Incidence, Risk Factors and Impact on Renal Function of the Appearance of Crystalluria Under High-dose Intravenous Amoxicillin
Incidence, Risk Factors and Impact on Renal Function of the Appearance of Crystalluria Under High-dose Intravenous Amoxicillin
Intravenous high-dose amoxicillin is a frequently prescribed antibiotic in intensive care units and infectious disease departments, in such indications as meningitis, endocarditis, listeriosis, or documented susceptible osteo-articular infections.
Among possible side effects is the occurrence of amoxicillin crystalluria. Its incidence is unknown. It is deemed uncommon but literature is poor and its incidence is probably under-estimated. It can be complicated with hematuria or acute renal failure and may require renal replacement therapy. measures that can limit the impact of this nephrology crystalluria.
Since the beginning of 2014, because of the occurrence of a case of macroscopic crystalluria in our unit and a pharmacovigilance alert, it has been decided to search systematically for amoxicillin crystalluria in several centers of the north-east of France. The investigators propose to study these patients, through an observational study of a multicenter longitudinal cohort, to assess the incidence of amoxicillin crystalluria during high dose intravenous amoxicillin treatment.
Amoxicillin is an antibiotic of the family of β-lactams and sub-family penicillin A. It has a time-dependent bactericidal activity. It is mainly active on many bacterial species frequently found in infectious diseases. It can be administered orally or parenterally. It is therefore a frequently prescribed antibiotic, alone or in combination. Its elimination is mainly urinary in its active form. It is frequently prescribed intravenously at high doses in infectious disease departments, in indications like meningitis, endocarditis, listeriosis, or documented susceptible osteo-articular infections.
Several amoxicillin treatment side effects have been reported, like hypersensitivity reactions, gastrointestinal adverse reactions, mucocutaneous candidiasis, nervous system toxicity and eventually kidney disease, namely the occurrence of crystalluria or acute interstitial nephritis of whose frequency is undetermined.
Literature concerning the occurrence of crystalluria is mainly based on Case reports and there are no incidence data. Crystalluria is characterized by the presence of amoxicillin trihydrate crystals in urine, after amoxicillin urinary supersaturation. The appearance is that of large needles, isolated or aggregated into "broom bush-like" crystal, 50-300µm, with birefringent under polarized light. Amoxicillin crystalluria may be of varying importance, microscopic or macroscopic, sometimes up to urinary tract obstruction and may cause acute renal failure (ARF) and require renal replacement therapy. A low urine output could be associated with the occurrence of crystalluria, as well as a urine pH between 5 and 7, a high urinary density and high doses of amoxicillin.
Amoxicillin crystalluria is deemed uncommon. In the national pharmacovigilance database, in November 2013, there were only 9 reported cases, from 12 to 78 years old, 5 of whom severe one of which life-threatening. There is a high underreporting in these data, and it seems that this impact is highly underestimated for several reasons: amoxicillin crystalluria is not sought in routine clinical practice in all centers, even with high dose administrated. It is not clinically detectable if microscopic and even if a urine culture is performed, amoxicillin crystals may be mistaken for other crystals. In the investigator's experience automated systems do not always detect these crystals. Moreover, patients likely to receive high doses of amoxicillin are generally subject to severe infections and are more at risk of acute renal failure either due to background or previously initiated therapies (like the association with aminoglycosides), and the imputability of a crystalluria in the occurrence of an episode of acute renal failure among these patients is seldom considered.
Finally, a recent increase in the number of reported cases in the national pharmacovigilance database is an incentive to pay more attention to this complication.
The investigators also think that the occurrence of an amoxicillin crystalluria, even apart from extreme cases of macroscopic crystalluria complicated by oligo-anuric ARF, could increase the risk of intra-hospital ARF. The demonstration of such an effect associated with an incidence of significant crystalluria might justify the research in routine practice among patients receiving intravenous high-dose amoxicillin (IHDA).
