Diversity and severity of adverse reactions to quinine: A systematic review
Abstract
Quinine is a common cause of drug-induced thrombocytopenia and the most common cause of drug-induced thrombotic microangiopathy. Other quinine-induced systemic disorders have been described. To understand the complete clinical spectrum of adverse reactions to quinine we searched 11 databases for articles that provided sufficient data to allow evaluation of levels of evidence supporting a causal association with quinine. Three reviewers independently determined the levels of evidence, including both immune-mediated and toxic adverse reactions. The principal focus of this review was on acute, immune-mediated reactions. The source of quinine exposure, the involved organ systems, the severity of the adverse reactions, and patient outcomes were documented. One hundred-fourteen articles described 142 patients with definite or probable evidence for a causal association of quinine with acute, immune-mediated reactions. These reactions included chills, fever, hypotension, painful acral cyanosis, disseminated intravascular coagulation, hemolytic anemia, thrombocytopenia, neutropenia, acute kidney injury, rhabdomyolysis, liver toxicity, cardiac ischemia, respiratory failure, hypoglycemia, blindness, and toxic epidermal necrolysis. One hundred-two (72%) reactions were caused by quinine pills; 28 (20%) by quinine-containing beverages; 12 (8%) by five other types of exposures. Excluding 41 patients who had only dermatologic reactions, 92 (91%) of 101 patients had required hospitalization for severe illness; 30 required renal replacement therapy; three died. Quinine, even with only minute exposure from common beverages, can cause severe adverse reactions involving multiple organ systems. In patients with acute, multi-system disorders of unknown origin, an adverse reaction to quinine should be considered. Am. J. Hematol. 91:461–466, 2016. © 2016 Wiley Periodicals, Inc.
Introduction
Quinine may be among the most common causes of severe drug-induced disorders. In our previous systematic reviews of published reports of drug-induced thrombocytopenia (DITP) and of drug-induced thrombotic microangiopathy (DITMA), quinine was the second most common cause of DITP, following only quinidine 1, 2, and the most common cause of DITMA, including reports describing TMA as thrombotic thrombocytopenic purpura (TTP) or hemolytic uremic syndrome (HUS) 2, 3. Quinine is also the most common cause of DITMA in the experience of the Oklahoma TTP-HUS Registry and the BloodCenter of Wisconsin 4. Recognizing the frequency and severity of quinine-induced thrombocytopenia and TMA, the U.S. Food and Drug Administration (FDA) has established stringent regulations related to medicinal quinine production and accessibility over the past 22 years 5.
Among the reports of quinine-induced TMA that we previously reviewed 3, many additional disorders were described in addition to the characteristic disorders of TMA (microangiopathic hemolytic anemia, thrombocytopenia, and acute kidney injury). They included chills, fever, and hypotension, neutropenia, disseminated intravascular coagulation (DIC), painful acral cyanosis, liver toxicity, respiratory failure, rhabdomyolysis, and systemic cutaneous reactions. The diversity of these unexpected additional disorders often caused diagnostic uncertainty and also may have obscured the quinine etiology. The sources of quinine were also diverse 3. These were not only “drug-induced” disorders; they were also caused by common quinine-containing beverages.
In this review, we evaluated the diversity and severity of adverse reactions to quinine as well as the diverse sources of quinine exposure. For each report, we determined the level of evidence supporting a causal association with quinine. The goal of this review was to inform clinicians about the extraordinary clinical spectrum and severity of adverse reactions to quinine. Diagnosis of a quinine etiology is essential to prevent recurrent adverse reactions.
Methods
Systematic review: Adverse reactions to quinine
Data sources and searches
Eleven databases were searched from their inception through July 2015 (Supporting Information Table I). The search was restricted to English language articles that described human subjects.
Article selection
One author (N.W.L.) initially screened titles and abstracts of all identified articles to select articles describing adverse reactions associated with quinine or cinchona products. Selected articles were categorized as reporting individual patients or group data. Articles reporting individual patients were also categorized as describing either immune-mediated or dose-dependent toxicity-mediated adverse reactions 6. Immune-mediated reactions were defined a priori as exposure to a therapeutic dose of quinine (or less, as in a quinine-containing beverage or lotion) daily for <4 weeks 7 or taken only intermittently. Toxic reactions were defined a priori as exposure to more than 2 g of quinine in <24 hr, or a serum quinine concentration more than 10 mg ml−1 8, 9. Reports that did not fulfill criteria for either idiosyncratic or toxic reactions were excluded. Excluded patients typically had been taking a daily therapeutic dose for many months or years, or the quinine regimen was not described. The complete texts of the included articles were then reviewed to determine if sufficient clinical data were reported to evaluate the level of evidence supporting a causal association with quinine.
