Hypersensitivity reactions to fluoroquinolones

INTRODUCTION — This topic will review hypersensitivity (allergic) reactions to fluoroquinolones. A general discussion of fluoroquinolones, including adverse effects other than hypersensitivity reactions, is found separately. (See "Fluoroquinolones".)

TYPES OF HYPERSENSITIVITY REACTIONS — Hypersensitivity reactions, or allergic reactions, are immunologic reactions to a drug. There are several types of hypersensitivity reactions, which are broadly divided into immediate and delayed reactions based on the time between administration and the appearance of signs or symptoms [1]. Immediate reactions begin within one hour of the first administered dose, and delayed reactions begin later than one hour (and sometimes days) after the first dose. The intent of this classification system is to distinguish immediate reactions from other types since immediate reactions are more likely to progress to potentially life-threatening anaphylaxis if the patient is reexposed to the causative drug. Unfortunately, patients often do not recall the chronology of drug reactions, and thus, it is often not possible to classify reactions based on timing alone. (See "An approach to the patient with drug allergy".)

Most common — The two most common types of hypersensitivity reactions to fluoroquinolones are:

●Delayed-onset maculopapular cutaneous eruptions

●Immediate reactions, which are characterized by urticaria, pruritus, flushing, angioedema, wheezing, nausea, abdominal cramping or diarrhea, and/or hypotension

These two types of fluoroquinolone-induced hypersensitivity reactions are relatively common, and, in some studies, the incidence of fluoroquinolone-related anaphylaxis is comparable with that due to beta-lactams [2,3]. The remainder of this topic review will focus on these two forms of hypersensitivity.

Other types — A variety of other uncommon hypersensitivity reactions have been attributed to fluoroquinolones, including fixed drug eruptions, drug fever, vasculitis, serum sickness-like syndromes, drug reaction with eosinophilia and systemic symptoms/drug-induced hypersensitivity syndrome (DRESS/DiHS), symmetrical drug-related intertriginous and flexural exanthema (SDRIFE), Stevens-Johnson syndrome (SJS), and toxic epidermal necrolysis (TEN) [4-9]. There is no diagnostic testing for these types of reactions, and avoidance of the causative drug is uniformly recommended.

SJS and TEN are severe and potentially life threatening, and any future reexposure, which can result in recurrence, is contraindicated, except in extreme circumstances. (See "Stevens-Johnson syndrome and toxic epidermal necrolysis: Pathogenesis, clinical manifestations, and diagnosis".)

The classification and pathogenesis of drug hypersensitivity reactions are reviewed in greater detail separately. (See "Drug hypersensitivity: Classification and clinical features" and "Drug allergy: Pathogenesis".)

APPROACH TO THE PATIENT — Evaluation of the patient with a past drug reaction always begins with a careful clinical history and an assessment of the type of hypersensitivity reaction the patient most likely experienced. Important questions for the patient with drug allergy are reviewed separately. (See "An approach to the patient with drug allergy", section on 'Clinical history'.)

A thorough history of the reaction is essential since many patients labeled as allergic to fluoroquinolones actually experienced nonallergic adverse reactions, such as tendon inflammation or damage, myalgias, neurologic side effects, or gastrointestinal symptoms. In some cases, such patients should continue to avoid the causative drug, but they are not candidates for allergy evaluation. Nonallergic adverse effects of fluoroquinolones are reviewed separately. (See "Fluoroquinolones", section on 'Adverse effects'.)

Multiple drug allergy syndrome — The term "multiple drug allergy syndrome" is applied to patients who report hypersensitivity reactions to two or more classes of medications. A subset of these patients reacts only to antibiotics. Typically, such patients have reacted to commonly used antibiotics, such as penicillins, sulfonamides, fluoroquinolones, and macrolides and have limited antibiotic options remaining to treat infections. Evaluation of one or more of these antibiotics to determine whether the same or related drugs are actually tolerated can be very helpful in avoiding the use of expensive broad-spectrum antibiotics and in reducing anxiety on the part of the patient and providers. In such individuals, the management goal is the identification of two or more classes of antibiotics that the patient can safely receive, focusing on those that are most clinically useful.

