Introduction

Fosfomycin, discovered in Spain in 1969, is a cell wall-acting antibacterial drug that inhibits the formation of N-acetylmuramic acid, a precursor of peptidoglycan. Fosfomycin has been available in Europe since the 1970s and was approved in the United States in 1996 [1]. In the US, it is approved as a 3-g sachet of fosfomycin tromethamine to be given orally as a single dose for uncomplicated cystitis due to susceptible strains of Escherichia coli and Enterococcus faecalis. In Europe, fosfomycin is also available as fosfomycin disodium for intravenous (IV) administration and is used for various infections at doses of 12–16 g/day (up to 24 g/day), divided into 3–4 doses.

In vitro, fosfomycin is active against Enterococci including vancomycin-resistant strains, E. coli, methicillin-resistant Staphylococcus aureus, Citrobacter diversus, Citrobacter freundii, Enterobacter aerogenes, Klebsiella oxytoca and Klebsiella pneumoniae, including some carbapenem-resistant strains, Proteus mirabilis and vulgaris, and Serratia marcescens [2]. Some extended spectrum beta-lactamase-producing strains of E. coli and K. pneumoniae are susceptible to fosfomycin [3, 4]. Fosfomycin is variably active against Pseudomonas aeruginosa and Acinetobacter spp. [2, 4, 5].

The oral formulation of fosfomycin is considered to have a favorable safety profile with gastrointestinal disturbances being the most commonly associated adverse event [1, 2]. An IV formulation of fosfomycin, fosfomycin disodium, is associated with a high sodium intake which could be a limitation in patients with heart failure or those on hemodialysis [2]. Adverse reactions such as angioedema, aplastic anemia, cholestatic jaundice, and hepatic necrosis have been reported postmarketing [1].

Because of the growing problem of antibacterial resistance, there has been an increased interest in fosfomycin use, especially its parenteral formulations [6]. This review summarizes the safety profile of fosfomycin using the Food and Drug Administration (FDA) Adverse Event (AE) Reporting System (FAERS) and published literature with an emphasis on AEs associated with its parenteral formulations.

Methods

The FAERS database was queried for domestic and foreign cases of AEs reported with oral or IV formulations. The FAERS contains information on AE and medication error reports submitted to FDA. AEs and medication errors are coded to terms in the Medical Dictionary for Regulatory Activities (MedDRA) terminology.

We conducted three FAERS searches. The first search included reports regardless of whether fosfomycin was implicated as the primary suspect. The period for the search was from 1996 through September 2012. Data retrieved by this search were subjected to a disproportionality analysis (henceforth referred to as a ‘data mining analysis’). This analysis aims to detect over-represented associations of drug–event combinations in the FAERS. The analysis used Empirica Signal® software (version 7.3.3.3.2.359, Oracle, Redwood Shores, CA, USA) and the Multi-item Gamma Poisson Shrinker (MGPS) data mining algorithm for all events associated with fosfomycin use [7]. MGPS generates adjusted relative reporting ratios, also known as Empirical Bayes Geometric Mean (EBGM) values, for the entire FAERS database. The EBGM is the value of a ratio of the observed to expected number of AEs indicating the strength of reporting relationships among all drug–AE combinations in the database. Importantly, the process does not adjust for reporting bias.

The EBGM value provides a stable estimate of the relative reporting ratio of any AE for a particular drug relative to all other drugs and AEs in FAERS. MGPS also calculates lower and upper 90% confidence intervals (CI) for the EBGM scores, denoted as EB05 and EB95, respectively. The higher the EBGM value for a particular drug–AE combination, the higher is the reporting association between that drug and AE in the database. Drug–AE pairs with an EB05 (lower bound of the 90% CI for the EBGM) >1 indicate AEs that occur above the expected rate. Furthermore, one may estimate that events with an EB05 >2 occur at least twice the expected ratio relative to the other drugs and events in the database. For our data mining analysis, we selected reports of events with an EB05 >1.

Subsequently, we conducted a FAERS search for reports of AEs where fosfomycin was considered to be the primary suspect [8].

A separate FAERS search was conducted for all cases of possible bone marrow toxicity with the terms of agranulocytosis, neutropenia, febrile neutropenia, aplastic anemia, bone marrow failure, aplasia pure red cell, and pancytopenia associated with the use of fosfomycin. This separate query was prompted by a case of aplastic anemia that was reported to the FDA. Aplastic anemia was defined as a combination of hemoglobin level <10 g/dl, segmented polymorphonuclear and band cells count <1.5 × 109/l, platelet count <100 × 109/l), and histological evidence of decreased cellularity, absence of infiltration and absence of significant fibrosis on bone marrow examination [9].