Levels of evidence for a causal association with quinine
Levels of evidence supporting a causal association with quinine were initially determined by two independent reviewers (N.W.L., E.E.P.); disagreements were adjudicated by a third independent reviewer (J.N.G.). All reports that were determined to have definite or probable evidence for a causal association with quinine were reviewed again by the third reviewer (J.N.G.) to confirm the level of evidence. For reports of individual patients with immune-mediated reactions, evidence levels were determined using both the Oklahoma criteria 1, 3 and the Naranjo criteria, an established and widely used algorithm [10]. For patients with toxic reactions, different criteria were used for determining levels of evidence. For reports of group data, levels of evidence were based on study design and outcome. Oklahoma criteria used for assessing levels of evidence for individual patients and group data are presented in Supporting Information Table II.
Article analysis
Articles reporting individual patients with immune-mediated reactions and a definite or probable causal associations with quinine were reviewed again by the third reviewer (J.N.G.) to record the clinical features and the severity of the adverse reaction. Less severe symptoms that are commonly associated with quinine were not recorded. These included tinnitus, nausea, vomiting, diarrhea, abdominal pain, fever, diaphoresis, headache, blurred vision, and syncope. Severity of the adverse reactions was documented by requirement for hospitalization, severity of clinical features (such as requirement of transfusions, treatment for suspected sepsis, treatment with plasma exchange for TTP, and requirement for renal replacement therapy), and by death.
Survey: Adverse reactions to drugs other than quinine
This part of our project is described as a survey to distinguish it from the rigorous methodology of a systematic review (Supporting Information Table III). To estimate whether drugs other than quinine may have comparable diverse and severe adverse reactions, one of the authors (A.L.L.) surveyed the FDA Adverse Events Reporting System (FAERS) quarterly data summaries, 2008–2012 11 and the quarterly reports from the Institute for Safe Medication Practices (ISMP) 12. From these databases, she selected 50 drugs that had diverse and/or severe adverse reactions. Fifty drugs were selected to provide a sufficient number to identify those with the most diverse and/or severe adverse reactions. To estimate the diversity of organ systems involved in the adverse reactions, she searched the Drugs@FDA database 13 for each of these 50 drugs, recording adverse reactions involving any of the 16 organ systems listed in this database. For the drugs with the greatest diversity of adverse reactions, she searched the MEDLINE database via OVID, from its inception through April 2015, to determine the severity of the reactions. She searched for severe reactions involving three major organ systems (hematologic, kidney, liver). These three organ systems were selected because they represent the prominent organ systems affected by immune-mediated adverse reactions to quinine.
Results
Systematic review: Adverse reactions to quinine
Our search criteria identified 2,534 unique articles. After review of titles and abstracts, 383 articles were selected for full-text review. Following the full-text review, 140 articles reporting 186 patients with immune-mediated reactions, 10 articles reporting 10 patients with toxic reactions, and 97 articles reporting group data were determined to have sufficient information to assess levels of evidence supporting a causal association with quinine (Fig. 1). All of these articles described adverse reactions to quinine itself; none described adverse reactions to the common health products containing cinchona, the natural tree bark that is a source for quinine.
Systematic literature search and article selection. The combined database searches (Supporting Information Table I) identified 3,228 articles, before duplicates were removed. Some articles reporting multiple patients described patients with an immune-mediated adverse reaction or patients with a toxic adverse reaction, who were included in our review, and also patients who did not meet criteria for either immune-mediated or a toxic reactions, who were excluded from our review. Therefore the number of articles reporting patients in these three categories (190) exceeds the number of articles included for analysis 184.
The evaluations of the reports of the 186 patients with immune-mediated reactions by both the Oklahoma and the Naranjo criteria agreed (weighted kappa statistic, 0.663). The major difference between the two methods was for patients who were classified as definite by Oklahoma criteria but only probable by Naranjo criteria. All patients classified as definite by Naranjo criteria were also classified as definite by Oklahoma criteria (Supporting Information Table IV). The basis for this difference was that identification of quinine-dependent antibodies reactive with platelets and/or other cells established a definite causal association by Oklahoma critieria if the additional criteria for a probable causal association had been met. This was a common occurrence for reports of patients with thrombocytopenia or TMA. For the Naranjo criteria, a positive laboratory test was only one of ten components that were assessed; identification of quinine-dependent antibodies did not determine definite evidence 10. The Oklahoma criteria were used for all data reported in this study.