Referral — Referral to an allergy expert with experience in evaluation of drug-allergic reactions is appropriate in any of the following situations:

●The patient has reactions to numerous antibiotics (eg, multiple drug allergy syndrome), and treatment options are becoming limited.

●The patient experienced a life-threatening reaction to a fluoroquinolone and requires advice about how best to prevent reexposure and whether related antibiotics can be safely administered.

●The patient experienced a serious reaction in the context of receiving multiple drugs (eg, perioperative anaphylaxis), and the cause of the reaction is not clear.

●The patient has a high likelihood of needing future treatment with a problematic drug due to the nature of their underlying medical problems (eg, fluoroquinolones in chronic urologic problems).

A careful clinical history can avoid unnecessary referrals. Many patients labeled as allergic to fluoroquinolones experienced nonallergic adverse reactions, such as tendon inflammation or damage, myalgias, neurologic side effects, or gastrointestinal symptoms. In some cases, such patients with adverse reactions should continue to avoid the causative drug, but they are not candidates for allergy evaluation. Nonallergic adverse effects of fluoroquinolones are reviewed separately. (See "Fluoroquinolones", section on 'Adverse effects'.)

Immediate-type skin testing and challenge procedures are often required in the evaluation of antibiotic hypersensitivity. These procedures should be performed by allergy specialists. Although these procedures are considered safe, they are not without risk and should be performed in a setting prepared to identify and treat possible allergic reactions, including anaphylaxis. In addition, any procedure that involves the deliberate exposure of a patient to a substance to which they may be allergic should be undertaken by someone with specific training in these techniques.

Unlike diagnostic tests for penicillin allergy, methods for evaluating fluoroquinolone hypersensitivity are not standardized or validated, and allergy experts may have different approaches to the use of fluoroquinolone skin testing and challenge.

DELAYED MACULOPAPULAR EXANTHEMA — The most common type of reaction to fluoroquinolones is a delayed-onset (defined as greater than one hour after the first administered dose) maculopapular eruption or exanthema that is nonpruritic. Such eruptions occur in 2 to 3 percent of patients treated with fluoroquinolones and are likely mediated by drug-specific T cells [10-13]. These reactions are self-limited and generally benign and do not feature urticaria, angioedema, fever, systemic symptoms, involvement of the mucous membranes or other organs, or skin desquamation.

Delayed exanthema may occur with any of the fluoroquinolones, although certain agents (eg, gemifloxacin) have particularly high rates, and the incidence is higher with prolonged durations of therapy and possibly in females [14]. As an example, the incidence of maculopapular rash in patients treated with gemifloxacin (who were receiving the drug for an infection) ranged from 1.2 percent for 5-day therapy to 7.4 percent for 14-day therapy [11]. Females <40 years old were found to have a higher incidence of exanthema compared with other groups of patients (2.1 and 22.6 percent for 5- and 14-day therapy, respectively) [11].

Cross-reactivity — The development of an exanthema with one fluoroquinolone does not necessarily predict recurrence with others. One of the most illustrative studies about delayed cutaneous eruptions to fluoroquinolones was an unpublished double-blind, double-dummy, crossover study conducted in healthy female volunteers <40 years old (not in need of antibiotic therapy) treated with 10 days of gemifloxacin or ciprofloxacin [15]. This study specifically focused on younger females and was described in a review but has not been published [11]:

●In the first part of the study, rashes developed in a total of 260 of 819 (32 percent) and 7 of 164 (4.3 percent) females treated with gemifloxacin and ciprofloxacin for 10 days, respectively [11]. The vast majority of the rashes were maculopapular, and the most common finding on biopsy was uncomplicated exanthem (mild superficial perivascular lymphocytic infiltrate). There was no evidence of vasculitis, bulla formation, or necrosis. Blood and urine tests were unremarkable, with no eosinophilia, hepatitis, or other abnormalities.

●In the second part of the same study, the 260 females who developed exanthema to gemifloxacin were given ciprofloxacin, and only 10.4 percent developed a rash [11]. It is unknown whether this low degree of cross-reactivity between gemifloxacin and ciprofloxacin can be generalized to other combinations of fluoroquinolones. Interestingly, some females who initially tolerated ciprofloxacin for 10 days went on to develop a rash when treatment was repeated for another 10 days.