We then supplemented the results of the FAERS searches by the literature review of the safety profile of fosfomycin. The review was conducted by searching articles in English with the term “fosfomycin” via PubMed. We conducted a systematic review aiming to evaluate for possible imbalances in the frequency of AEs associated with fosfomycin as compared to other antibacterial drugs with the focus on fosfomycin AEs over-represented in the data-mining analysis and on safety signals associated with IV administration of fosfomycin.

The reports of clinical trials, meta-analyses, systematic reviews, and case reports of AEs associated with fosfomycin were selected. Individual trials were selected if they included greater than 10 patients. For parenteral formulations of fosfomycin, all clinical trials were selected for analysis and for oral formulations only prospective comparative trials were selected. References of the selected articles were also reviewed. The database was most recently accessed on July 9, 2015.

Compliance with Ethics Guidelines

This article is based on previously conducted studies and does not involve any new studies of human or animal subjects performed by any of the authors. No identifiable patient data were provided or accessed.

Results

Data mining for all AEs with EB05 >1 associated with fosfomycin that were reported to FAERS identified a total of 559 reports for 69 events. Two events had an EB05 score greater than 8—pseudomembranous colitis, 5 reports, EBGM 57.4 and agranulocytosis, 16 reports, EBGM 25.8. Two events had an EB05 score from >4 to 8—toxic skin eruption, 7 events, EBGM 16.1 and hepatocellular injury, 10 reports, EBGM 10.

Eighteen events had EB05 score from >2 to 4. These AEs were mainly related to hypersensitivity reactions, liver and bone marrow toxicity. AEs related to hypersensitivity reactions included drug reaction with eosinophilia and systemic symptoms, 6 reports, EBGM 9.2; Stevens Johnson syndrome, 11 reports, EBGM 6.7; eosinophilia, 8 reports, EBGM 6.7; drug eruption, 6 reports, EBGM 5.4; urticaria, 18 reports, EBGM 3.6.

AEs with EB05 score from >2 to 4 related to liver toxicity included hepatitis, 14 reports, EBGM 6.1; hepatic function abnormal, 12 reports, EBGM 4.8; alanine aminotransferase increased, 16 reports, EBGM 4.6; aspartate aminotransferase increased, 13 reports, EBGM 3.8.

AEs with EB05 score from >2 to 4 related to bone marrow toxicity included pancytopenia, 13 reports, EBGM 5.7; leukopenia, 16 reports, EBGM 5.4; thrombocytopenia, 19 reports, EBGM 4.4; neutropenia, 13 reports, EBGM 4.

The other AEs with EB05 score from >2 to 4 included blood urea increased, 10 reports, EBGM of 4.6; blood lactate dehydrogenase increased, 8 reports, EBGM 4.2; disseminated intravascular coagulation, 7 reports, EBGM 4.2; hypokalemia, 8 reports, EBGM 3.9; and pyrexia, 33 reports, EBGM 3.1.

Review of the events with EB05 score from >1 to 2 revealed that many of these events were related to hypersensitivity reactions including anaphylactic reaction (n = 5), anaphylactic shock (n = 4), toxic epidermal necrolysis (n = 4), dermatitis exfoliative (n = 6), acute generalized exanthematous pustulosis (n = 3), drug hypersensitivity (n = 6), hypersensitivity (n = 8), vascular purpura (n = 3), rash maculopapular (n = 6), rash erythematous (n = 6), dermatitis allergic (n = 3), and face edema (n = 8).

Subsequent FAERS query for events where fosfomycin was considered to be the primary suspect identified a total of 146 reports. There were 31 domestic and 115 foreign reports; in 132 cases fosfomycin was given by oral and in 5 cases by IV or intramuscular routes; the route of administration was not reported in 9 reports. In the majority of cases, fosfomycin was used for the treatment of urinary tract infections (UTI) (n = 131).

Table 1 presents the most frequently reported AEs (case count of >3). Events related to lack of efficacy or those that were deemed non-significant are not included. Gastrointestinal disorders and allergic reactions were reported most frequently.

Table 1 Most frequently reported MedDRA preferred terms
Full size table

There were a total of 38 reports of allergic reactions associated with fosfomycin, although no reports of severe skin reactions such as Stevens Johnson syndrome or drug rash with eosinophilia and systemic symptoms were retrieved by this query.