Definite or probable evidence supporting a causal association with quinine was established for 142 (76%) of 186 patients with immune-mediated reactions, for six (60%) of 10 patients with toxic reactions, and for 25 (26%) of 97 articles reporting group data (Table 1). The citations and the descriptions of all individual patients and group data with definite or probable evidence for a causal association with quinine are presented in Supporting Information Table V.
| Individual patients | Group data (97) | ||
|---|---|---|---|
| Evidence level | Immune-mediated (186) | Toxic (10) | |
| Definite | 112 | 6 | 22 |
| Probable | 30 | 0 | 3 |
| Possible | 43 | 4 | 66 |
| Unlikely | 1 | 0 | 6 |
- For adverse reactions in individual patients, the reactions were distinguished as immune-mediated or toxic as described in Methods. Levels of evidence for reports of individual patients with immune-mediated adverse reactions were determined using the Oklahoma criteria. Criteria for determining levels of evidence are presented in Supporting Information Table II.
Individual patients: Quinine-induced immune-mediated reactions
One hundred forty-two patients (114 articles) with immune-mediated reactions had definite or probable evidence supporting a causal association with quinine (Supporting Information Table Va). The reports of 112 patients provided definite evidence, which included the occurrence of the reaction within hours of quinine exposure, the presence of no other drugs, herbal remedies or other potential etiologies, and the occurrence of recurrent acute episodes following recurrent quinine exposures (86 patients) and/or the documentation of quinine-dependent antibodies (42 patients). The reports of 30 patients provided probable evidence, which included the occurrence of the reaction within hours of quinine exposure and the presence of no other drugs or herbal remedies exposure or other potential etiologies. For patients with probable evidence, recurrent acute episodes following recurrent quinine exposures or the documentation of quinine-dependent antibodies were not reported.
These 114 articles were published from 1865 14 through 2015. Forty-three (38%) of the 114 articles, describing 47 (33%) of the 142 patients, were published in 2000-2015. Patients were reported from 14 counties; 65 (46%) patients were from the US. Seventy-six (54%) of the 142 patients were women; their median age was 50 years (range, newborn infant to 84 years).
Table 2 illustrates the diversity of the organ system involvement in the 142 patients with immune-mediated reactions. Hematologic disorders were the most common and thrombocytopenia was the most common hematologic abnormality. Many patients had multiple organ systems affected. For example, one patient had fever, chills, and hypotension, microangiopathic hemolytic anemia, thrombocytopenia, disseminated intravascular coagulation, acute kidney injury, rhabdomyolysis, and chest pain with T-wave inversion 15. She, like many other patients with quinine-induced idiosyncratic reactions, was initially treated for sepsis.
| Hematologic (88 patients): thrombocytopenia (79), microangiopathic hemolytic anemia ( 33), neutropenia ( 24), disseminated intravascular coagulation ( 17), eosinophilia ( 4), autoimmune hemolytic anemia ( 2), lymphopenia ( 1), methemoglobinemia ( 1) |
| Dermatologic (44 patients): photosensitive eczema ( 10), photosensitive erythema ( 3), photosensitive dermatitis ( 2), photosensitive vasculitis ( 1), photosensitive lichen planus ( 1), fixed drug eruption ( 9), contact dermatitis ( 8), contact eczema ( 1), lichen planus ( 2), systemic erythematous rash ( 3), urticaria ( 1), toxic epidermal necrolysis ( 1), exfoliative dermatitis ( 1), mimic of mycosis fungoides ( 1) |
| Kidney (37 patients): thrombotic microangiopathy ( 33), interstitial nephritis ( 3), acute kidney injury ( 1) |
| Systemic reactions (15 patients): chills, fever, hypotension ( 11), chills, fever ( 2), painful acral cyanosis ( 2), anaphylaxis ( 1) |
| Liver toxicity (13 patients): elevated serum transaminases ( 13) (5 patients were biopsied documenting granulomatous hepatitis) |
| Neurologic (3 patients): stupor/obtunded ( 2), headache/meningismus ( 1) |
| Pulmonary (7 patients): pulmonary edema ( 6), hypoxemia requiring intubation ( 1) |
| Endocrine (3 patients): hypoglycemia ( 3) |
| Musculoskeletal (2 patients): rhabdomyolysis ( 2) |
| Cardiac (2 patients): chest pain, T wave inversion ( 1), pericarditis ( 1) |
| Rheumatologic (2 patients): systemic lupus erythematosus ( 1), Henoch-Schönlein syndrome ( 1) |
| Ophthalmologic (1 patient): Blind ( 1) |
- These 142 patients had definite (112) or probable ( 30) evidence for a causal association with quinine. The sum of patients described for the 12 organ systems (including the systemic reactions) (217) exceeds the total number of patients with idiosyncratic adverse reactions (142) because many patients had adverse reactions affecting multiple organ systems. Within each organ system, the number of patients with each type of adverse reaction is listed in parentheses. For the hematologic adverse reactions, the sum of the types of reactions (161) exceeds the number of patients (88) because many patients had multiple types of hematologic abnormalities. For the systemic adverse reactions, one patient had chills, fever, and hypotension and also painful acral cyanosis.