Treatment — Discontinuation of the culprit fluoroquinolone is the primary treatment for maculopapular exanthema. The median time to resolution of the rash is five days [11], and symptoms may worsen for a few days, even after the drug is stopped [16]. Oral antihistamines may be administered at age-appropriate doses if there is a pruritic component. (See "Exanthematous (maculopapular) drug eruption", section on 'Management'.)

Evaluation — After resolution of an exanthema to a fluoroquinolone, evaluation is appropriate, although there is a lack of consensus about the best approach. In the United States, these reactions are usually diagnosed based on the clinical history (and physical findings if the reaction is witnessed by a clinician). Photographs or detailed descriptions in the medical record are invaluable for future reference. Further evaluation is not usually performed.

The largest study to undertake challenge of patients with history of delayed reactions to fluoroquinolones evaluated 66 patients [17]. No skin testing was performed, and all of the historical reactions occurred less than one year prior to evaluation. Upon being challenged with the same quinolone that caused the previous reaction, 63 of 66 patients tolerated the antibiotic.

Patch testing for the evaluation of delayed maculopapular exanthema from fluoroquinolones is not useful. In a study of 101 patients with fluoroquinolone allergy, 37 had delayed reactions (defined in this study as symptoms appearing greater than six hours after the last dose) [5]. Approximately one-half of these reactions began within 24 hours of the first dose. All 37 patients underwent patch testing, and all had negative results. Twenty-eight underwent fluoroquinolone graded challenges with the culprit drug, and two patients reacted with recurrence exanthema, both after the fourth dose. Thus, patch testing was not helpful for identifying patients at risk for this type of reaction.

Graded challenge — Since cross-reactivity among fluoroquinolones for delayed maculopapular reactions appears to be low [11], it is reasonable to perform graded challenge in patients with a history of a self-limited maculopapular exanthema, with a fluoroquinolone different from the one that caused the reaction. However, it is important to understand the limitations of challenge procedures for delayed maculopapular exanthema. Many of these reactions only recur after several days of therapy, and there is no practical way to elicit such reactions without administering a full therapeutic course. This should be explained to patients.

The author's approach to challenge in this setting is to administer 1/10^th of a full dose, wait one hour, and, if there is no adverse reaction, then administer the remaining 9/10^ths of the dose and observe for an additional hour. The primary rationale for giving the initial dose under observation is to reassure the patient that the drug can be safely ingested. Otherwise, there may be continued reluctance to use it in the future. A delayed reaction may occur during the 24 hours following the initial dose, and patients should be instructed to inform the clinician if this occurs. In the study mentioned previously, 28 patients with delayed reactions were challenged, and the two who reacted developed symptoms at 15 and 20 hours after the last of the graded doses [5].

After receiving a full dose under observation, patients who do not need the antibiotic at that time can simply take it normally in the future. Patients who need fluoroquinolone therapy at the time can continue the treatment for a full course.

If a patient develops a mild exanthema during future courses of therapy, it may be reasonable to attempt to continue the drug and add antihistamine therapy provided that the reaction lacks any worrisome features. This approach has not been formally studied and is only appropriate when the benefit of continued treatment clearly outweighs the risk of further reaction. Such situations require clinical judgement. (See "Exanthematous (maculopapular) drug eruption", section on 'Management'.)

REACTIONS OF UNCERTAIN TYPE — If a previous reaction to a fluoroquinolone was poorly documented, if the clinical history is uncertain, or if the reaction involved a pruritic skin component, then the possibility of an immediate reaction should be considered. (See 'Immediate reactions' below.)

IMMEDIATE REACTIONS — Immediate reactions are characterized by symptoms that begin within one hour of the first administered dose. Immediate reactions result from activation of mast cells and basophils. Common signs and symptoms include urticaria; pruritus; flushing; angioedema of the face, extremities, or laryngeal tissues (leading to throat tightness with stridor or, rarely, asphyxiation); wheezing; nausea; abdominal cramping or diarrhea; hypotension; and/or anaphylactic shock. Fatal immediate reactions to fluoroquinolones have been reported [18].