The seven cases of fetal toxicities reported included stillbirth (1), pyelocalyceal dilation (1), interventricular septal defect and micrognathia (1), spina bifida (1), hydrocele (1), congenital pulmonic stenosis and patent foramen ovale (1), and cranial bone defect with exencephaly (1).

The FAERS search for all cases of cytopenia associated with fosfomycin identified 50 reports. Analysis of these cases identified one case that met criteria for aplastic anemia.

The majority of cases described isolated decreases in white blood cell count (n = 25). Time to onset of cytopenia ranged from 1 to 116 days with a median of 11 days. All patients received other medications that could be implicated in the development of bone marrow toxicities. There was no obvious correlation between development of cytopenia and a total dose of fosfomycin and in several cases cytopenia developed after a single dose of the drug.

One case involved a 68-year-old female who took one fosfomycin 3-g sachet for UTI. Concomitant medications included itraconazole and ciprofloxacin. One week after receiving fosfomycin the patient was found to have a decrease in platelet count to 70 × 109/l. On day 17 after fosfomycin administration platelet count was 27 × 109/l, WBC was 1.7 × 109/l and hematocrit was 33. Bone marrow aspiration reported changes “consistent with aplastic anemia”. Filgrastim and epoetin alfa were administered and hematological abnormalities resolved by day 44 after fosfomycin dosing.

Literature Search

The search resulted in selection of 20 trials of parenteral fosfomycin and 25 trials of oral fosfomycin. Details of the retrieved articles are presented in Tables 2 and 3.

Table 2 Adverse events associated with parenteral use of fosfomycin in individual trials
Full size table
Table 3 Adverse associated with oral use of fosfomycin in individual trials
Full size table

The trials of parenteral fosfomycin reported on 1242 patients enrolled in 8 comparative, 7 prospective non-comparative, and 8 retrospective non-comparative trials. A total of 2052 patients were enrolled in prospective comparative trials of oral formulations of fosfomycin. Our review included 6 trials in 254 pediatric patients (3 trials of parenteral and 3 of oral fosfomycin) and 5 trials in 291 pregnant patients (all received oral fosfomycin).

The eight comparative trials of parenteral fosfomycin enrolled 664 patients in the fosfomycin and 626 patients in the comparator arms [10–17]. Seven trials were randomized controlled trials [10, 12–17] and one was a retrospective study [11]. Fosfomycin was used for the treatment of pneumonia including ventilator-associated pneumonia [12, 17], primary bacteremia [17], complicated intra-abdominal infection, UTI [17], skin and soft tissue infections [17], acute hematogenous osteomyelitis in children [11], prevention of infection after urological surgery [13], prophylaxis of surgical site infection after elective total knee arthroplasty [14] and colorectal surgery [10, 15, 16]. Fosfomycin was administered up to 3 weeks at a dose up to 12 g but the majority of patients in these trials (530/664) received fosfomycin for 1–3 days as prophylaxis for surgical site infections.

Parenteral fosfomycin was given in combination with other antibacterial drugs in 6 comparative trials [10–12, 15–17]; antibacterial drugs that were used in combination with fosfomycin and those that were used as comparators are presented in Table 2. The cumulative rate of AEs was similar in fosfomycin and comparator arms. The most common AEs observed in these trials are presented in Table 4.

Table 4 Adverse events associated with parenteral fosfomycin in 8 comparative trials [10–17]
Full size table

AEs observed in 15 non-comparative trials of parenteral fosfomycin [18–32] are presented in Table 5. These trials enrolled 578 patients, mainly adults. in combination with other antibacterial drugs in 9 trials (Table 5). Fosfomycin was used at a daily dose of up to 24 g for the treatment of various infections including bloodstream, central nervous system infections and ventilator-associated pneumonia. The duration of treatment ranged from 4 days to 2 months.

Table 5 Adverse events associated with parenteral administration of fosfomycin in 15 non-comparative trials [18–32]
Full size table

AEs associated with oral administration of fosfomycin are presented in Table 6. The majority of these trials studied a single dose of fosfomycin for the treatment of uncomplicated UTIs. In two trials of prophylaxis of recurrent UTI, fosfomycin was administered for 3–6 months [33, 34]. The most frequently reported adverse reactions (8%) were gastrointestinal disorders. More details are provided in Table 6.