Table 3 presents the immune-mediated adverse reactions in the 142 patients according to their principal presenting features, as described in the case report. Eighty-six (60%) of the 142 patients had more than one episode of adverse reaction before the quinine etiology was suspected and diagnosed. Among the 41 patients who were reported because of their dermatologic reactions, none had other organ system involvement, only six were hospitalized, none died. Among the other 101 patients, 92 (91%) were hospitalized. The median nadir platelet count in patients with isolated thrombocytopenia was 3000/µL; five patients received red cell transfusions for severe gastrointestinal and/or vaginal hemorrhage. Among the 33 patients with quinine-induced TMA, 20 were treated with plasma exchange for the diagnosis of TTP. However 22 of these 33 patients had clinical features that were not typical for TTP or HUS; sepsis was the most frequent initial diagnosis in these patients, suggested by the presence of chills, fever, hypotension, neutropenia, DIC, and acral cyanosis. Sepsis was also the initial diagnosis in patients presenting with chills, fever, hypotension, and DIC without acute kidney injury. Thirty-seven patients were described as having acute kidney injury; 30 (81%) required renal replacement therapy. Three (3%) patients died related to their adverse reaction to quinine; all three had had previous serious adverse reactions for which a quinine etiology was not suspected. Two patients died from pulmonary hemorrhage during their second episode of quinine-induced thrombocytopenia 16, 17. One patient died during her third episode of quinine-induced TMA associated with DIC and rhabdomyolysis 18.
| Clinical presentation | Patients | Exposure | Additional clinical features |
|---|---|---|---|
| Thrombocytopenia | 38 | Pill ( 27), beverage ( 9), transplacental ( 1), IV heroin ( 1) | Neutropenia ( 6) |
| Thrombotic microangiopathy (includes thrombocytopenia, microangiopathic hemolytic anemia, acute kidney injury) | 33 | Pill ( 30) beverage ( 3) | DIC ( 11), neutropenia ( 10), chills, fever, hypotension ( 8), liver toxicity ( 5), rhabdomyolysis ( 2), painful acral cyanosis ( 2), pulmonary edema ( 2), hypoxemia requiring intubation ( 1), stupor/obtunded ( 2), headache/meningismus ( 1), chest pain with T wave inversion ( 1), methemoglobinemia ( 1), systemic erythematous rash ( 1) |
| Liver toxicity | 6 | Pill ( 5), beverage ( 1) | Neutropenia ( 3), thrombocytopenia ( 1), lymphopenia ( 1), eosinophilia ( 1), exfoliative dermatitis ( 1). |
| DIC | 5 | Pill | Neutropenia ( 3), chills, fever, hypotension ( 2) |
| Interstitial nephritis | 3 | Pill | Liver toxicity ( 2), eosinophilia ( 1) |
| Pulmonary edema | 3 | Pill | Neutropenia ( 1) |
| Hypoglycemia | 3 | Pill | None |
| Autoimmune hemolytic anemia | 2 | Pill | DIC ( 1), acute kidney injury ( 1), pulmonary edema ( 1), systemic erythematous rash ( 1) |
| Neutropenia | 2 | Pill | Chills, fever, hypotension ( 1) |
| Chills, fever | 2 | Pill | Eosinophilia ( 1) |
| Anaphylaxis | 1 | Beverage | None |
| Systemic lupus erythematosus | 1 | Pill | Thrombocytopenia, neutropenia, pericarditis |
| Henoch-Schönlein syndrome | 1 | Pill | Thrombocytopenia, eosinophilia |
| Blindness | 1 | Pill | None |
| Dermatologic | 41 | Pill ( 17), beverage ( 14), topical lotion ( 7), factory exposure ( 2), pessary ( 1) | None |
- These 142 patients had definite (112) or probable ( 30) evidence for a causal association with quinine. They are presented according to their principal clinical presentation as described in the original case reports. Many patients had multiple additional clinical features. Among patients with thrombocytopenia, exposure to “pill” includes three patients taking a liquid quinine remedy in 1865 ( 14). Beverages included tonic water, gin and tonic, Schweppes Bitter Lemon, and Dubonnet.
