular potassium secretion, might have contributed to this condition37.
The FDA (Food Drug Administration) recommends a daily dose of 800 mg for the treatment of common infections in adults, while the recommended dose for pneumocystis pneumonia is 75–100 mg/kg/day of SMX, with dose adjustments based on kidney function.
In our study, all patients were initially treated with intravenous SMX, with a median daily dose of 55.81 mg/kg/day at introduction. The risk of AKI associated with CMX use has been studied, but the literature has reported inconsistent findings regarding the relationship between dosage and AKI risk.
Fraser’s and Shimizu’s studies provided conflicting results regarding a dose-dependent relationship between SMX and AKI12,13. Fraser et al. did not find a significant relationship whereas Shimizu et al. identified a dose-dependent effect12.
In our study, the median time to onset of AKI after SMX introduction was 4 days (IQR 3.5–5 days), shorter than in Fraser’s study (6.5 days) and Shimizu’s study (8 days [IQR, 5–16 days])12,13.
However, these studies did not distinguish between the different mechanisms of acute kidney injury secondary to CMX use. They probably included patients with SMX-induced crystalline nephropathy, tubulo-interstitial nephropathy due to an immuno-allergic mechanism and, possibly, functional increases in plasma creatinine. Leucocytes were identified in the urine of 5 patients with probable crystalline nephropathy. No eosinophilic cells count was performed. We cannot rule out the participation of immuno-allergic processes but the rapid improvement of kidney function after SMX withdrawal does not support this hypothesis.
Our study focused specifically on SMX-induced crystalline nephropathy, necessitating consideration of this distinct parameter in comparisons with previous studies. According to our observations, risk factors for SMX-related AKI previously identified in former studies are actually risk factors for SMX-related crystalline nephropathy.
We observed NASM crystals in 5 patients who were not included for two reasons (Supplemental Table 1). First, two patients developed AKI but were treated with long-term CMX (Supplemental Table 1).
Second, three patients did not develop AKI: one was treated with short-term oral CMX for an exacerbation of COPD, another received long term prophylaxis for pneumocystosis, and the last patient was treated for suspected pneumocystis intravenously (Supplemental Table 1).
NASM crystals may therefore be observed in urine in the absence of clinically evident kidney lesions. Further prospective studies are required to characterize whether NASM-positive crystalluria is an independent risk factor for AKI, as recently evidenced for amoxicillin38,39. One may hypothesize that intermittent and sparse crystals may be less toxic for kidney tubule than a huge number of crystals: the number of crystals in urine and the global crystalline volume should be assessed in future studies.
Overall, our findings suggest the importance of very early monitoring of kidney function in patients receiving high doses of SMX for pulmonary pneumocystosis, and the need for further prospective studies evaluating the predictive value of crystalluria in patients exposed to SMX.
A study by Berglund et al. in 1975 showed that TMP caused a significant increase in plasma creatinine levels without affecting GFR10. Lee observed a similar effect studying the influence of TMP on tubular transport of organic anions and cations in rat kidney tissue11. Shouval noted that CMX caused transient changes in plasma creatinine and creatinine clearance8. In 1981, Kainer found no significant changes in GFR measured by EDTA clearance in healthy adults treated with CMX9.
In our study, we demonstrate that patients with SMX-induced crystalline nephropathy exhibit acute kidney injury, confirmed by the comparable evolutionary profiles of urea and plasma creatinine levels. The inhibitory effect of TMP on tubular creatinine secretion exists and is confirmed by the dissociation between creatinine and urea levels observed with CMX, as shown in individuals in Fig. 3A–C, but the simultaneous increase in serum creatinine and urea confirms the onset of AKI. Of note, the increase in urea cannot be attributed to the use of corticosteroids, as it coincides with the introduction of CMX and decreases significantly after the drug is discontinued. The evolution of kidney function in Fig. 3A–C illustrates this trend.
In the Shimizu study, univariate logistic regression analysis revealed that concomitant use of loop diuretics, β-lactams, glycopeptides, and aminoglycosides were independent variables associated with the development of acute kidney injury (AKI) under CMX therapy13.