Table 6 Adverse events associated with oral administration of fosfomycin in 28 prospective comparative trials [33–59, 71]
Full size table

One placebo-controlled trial for prophylaxis of recurrent uncomplicated UTI evaluated safety of a prolonged exposure to oral fosfomycin [33]. The trial compared 166 females treated with 3 g of oral fosfomycin every 10 days for 6 months with 155 subjects treated with placebo. After a follow-up of 360 days including 6 interim evaluations there were 2 (1%) AEs (rash and mild dyspnea) in the treatment arm as compared with 4 events in the placebo arm. Hematology and chemistry laboratory parameters at the end of the study did not show any significant difference between the two arms.

Trials conducted in pediatric population included 3 retrospective trials of parenteral fosfomycin (n = 118) [11, 25, 29] and 3 prospective randomized trials of oral fosfomycin (n = 134) [44, 49, 50]. In trials of parenteral fosfomycin, the drug was given up to 4 weeks for the treatment of acute hematogenous osteomyelitis, bacteremia, and lung infection. Oral fosfomycin was administered as a single time dose for the treatment of UTI. There were also several studies of parenteral [18, 24, 26–28] and oral [28, 59, 60] fosfomycin that included children that did not present safety data in children separately. Four pediatric trials were comparative [11, 44, 49, 50]. Overall, no specific safety issues related to the use of fosfomycin in children were identified.

Five trials of a total of 291 patients reported on the use of fosfomycin in pregnancy for the treatment of asymptomatic bacteriuria [36, 51, 52, 54, 56]. One of these trials in 153 patients specifically indicated that no serious fetal AEs were observed [36].

Our review of the literature also identified several reviews reporting on the efficacy and safety of fosfomycin. A meta-analysis of 27 randomized controlled trials compared fosfomycin (n = 2188) with other antibacterial drugs (n = 2052) for the treatment of cystitis [61]. The authors reported that fosfomycin had a comparable safety profile with the evaluated comparators in non-pregnant women, mixed and pediatric populations, and was associated with fewer AEs in pregnant women [61].

Information on the efficacy and safety of fosfomycin was provided by the Japan Research Committee of Fosfomycin in a report evaluating 1191 and 503 patients who received oral and IV fosfomycin, respectively [62]. For the oral formulation, AEs were observed in 122 of 1191 cases (10%) and included mainly gastrointestinal disturbances.

The report notes that side effects were observed in 17% of patients and included aspartate aminotransferase and glutamic-pyruvate transaminase elevations (although infrequent), pain at the injection site, gastrointestinal disorders, palpitations, and rash.

Mayama et al. analyzed postmarketing experience with oral fosfomycin in Japan by reviewing clinical records of 35,481 patients who were prescribed fosfomycin calcium capsules in 1981–1986 [63]. The overall incidence of side effects was 3.5%. The incidence of side effects was higher at a daily dose >3 g and was not dependent on the duration of administration. The most common side effects were related to gastrointestinal disorders (2.7%), hepatobiliary disorders (0.2%), and skin disorders (0.2%). There was one case of thrombocytopenia and one case of anemia.

There have been a few reports of liver toxicity associated with fosfomycin [64, 65]. An acute painful hepatomegaly, transaminase elevation and a hyperechogenic liver on ultrasound were reported in a 30-year-old female with cystic fibrosis and normal baseline liver function tests and liver ultrasound [64]. The liver abnormalities reoccurred three times during repeated courses of IV fosfomycin for Pseudomonas aeruginosa bronchitis. Laboratory and ultrasound abnormalities resolved every time after discontinuation of fosfomycin.

Another report of fosfomycin-induced liver injury described a 50-year-old male with no underlying liver diseases who developed mixed hepatocellular and cholestatic liver injury 3 days after being started on fosfomycin (route is not reported) [65]. His alanine aminotransferase peaked at 12.8 times the upper limit of normal and total bilirubin increased to 42 μmol/L (normal range <26 µmol/L). Liver function tests normalized 1 week after the withdrawal of the drug.

There have been four case reports of anaphylaxis associated with fosfomycin administration [66–68]. All reported cases were associated with oral formulations of fosfomycin and occurred within minutes to an hour after fosfomycin intake.

Discussion

Fosfomycin has been in clinical use for about 25 years and its safety profile has been previously described [6, 61, 69]. Our review did not identify new safety concerns related to fosfomycin. The most frequent AEs associated with parenteral administration of fosfomycin included rash, peripheral phlebitis, hypokalemia, and gastrointestinal disorders. Gastrointestinal disorders were the most common AEs associated with oral administration.