Pathogenesis — There is evidence for both immunoglobulin E (IgE) and non-IgE mechanisms for immediate reactions caused by fluoroquinolones. Animal studies have shown that fluoroquinolones at supratherapeutic doses can cause histamine release and/or generalized rash [19,20]. This nonspecific mast cell degranulation caused by fluoroquinolones has been shown to be due to interaction with the surface receptor Mas-related G protein-coupled receptor X2 (MRGPRX2) [21]. Unlike IgE-mediated allergic reactions, non-IgE-mediated reactions may occur with first exposure since prior sensitization is not necessary. Apart from this difference, IgE-mediated and non-IgE-mediated reactions may be clinically indistinguishable, and their relative incidence is not known.

IgE-mediated reactions — There is very limited evidence of fluoroquinolone-specific IgE in patients with past immediate reactions. In the most convincing study, fluoroquinolone-specific IgE was detected in 30 of 55 patients by radioimmunoassay (RIA), although at very low levels [22]. Inhibition studies were performed to demonstrate that the IgE was functional, although only on a few sera.

Natural history — There is some evidence that fluoroquinolone-specific IgE declines over time, implying that some patients could lose the allergy if not reexposed for a period of time. In the study (mentioned above) of 55 patients with convincing histories of immediate-type reactions to fluoroquinolones occurring 1 to 48 months prior to evaluation, 30 of the 55 had detectable drug-specific IgE by RIA [22]. In those patients who had detectable drug-specific IgE, an average of 9.7 months had elapsed since the reaction compared with 15.6 months in those without detectable IgE. Also, among the 30 IgE-positive patients, those with recent reactions (defined as less than eight months prior to evaluation) had significantly higher RIA values than those with more distant reactions. Although none of the patients were challenged with fluoroquinolones to confirm allergic status, these data suggest that fluoroquinolone allergy will eventually resolve in some patients.

Based on a small number of studies of patients with immediate reactions to fluoroquinolones who underwent graded challenge with the culprit drug, most of the evidence indicates that a majority of patients tolerated the drug upon reexposure [5,23,24]. Unfortunately, the time elapsed between the reaction and challenge was only specified in one [23]. However, another relatively large study demonstrated the opposite finding (ie, that all patients reacted), which may have been due to the fact that the evaluations were performed soon (median of five months) after the allergic reactions [25]. These studies are described below:

●In a study of 33 patients who were evaluated only with challenge, 24 (73 percent) tolerated the culprit drug [23]. Reactions occurred 3 to 120 months before evaluation (median 12 months).

●In the study of 101 patients discussed above, 51 patients with immediate-type reaction histories were challenged, and 45 (88 percent) tolerated the drug (most of whom were skin test negative) [5]. The median time elapsed between the initial reaction and evaluation was not specified.

●In the study of 71 patients already discussed, 17 were challenged, and 11 tolerated the drug [24]. The median time elapsed between reaction and evaluation was not specified.

●In a study of 152 patients, a basophil activation test (BAT) was performed, and only those patients with negative BAT results were challenged [17]. Skin testing was not performed. Twenty-four patients had a positive BAT and were not challenged. The remaining 128 patients underwent graded challenge with the culprit fluoroquinolone, and 42 reacted. Therefore, a total of 66 of 152 (43 percent) of the patients were confirmed to be allergic (42 via challenge and 24 via positive BAT). This is a surprisingly low value given that all of the patients had a convincing reaction history less than one year prior to evaluation and that approximately one-half of the historical reactions were anaphylaxis.

●In a study of 120 patients, BAT and prick skin testing was performed, and only those patients with both negative BAT and skin test results were challenged. Twenty-three patients underwent graded challenge with the culprit fluoroquinolone, and all reacted (pruritus or urticaria), but none of the reactions required epinephrine. The median time elapsed between the initial reaction and evaluation was five months [25].

Incidence — Fluoroquinolone-induced anaphylactic reactions are increasingly reported, although the incidence is not precisely known [5,22,26-30].