Our aim was to rule out the involvement of other factors in the development of AKI, such as other nephrotoxic agents (including aminoglycosides, vancomycin….) or hemodynamic alterations. Comprehensive data collection, facilitated by the fact that patients were hospitalized in intensive care units, enabled in-depth analysis of each case, including drug dosages, administration timing and other relevant clinical parameters.
The Naranjo score was used to assess the imputability of SMX crystals in cases of AKI, while ruling out other possible causes. Key factors that contribute significant points to Naranjo’s causality assessment score include the time interval between the introduction of the drug and the onset of the adverse reaction, and the absence of other potential causes or other drugs responsible for the adverse reaction33. The presence of other nephrotoxic factors, including drugs, was therefore taken into account, and the SMX was considered as the probable cause (score ≥ 5) of AKI for 11 patients. In most of patients, no other potential cause of AKI was identified. Some patients received other potential nephrotoxic drugs but the timescale analysis and the evolution of kidney function did not support the role of these treatments in the onset of AKI. Only one patient received vasopressor amines, in low doses and briefly.
The kidney prognosis of the patients in our study was generally favorable, as only one patient developed end-stage renal disease, and this patient was already in stage 4 chronic kidney disease prior to the hospitalization. Obviously, the identification of crystals in urine led to SMX withdrawal and may have modified the evolution of kidney function.
The use of sulfonamides has been linked to the presence of crystals in urinary sediments since decades21,23,24,25. The study of crystal structures has been based primarily on individual clinical cases and it was shown that crystal morphology varies with the type of sulfonamide, and sometimes for a same type of sulfonamide21,40.
Prien and Frondell already identified various types of crystals in the urine sediments of patients receiving sulfonamides in 194021. Different shapes were found such as arrowheads, thick convex lenses, rosettes, hexagons, millstones, sheaves (or shocks) of wheat, or dumbbells. The appearance of sulfonamide crystals in the urine is closely related to the solubility of their compounds and acetylated derivatives21,23.
Daudon et al. conducted a retrospective study to determine the frequency and nature of spontaneous crystalluria in a population without uronephrologic pathology41. They observed 466 positive crystalluria (7.6%) out of 6100 urine samples analyzed. Drug crystals accounted for 8.5% of all positive crystalluria. NASM crystals were largely predominant, found in 27 of 466 samples (5.8%), mostly in pure form (4.9%, vs. 0.9% mixed forms), and equally distributed between the sexes. The mean urinary pH of the urine samples was 5.7. They were rarely associated with urinary stones (0.1% of cases).
Few other cases of NASM crystalluria have been reported in the literature25,26,27,30,42,43,44,45,46. The largest retrospective case series reported on NASM-positive crystalluria to date was published in 2018 by Castiglione et al.42. It included seven patients with various bacterial infections. FTIR spectrophotometric analysis authenticated the NASM crystals, as in the studies of Daudon et al.30,31,41. Lower doses of SMX were associated with rectangular-shaped crystals, whereas higher doses resulted in larger, irregularly shaped crystals32.
We found no correlation between the dosage of SMX and crystal morphology. However, we frequently observed various crystal shapes present at the same time in an individual.
The majority of crystal morphologies align with those documented in previous cases, including lozenges, hexagons, flower-like, globules or spheroids, mushrooms or dumb-bells, shocks or sheaves of wheat (which appear to describe a similar morphology using different terms)27,30,32. Interestingly, we also observed in some cases intracellular crystals, probably in macrophages, suggesting that NASM crystals may induce an inflammatory response.
The main potential risk factors identified for SMX-induced crystal nephropathy are acidic urine pH, hypovolemia resulting in low diuresis, hypoalbuminemia, high urine concentrations of SMX and NASM, and underlying chronic kidney disease26,30,32.
Only half of the patients had low urinary pH levels, with a median value of 5.9. This observation is in line with some cases reported by Castiglione32. Hence, ruling out SMX-induced crystalline nephropathy solely based on non-acidic urine criteria may not be justified.