While the initial results of data mining for all reports of AEs associated with fosfomycin in the FAERS database revealed a higher than expected frequency of reports of agranulocytosis, liver injury, severe skin reactions, and pseudomembranous colitis, subsequent searches for adverse reactions implicating fosfomycin as the primary suspect as well as the literature review did not suggest an association of the drug with these AEs. Serious AEs of cytopenias, liver toxicities, and hypersensitivity reactions were not more frequent in fosfomycin-treated patients in comparative trials and were not common in non-comparative trials.

A separate search and detailed analysis of all FAERS reports of cytopenia identified one case that met pre-defined criteria of aplastic anemia that was associated with the use of oral fosfomycin. No cases of aplastic anemia associated with either oral or parenteral administration of fosfomycin were found by literature search. Considering more than 40-year experience with the drug, fosfomycin-associated aplastic anemia seems to be a rare event.

With regard to liver toxicities, we did not see an imbalance between fosfomycin- and comparator-treated patients in the comparative trials. However, several cases of liver toxicities associated with fosfomycin have been reported and monitoring of liver functions in patients receiving parenteral fosfomycin when the drug is administered for several days may be warranted.

The rates of hypersensitivity reactions were also comparable in fosfomycin- and comparator-treated patients in comparative trials and also four cases of anaphylactic reactions associated with fosfomycin have been reported. Of note, the package insert for oral fosfomycin lists the AEs of aplastic anemia, angioedema, cholestatic jaundice, and hepatic necrosis in the Adverse Reactions/Postmarketing Experience section [1].

Important safety consideration for parenteral fosfomycin is its high sodium content which may result in sodium overload and heart failure [23, 70]. Thus, each gram of fosfomycin contains 14.35 mEq (330 mg) of sodium. Considering that an average dose of fosfomycin is about 12 g, the sodium load associated with its use may be significant, especially for patients with underlying heart failure. In comparison, piperacillin and tazobactam, the antibacterial drug that is considered to have high sodium content contains 2.36 mEq (54.28 mg) of sodium per gram of piperacillin. A pooled analysis of comparative trials in our review did not demonstrate a higher rate of heart failure in fosfomycin-treated patients. However, a publication that specifically addressed AEs associated with IV fosfomycin reported that 6% of patients experienced cardiovascular AEs related due to sodium overload [23]. Hypokalemia was the most common AE in this study (26%). Another publication reports on the case of fosfomycin-associated heart failure that promptly resolved after the discontinuation of the drug [70].

Although, fosfomycin was suspected to be the cause of fetal toxicities in 7 cases in the FAERS database, the data mining search demonstrated EB05 scores of 1–2 for these types of AEs, indicating an overall expected frequency of reporting of this adverse reaction. Our review of the literature also did not identify cases of fetal toxicities. Of note, oral fosfomycin is labeled as pregnancy category B.

Our analyses have several limitations. FAERS contains spontaneously submitted data on AEs by the public and reporting biases such as under- and over-reporting of drug events can occur. Because of the spontaneous nature of reporting, the results of this analysis should not be interpreted as a formal comparison of treatment groups or of their relative risks and does not allow one to determine adverse event incidence rates.

The effects of concomitant illnesses or therapy cannot be fully controlled in this data mining analysis. Other factors such as the length of time of marketing, drug usage, and changes in coding practices over time should also be considered when interpreting these data mining results. Given the limitations inherent in the FAERS data, the high scores do not prove causality or an increased relative risk of the drug–AE in all patients exposed to fosfomycin. Another limitation of the FAERS search is that only 5 reports with parenteral administration of fosfomycin were retrieved.

Limitations of the literature review are related to a small number of randomized comparative prospective trials for parenteral fosfomycin and that in 5 out of 7 of these trials the drug was given for up to 3 days for prophylaxis of surgical site infection [10, 13–16]. Only 64 patients were treated with fosfomycin in randomized prospective comparative trials [12, 17]. Another comparative trial was retrospective and enrolled 70 patients in the fosfomycin arm [11].

Conclusion

In conclusion, no new safety concerns related to fosfomycin have been identified by this review. In comparative trials serious AEs such as cytopenia, anaphylaxis, and liver toxicities were not more common in fosfomycin-treated patients as compared with those treated with comparator. The most frequently reported AEs associated with parenteral administration of fosfomycin were rash, peripheral phlebitis, hypokalemia, and gastrointestinal disorders. High sodium content of parenteral fosfomycin should be taken into consideration in treating patients with underlying heart disease. An important limitation of the available safety data regarding the parenteral formulation of fosfomycin is that only in a very small number of randomized comparative trials was parenteral fosfomycin administered for a duration that would be expected in the treatment of the majority of infections. Oral formulations of fosfomycin were mostly associated with gastrointestinal disturbances.