●The incidence of these reactions was estimated to be between 1 and 5 per 100,000 prescriptions filled in a large Unites States health insurance claims database, in which anaphylaxis was confirmed by review of medical records [2]. The rates for anaphylaxis to penicillins and cephalosporins were comparable (2 of 100,000 and 1 of 100,000, respectively). However, the nature of this study did not allow confirmation that the antibiotic was responsible for the anaphylactic events. Also, it is possible some of the reactions to fluoroquinolones were anaphylactoid (non-IgE mediated), rather than anaphylactic (IgE mediated).

●An estimate of the number of anaphylactic reactions to ciprofloxacin during its first year of use (1987 to 1988) in the United States was 12 reactions among 972,000 prescriptions or 1.2 in 100,000 prescriptions filled [31].

Another study of a database of voluntary reports estimated the incidence to be lower, at 0.2 to 3.3 per 1 million doses [32]. However, voluntary reporting of adverse drug reactions likely underestimates the true incidence of reactions.

Cross-reactivity among drugs — Cross-sensitization and cross-reactivity to other fluoroquinolones in patients with previous immediate reactions has been demonstrated, but no clear patterns of cross-sensitization have been identified:

●In the study discussed previously, there were 30 of 55 patients with immediate reactions and demonstrable drug-specific IgE by RIA, of whom 24 (80 percent) were positive to more than one fluoroquinolone, and nine reported historical allergic reactions to more than one fluoroquinolone [22]. Although subjects were not challenged in this study to confirm reactivity, this suggests cross-sensitivity. Another study using similar methods found a comparable rate of fluoroquinolone cross-reactivity based on in vitro IgE testing [26].

●In two large clinical studies assessing potential fluoroquinolone cross-reactivity using identical retrospective analyses of hospital electronic records (and by the same authors), 182 inpatients with a history of fluoroquinolone allergy were treated with a different fluoroquinolone (for example, a patient who reported an allergy to ciprofloxacin was treated with levofloxacin) [33,34]. These patients did not undergo allergy testing to confirm allergy to the index drug. Of these, 12 (6.6 percent) experienced a reaction, and none were severe or required administration of epinephrine. Approximately 80 percent of the index reactions were urticarial, but 13 patients had previous fluoroquinolone-related anaphylaxis. The only severe cutaneous adverse reaction (SCAR) was a single patient with acute generalized exanthematous pustulosis (AGEP). There was no clear pattern of cross-reactivity.

●In a number of case series, patients with past reactions to one fluoroquinolone were challenged with other fluoroquinolones [25,27,28,35-39]. Of 78 total reported patients, 40 reacted to at least one other fluoroquinolone, whereas 38 patients tolerated one or more different fluoroquinolones. Similarly, overall, no clear patterns of cross-reactivity among the drugs were detectable.

Given that even patients with a history of allergic reactions to a given fluoroquinolone are often able to tolerate the same quinolone when later evaluated (see 'Evaluation' below), it is not clear whether tolerance of dissimilar fluoroquinolone in these studies indicates lack of cross-reactivity or rather proof of no allergy to any fluoroquinolone.

Management — The management of immediate reactions to fluoroquinolones involves discontinuation of the drug, intramuscular epinephrine for symptoms suggestive of anaphylaxis, intravenous fluid resuscitations, and antihistamines for pruritus and urticaria. This is reviewed in detail separately. (See "Anaphylaxis: Emergency treatment", section on 'Immediate management'.)

Evaluation — Patients with previous immediate reactions to fluoroquinolones are challenging to evaluate. The pathophysiology of these reactions is not well understood, immediate-type skin testing with these drugs is not validated, and practice parameters are lacking. In addition, there is evidence for both of the following statements:

●Patients with past immediate reactions to one fluoroquinolone may react to others, although the rates of cross-reactivity are unknown, and clear patterns of cross-reactivity have not been demonstrated.

●A substantial proportion of patients with prior immediate reactions to a specific fluoroquinolone will tolerate the culprit drug if they are given it again [5,23,24]. Potential explanations for this observation include the possibility that the patient's reaction was not IgE mediated and was instead dependent on other variables or that the patient's reaction was IgE mediated but drug-specific IgE declined over time [5,23,24]. (See 'Natural history' above.)