The search for amoxicillin crystalluria is rather simple, by performing a direct microscopic examination of the urine. If in doubt about the diagnosis amoxicillin composition of these crystals, due to an atypical appearance or concomitant administration of other drugs that may generate similar crystals, the sample may be sent to a specialized laboratory for infrared spectroscopy to confirm that it is amoxicillin crystals.
IHDA treatments would be the situations with a higher risk for crystalluria. The highlighting of such an effect associated to a significant incidence of crystalluria might justify systematic research recommendations in routine practice with patients receiving IHDA. Following the occurrence in the infectious disease department of the Reims university hospital of a case of macroscopic crystalluria complicated by ARF requiring renal replacement and a pharmacovigilance alert, the search for amoxicillin crystalluria has been systematic as well as in several centers of the Grand East of France with patients receiving IHDA to have the possibility to watch these patients carefully at kidney level.
The investigators propose, through an observational study of a multicenter longitudinal cohort, to assess the incidence of amoxicillin crystalluria during IHDA, to search for the factors associated with the occurrence of this crystalluria and measure the impact of this crystalluria on the occurrence of an episode of intra-hospital ARF.
Data will be collected in patients' paper or computerized record at the Center.
Sociodemographic data.
Weight and size
Medical history
Date of beginning and end of hospitalization, date of beginning and end of antibiotic therapy
Ongoing treatments / Absorption of nephrotoxics or diuretics / Associated antibiotics
Indication of amoxicillin treatment
intravenous intake and urine output
Amoxicillin dosage and amoxicillin brand used and administration route
Biological data
Characteristics of a possible episode of acute renal failure occurring within 28 days following IHDA
Statistical aspects:
Justification of the number of subjects needed The impact of drug related crystalluria during amoxicillin treatment is unknown. Literature on this subject is composed of case reports. However a frequency lower than 5% during IHDA would not be of significant clinical interest. The number of subjects needed was then calculated with a 5% alpha risk, an expected frequency of 10% with an accuracy of 5%, leading to a required number of subjects of 139 patients. Taking into account 10% of possible exclusions, we plan to include 153 patients.
According to the observed incidence of crystalluria, an extension of the number of subjects needed will be discussed secondarily by investigators to obtain the necessary power to assess the impact of crystalluria on the risk of acute renal failure.
Statistical analysis :
Flowchart :
The synthesis of the patients included, excluded and analyzed will be expressed in the form of a flowchart.
Analysis of the primary endpoint:
It is a descriptive analysis of the cumulative incidence of amoxicillin crystalluria under IHDA. It will be expressed as a percentage and its 95% confidence interval.
c. Secondary endpoints : i. Descriptive analysis : Quantitative variables of normal distribution are described by their mean and standard deviation, and quantitative variables of non-normal distribution are described by their median and interquartile range. Qualitative variables are described by their frequency and percentages.
ii. Univariate analysis : Depending on the secondary endpoint, the variable to be explained (dependent variable) is the occurrence of crystalluria or the occurrence of acute renal failure. The measure of associations will be made by calculating odds ratios and their 95% confidence intervals.
iii. Multivariate analysis : Depending on the number of events, multivariate analysis with a binary logistic regression model will be proposed.
On the factors favoring the occurrence of crystalluria, the variables are those with a p-value less than 0.20 in univariate analysis. Amoxicillinemia will be forced into the model.
On the factors favoring the occurrence of acute renal failure, the variables are those with a p-value less than 0.20 in univariate analysis. The occurrence of crystalluria will be forced into the model.
The associations measurements will be made by calculating adjusted odds ratio and their 95% confidence interval.
Inclusion criteria:
The patients included will be all patients aged 18 years or more consecutive for which treatment by IHDA (> 150mg / kg / day) has been initiated in a unit participating in the study or within 48 hours before admission to this Service .
Criteria for non-inclusion:
Patients on dialysis or kidney transplant receiver will not be included.
Exclusion criteria:
Patients who did not receive treatment until at least the first urine direct examination and patients for whom it is not feasible for technical reasons will also be excluded.