In 102 (72%) of the 142 patients, the source of quinine exposure was a pill, taken for leg cramps or other minor symptoms. In 28 (20%) patients the source was a beverage, typically gin and tonic or tonic water itself. Bitter lemon and Dubonnet were other reported beverage sources of quinine. Quinine exposures were diverse among the other 12 patients. One patient with thrombocytopenia following intravenous heroin had quinine detected in his urine and blood; quinine-dependent, platelet-reactive antibodies were subsequently identified. It was then learned that quinine is a commonly used adulterant for heroin 19. One newborn infant had quinine-induced thrombocytopenia caused by quinine pills given to her mother to induce labor; her mother also had quinine-induced thrombocytopenia 20. In seven patients who were reported because of dermatologic reactions, quinine-containing topical lotions were responsible. In one man, perineal and genital eczema were attributed to his wife's quinine-containing contraceptive pessary 21. A factory that purified quinine had an outbreak of dermatitis attributed to quinine in 13 workers 22; two workers who were described in detail were included in our review.
Individual patients: Quinine-induced toxicity reactions
Among the six patients with definite evidence supporting a causal association with a dose-dependent toxic reaction (no reported patients had probable evidence), blindness occurred in five; partial vision recovered in all patients (Supporting Information Table Vb). One patient developed acute kidney injury, anemia, thrombocytopenia, and seizures; her recovery was apparently complete.
Group data: Quinine-induced adverse reactions
Descriptions of adverse reactions to quinine as group data were primarily in reports of clinical trials of antimalarial treatments (Supporting Information Table VC). Symptoms were those characteristically described as cinchonism 8: tinnitus, dizziness, diaphoresis, nystagmus, hypoglycemia, and vertigo. Three reports described cardiac abnormalities (increased QRS or QT intervals).
Survey: Adverse reactions to drugs other than quinine
To estimate whether other drugs may have diverse and/or severe adverse reactions comparable to quinine, we selected 50 drugs from the FAERS quarterly data summaries 11 and the ISMP quarterly reports 12 that had been reported with diverse and severe adverse reactions. Fifteen of these 50 drugs, in addition to quinine, were reported in the Drugs@FDA database 13 to cause adverse reactions in 12 or more organ systems (Supporting Information Table IIIa). Six of these 16 drugs, in addition to quinine, (ciprofloxacin, clozapine, minocycline, omeprazole, trimethoprim-sulfamethoxazole, valproic acid) have been reported to cause severe adverse reactions involving one or more of three major organ systems: hematologic, kidney, liver. Quinine had the most reports of hematologic adverse reactions; ciprofloxacin had the most reports of adverse reactions involving kidney function; valproic acid had the most reports of liver toxicity (Supporting Information Table IIIb).
Discussion
The goal of our systematic review was to document the diversity and severity of adverse reactions to quinine. We reviewed reports of both individual patients and group data. Among the individual patient reports, we reviewed both immune-mediated and dose-dependent, toxic adverse reactions. The adverse reactions reported in group data were toxic reactions, related to the narrow margin between therapeutic and toxic blood levels 8, 9. Dose-dependent toxic reactions, often described as cinchonism 8, are well known, typically mild and reversible. Therefore we focused our review on reports of patients with immune-mediated reactions for which there was definite or probable evidence, determined by the Oklahoma criteria 1, 3, for a causal association with quinine. The Oklahoma criteria were validated by documentation of agreement with the Naranjo criteria, an established and widely used algorithm [10].