In most cases, evaluation consists of immediate-type skin testing and graded challenge, although there is legitimate debate about the utility of skin testing [5]. The approach of the author is presented here.

Other types of tests remain investigational. Specifically, we are aware of no published studies reporting the accuracy of commercial assays for fluoroquinolone-specific IgE, and BATs and lymphocyte stimulation tests are either not widely available or have not been adequately validated.

Situations in which allergy testing is helpful — Evaluation for fluoroquinolone allergy is not indicated in every patient, as it may be adequate to simply avoid this class of antimicrobials altogether in some patients. The author's experience is that evaluation is most helpful in patients who report immediate reactions that were less severe than full anaphylaxis, such as pruritic rashes, urticaria, or angioedema and who are likely to require fluoroquinolone therapy in the future. As examples:

●Evaluation would be helpful in a 70-year-old female with recurrent, frequent urinary tract infections who has been hospitalized for pyelonephritis and reports an itchy rash while on levofloxacin.

●Evaluation is not necessary in a 40-year-old male with no significant medical problems who developed an urticarial reaction while on ciprofloxacin for a vasectomy-related infection but who otherwise rarely requires antibiotic treatment.

For patients with past fluoroquinolone-associated anaphylaxis, one can distinguish, somewhat arbitrarily, between those with recent reactions (eg, within the past five years) and those with more remote reactions:

●We do not perform skin testing or challenge for patients with recent anaphylaxis, since skin testing is not sufficiently reliable and a challenge would pose risk to the patient. If such a patient required treatment with a fluoroquinolone antibiotic and there were no acceptable alternatives, it should be administered using rapid desensitization protocol (also known as rapid induction of drug tolerance) [40,41]. (See "Rapid drug desensitization for immediate hypersensitivity reactions".)

●For patients with anaphylaxis more than five years ago, it is reasonable to perform skin testing based on limited evidence that drug-specific IgE declines with time, and, thus, the patient may no longer be allergic [22]. (See 'Natural history' above.)

Immediate-type skin testing — Immediate-type skin testing with fluoroquinolones is not validated or standardized, and it may not distinguish between IgE-mediated and non-IgE-mediated immediate reactions, since fluoroquinolones can cause nonspecific mast cell degranulation. Further, the allergenic determinants are not known, and some may be reactive metabolites. Because of these limitations, some experts do not advocate skin testing with fluoroquinolones [5,39].

Despite the above, it is the author's opinion that the negative predictive value of skin testing is sufficiently high that it can be clinically useful in some cases, as described in the following section. (See 'Interpretation of results' below.)

If skin testing is performed, prick-puncture skin testing should be performed first. Positive histamine and negative saline controls should also be applied. If prick-puncture testing is negative, intradermal skin testing is then performed. Skin test results are read at 15 minutes.

Concentrations for skin testing — It is very important to use nonirritating concentrations for intradermal testing to avoid false-positive results since these drugs can cause nonspecific activation of mast cells. It is the author's approach to perform skin testing with commonly used fluoroquinolones for which nonirritating concentrations are available. Thus, the author tests with both levofloxacin and moxifloxacin in most cases. In other countries, different fluoroquinolones may be more commonly used, and these could be included in skin testing.

The author uses the following maximal concentrations for both prick-puncture and intradermal testing:

●Levofloxacin – An intravenous preparation of levofloxacin diluted to a concentration of 0.025 mg/mL (a 1:1000-fold dilution of the commercially available 25 mg/mL solution) was nonirritating in 25 subjects with no history of drug allergy [42].

●Moxifloxacin – The best concentration of moxifloxacin for use in skin testing is not entirely clear. In 16 patients without a history of quinolone allergy, prick testing with full-strength commercially available intravenous moxifloxacin (1.6 mg/mL) was negative in all subjects. Intradermal skin testing with serial 10-fold dilutions was negative in all 16 subjects at 1:1000 (0.0016 mg/mL); positive in 2 of 16 at 1:100 (0.016 mg/mL); and, among the remaining 14 skin test-negative subjects, 2 of 14 were positive at 1:10 (0.16 mg/mL) [43]. The author uses 0.016 mg/mL as a reasonable compromise.