We observed consistently high urinary gravity exceeding 1015 for all patients (median 1025), indicating concentrated urine and reduced urine output.
Hypoalbuminemia may increase the risk of crystalluria by increasing plasma and free urinary concentrations of NASM5,26,30,47. In our study, all patients with available albumin levels (missing data for 5 patients out of 11) were hypoalbuminemic.
Furthermore, it is noteworthy that patients in our cohort received notably high initial doses of SMX, all administered intravenously. For three patients, a transition to oral administration was initiated as kidney function began to decline.
Patients with pre-existing chronic kidney disease are also at high risk, as confirmed by our observations27,30,42.
In summary, our findings highlight the importance of low diuresis volume, hypoalbuminemia, and notably, high doses of SMX, alongside pre-existing chronic kidney disease, as key indicators for potential SMX-induced crystalline nephropathy risk. By contrast, low urine pH does not seem mandatory to develop AKI in this context. Clinicians should pay special attention to these factors when initiating CMX therapy.
Crystalluria, although not commonly conducted in all centers, holds potential as a diagnostic tool for identifying SMX-induced crystalline nephropathy. The wide range of morphologies exhibited by NASM crystals, which can occasionally resemble other crystals like uric acid, cystine, or calcium oxalate monohydrate, emphasizes the necessity of an experienced biologist and the utilization of FTIR spectrophotometry analysis to definitively ascertain their composition30.
Our study has several limitations. First, it is important to acknowledge that our study design is retrospective, which poses inherent limitations. A prospective study, similar to those conducted for Amoxicillin, would provide more robust evidence38,39. This type of study would involve systematically conducting crystalluria assessments on all hospitalized patients receiving therapeutic doses of SMX, allowing for the observation of positive crystalluria and the early characterization of patients who develop acute kidney injury. By incorporating a prospective approach, we could gather more reliable data on the incidence and progression of SMX-induced crystalline nephropathy.
In addition to the retrospective nature of our study, another limitation is the small sample size, which restricts our ability to perform comprehensive data analyses to confirm the risk factors associated with SMX-induced crystalline nephropathy.
Intriguingly, despite the Naranjo score indicating probable crystalline nephropathy, there are no known reports of intratubular NASM crystals in kidney biopsies from patients with acute kidney injury. This might be attributed to biopsy preparation methods. Recently, the presence of Amoxicillin crystals in kidney biopsies has been demonstrated, even though the link between Amoxicillin crystals and acute kidney injury was already established48. Amoxicillin crystals tend to disappear during paraffin embedding but can be identified in frozen sections. A similar hypothesis could apply to NASM crystals, suggesting that frozen sections in patients with a high likelihood of SMX-induced crystalline nephropathy would be worthwhile to investigate. Moreover, most of patients with presumed SMX-induced crystalline nephropathy do not usually undergo kidney biopsies due to the mostly favorable outcomes upon discontinuation of the medication, this was the case for all patients included in this study.
Despite these limitations, our study provides valuable clinical information that will be of help to design future prospective studies. SMX is a widely used drug and caution should be exercised when administering high doses, particularly in patients with chronic kidney disease (CKD), as they appear to be at increased risk of SMX-induced crystalline nephropathy. Close monitoring of kidney function and dosage adjustments are essential in these cases. Thus, in the context of positive crystalluria, dose reduction of SMX is probably appropriate, particularly if crystalline nephropathy is suspected in patients receiving high doses of CMX. Readjustment of the patient’s fluid balance by appropriate hydration, possibly alkaline if urinary pH is acidic, would in fact improve crystal solubility and minimize intratubular precipitation24,29. If crystalline nephropathy is strongly suspected, switching to pentamidine would be an option.
These observations emphasize the significance of utilizing early crystalluria as an important warning sign in the management of these patients, and to develop a prospective evaluation of crystalluria value in patients exposed to high doses of SMX.
- The Renal Warrior Project. Join Now
- Source: https://www.nature.com/articles/s41598-024-56322-9