●Ciprofloxacin – A nonirritating concentration of ciprofloxacin has not been identified, and the author does not perform skin testing to ciprofloxacin in his practice. In the study of 25 patients mentioned previously, at least one subject had a positive skin test response with even a 10,000,000-fold dilution [42]. However, in another study, a concentration of 0.0067 mg/mL was found to be nonirritating, although only 16 patients were tested, and the evaluation of more patients may have revealed irritant reactions [44].

Utility of skin testing — A small number of studies provide information about the utility of immediate-type skin testing in identifying patients at risk for immediate fluoroquinolone reactions [5,24]. In the largest, fluoroquinolone prick-puncture skin testing had a negative predictive value of 93 percent, sensitivity of 50 percent, and specificity of 93 percent [5]. This study included 101 patients with past hypersensitivity reactions to fluoroquinolones within one year of evaluation, of whom 64 had immediate reactions [5]. These 64 underwent prick-puncture skin testing using moxifloxacin at 1.6 mg/mL and levofloxacin at 5 mg/mL. Intradermal skin testing was not performed, since the investigators were unable to identify nonirritating concentrations in control subjects for several of the drugs. Of 64 patients with immediate reactions, 12 were skin test positive, and 52 were negative. Six of the 12 with positive skin testing underwent fluoroquinolone graded challenges (including to the culprit drug), and three developed reactions (mostly urticarial, none of which required epinephrine). Among the other 52 skin test-negative patients with immediate reactions, 45 were challenged, and 3 reacted. Thus, 6 of 64 patients (9 percent) with past immediate reactions were confirmed to be persistently allergic (although 13 were not challenged).

Interpretation of results — Based on the studies described above, most patients with immediate reactions who undergo fluoroquinolone skin testing will have negative results provided nonirritating doses of drug are used. We use the information obtained from skin testing as follows:

●A positive skin test indicates drug-specific IgE may be present, and the available studies indicate that the positive predictive value is approximately 50 percent [5,24,45]. It is the author's approach to advise patients with positive results to avoid all fluoroquinolones because of possible cross-reactivity among fluoroquinolones in causing immediate reactions. If the patient requires treatment with that fluoroquinolone and there are no acceptable antibiotic alternatives, it should be administered via a rapid desensitization. (See 'Cross-reactivity among drugs' above and 'Desensitization' below.)

●A negative skin test provides some reassurance but does not definitively exclude IgE-mediated allergy, because the allergenic determinants have not been defined and some active metabolites may cause reactions [45]. History-positive patients with negative fluoroquinolone skin testing should undergo graded challenge as described in the next section. Based on the study reviewed above, approximately 7 percent of patients would be expected to react upon challenge [5].

Graded challenge — Graded challenge is the cautious administration of a drug to a patient who is unlikely to be allergic to it [45]. Negative fluoroquinolone skin testing should be followed by graded challenge in order to confirm lack of allergy (or, if the challenge is positive, to demonstrate that the skin test was falsely negative). Otherwise, patients or other clinicians may misinterpret the skin test result to mean that fluoroquinolone allergy has been conclusively ruled out, which is not the case.

Because of the uncertain utility of skin testing with fluoroquinolones, another option for evaluating patients with immediate reactions is to forego skin testing and perform a graded challenge with a fluoroquinolone other than the one that caused the reaction. The author's approach is to challenge with an alternate equally efficacious fluoroquinolone since cross-reactivity is approximately 50 percent based on a limited number of challenges reported in the literature [27,28,35-39]. This is a reasonable approach in patients with isolated urticaria and/or angioedema. The author does not do this in patients with anaphylaxis, although others have reported performing challenges in such patients safely [5,37].

Fluoroquinolone graded challenge typically starts with 1/100^th of the full dose and includes two to four steps until the full dose is reached. The degree of caution exercised is determined by the severity of the historical reaction and the time elapsed since the reaction. For example, an outpatient graded challenge with levofloxacin for a patient with immediate urticaria in the past could be carried out under continuous supervision with 5 mg, 50 mg, and then 500 mg of levofloxacin (oral solution), each given one hour apart. The patient is observed for two hours after the final dose.

Desensitization — Desensitization is appropriate for patients who require retreatment with a fluoroquinolone and had [40,41]:

●Anaphylaxis to a fluoroquinolone within the recent past (ie, five years)

or

●A prior immediate reaction that was less severe than anaphylaxis but skin testing to the culprit fluoroquinolone was positive

Desensitization involves the administration of a drug beginning with very small doses (usually 1/1000^th of a normal dose or less) and gradually increasing in a stepwise manner until a full therapeutic dose is reached. It is often performed in a hospital setting when the patient requires a fluoroquinolone for a serious infection. Desensitization alters a patient's adverse immune response to a drug and results in temporary tolerance, allowing the patient with a hypersensitivity reaction to receive an uninterrupted course of the medication safely. Once the medication is discontinued, the patient's hypersensitivity to the medication returns. Therefore, desensitization must be repeated each time the patient requires a course of the drug in question. (See "Rapid drug desensitization for immediate hypersensitivity reactions".)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Drug allergy and hypersensitivity".)

SUMMARY AND RECOMMENDATIONS

●Types of hypersensitivity reactions – The two most common types of hypersensitivity reactions to fluoroquinolones are delayed-onset maculopapular exanthema, which are generally benign, and immediate reactions (urticaria, pruritus, flushing, angioedema, wheezing, nausea, abdominal symptoms, and/or hypotension/shock), which can be life threatening. (See 'Types of hypersensitivity reactions' above.)

●Whom to evaluate and refer – Evaluation of a patient with a past drug reaction always begins with a careful clinical history and an assessment of the type of hypersensitivity reaction the patient most likely experienced. Referral to an allergy expert is appropriate if the patient has experienced a serious reaction, has an underlying medical condition that will likely require fluoroquinolone therapy in the future, or reports multiple drug allergies and has limited drug options left. (See 'Approach to the patient' above.)

●Delayed maculopapular exanthema – Fluoroquinolones cause a delayed-onset maculopapular, nonpruritic rash in 2 to 3 percent of patients. These reactions are likely mediated by drug-specific T cells. In most cases, no specific testing is useful. Cross-reactivity among the fluoroquinolones for this type of reaction appears to be low. For patients who require fluoroquinolone therapy in the future, we suggest identifying a safe alternative using a graded challenge with a fluoroquinolone, other than the one that caused the reaction (Grade 2C). (See 'Delayed maculopapular exanthema' above.)

●Rare immediate reactions – Fluoroquinolones can cause rare immediate reactions, including anaphylaxis, through both immunoglobulin E (IgE) and non-IgE-mediated mechanisms. There is evidence of cross-reactivity among fluoroquinolones in patients with immediate reactions, but no clear patterns have been distinguished. Diagnostic tests for immediate fluoroquinolone hypersensitivity are not validated or standardized, in contrast to tests for penicillin hypersensitivity. If skin testing is performed, it is critical to use nonirritating concentrations of the drugs in question since fluoroquinolones can cause irritant false-positive reactions. (See 'Immediate reactions' above.)

•We suggest that patients with anaphylaxis in the recent past (ie, less than five years) empirically avoid all fluoroquinolones (Grade 2C). We do not perform skin testing or challenge in such patients. If a fluoroquinolone is absolutely required in the near future, it should be administered using a desensitization protocol. (See 'Desensitization' above.)

•Skin testing for immediate reactions is not standardized or well validated, although we find it helpful in some clinical circumstances. Skin testing and challenge procedures are most useful in patients who experienced relatively mild immediate reactions, such as pruritic rashes, urticaria, or angioedema or whose reactions occurred in the remote past (more than five years) and who are likely to require fluoroquinolone therapy in the future. (See 'Evaluation' above.)

-For patients with negative skin test results, we suggest a graded challenge be performed to confirm that the patient tolerates the drug (Grade 2C). (See 'Interpretation of results' above and 'Graded challenge' above.)

-For patients with positive skin test results, we suggest avoidance of all fluoroquinolones (Grade 2C). If a fluoroquinolone is absolutely required in the near future, it should be administered using a desensitization protocol. (See 'Interpretation of results' above and 'Desensitization' above.)