The immune-mediated adverse reactions to quinine were remarkably diverse. They involved twelve organ systems. Often multiple organ systems were involved in individual patients. The clinical diversity of quinine-induced disorders creates a diagnostic dilemma. In very few of the published reports was a quinine etiology described as the initial diagnosis. Most reports described repeated systemic illnesses that were only retrospectively appreciated to be associated with repeated quinine exposure. Even in patients who were suspected of having an adverse drug reaction, explicit questions regarding quinine exposure were essential. Patients did not consider that a quinine pill taken only occasionally was important information for their physician 23, 24 and did not consider that quinine in beverages was relevant. The difficulty of diagnosis suggests the possibility that the quinine etiology of severe systemic disorders may often be unrecognized.
Not only was there diversity of clinical features, there was also diversity of the exposures to quinine. Although most reports described reactions following a quinine pill, quinine-containing beverages were the cause of the adverse reactions in 28 (20%) of the 142 patients. The adverse reactions to quinine-containing beverages were not different from reactions to quinine pills regarding diversity of clinical features and severity of the adverse reactions. The diversity of exposures, which may be unique for quinine, can also obscure recognition of a quinine etiology.
Immune-mediated adverse reactions to quinine were not only diverse, they were also very severe. Excluding the 41 patients with dermatologic reactions, 92 (91%) of 101 patients were hospitalized. Many patients were initially diagnosed with sepsis and treated with parenteral antibiotics. Severe thrombocytopenia caused severe hemorrhage requiring red cell transfusions in five patients. Twenty patients were treated with plasma exchange for a diagnosis of TTP. Thirty-seven (37%) of these 101 patients had acute kidney injury; 30 (81%) of these 37 patients required renal replacement therapy. Acute kidney injury requiring renal replacement therapy increases the risk for progressive chronic kidney disease by 28-fold and doubles the long-term risk for death 25. Three patients died as a result of their adverse reaction to quinine. All three patients had had previous episodes of their quinine-induced disorder: two had had previous quinine-induced thrombocytopenia; one had had two previous episodes of quinine-induced TMA and DIC.
The clinical features of these patients were consistent with an immunologic etiology. There was a prior exposure to quinine and the symptoms began suddenly, within hours after the triggering exposure to quinine. Even the small amounts of quinine in beverages could trigger severe reactions. Quinine-dependent antibodies reactive with platelets, neutrophils, and lymphocytes have been documented to be associated with thrombocytopenia, neutropenia, and lymphopenia 26-28. One patient had quinine-induced thrombocytopenia followed 6 months later by quinine-induced TMA. Quinine-dependent antibodies at the time of thrombocytopenia reacted only with platelets while at the time of the TMA, quinine-dependent antibodies reacted also with endothelial cells 29. Another report described a patient with TMA and severe neutropenia who had quinine-dependent antibodies that reacted with neutrophils, causing increased adhesion to endothelial cells 26. These observations are consistent with observations that immune-mediated endothelial dysfunction may contribute to the acute kidney injury of quinine-induced TMA 30.
Quinine has been commonly studied to understand the molecular mechanisms of drug-dependent antibody formation 31-33. Quinine-dependent antibodies may be derived from a pool of naturally-occurring antibodies that are weakly reactive with autologous proteins 31, 33. Quinine may have unique amphipathic properties that allow it to become integrated into complementarity-determining regions of naturally occurring antibodies, creating a hybrid paratope that greatly increases binding affinity to cell surface antigens 32. Quinine may also integrate into quasi-stable regions of cell surface integrins to alter the conformation of a target epitope, causing increased binding affinity of the naturally-occurring antibodies 34. Quinine's molecular properties may explain why it is a common cause of drug-induced disorders.
Because quinine was an established medicine before enactment of the US Food, Drug, and Cosmetic Act in 1938, it was not initially regulated by the FDA and was available without a prescription. Because of reports of severe hematologic reactions, the FDA prohibited over-the-counter marketing of quinine in 1994 and subsequently determined that quinine would require a new drug application. In 2005 Qualaquin® was approved as the only medicinal quinine, indicated only for the treatment of malaria. The FDA ordered the manufacturing of unapproved medicinal quinine products to be stopped 5. These regulations have dramatically decreased the number of prescriptions for quinine pills in the US 5. However similar regulations have not occurred in other countries. For example, in Canada pills containing 50 mg or less of quinine are considered as a natural health product and are available without a prescription 35.
FDA regulation of quinine pills does not affect quinine-containing lotions and shampoos 36 and quinine-containing beverages, which remain commonly available. Quinine is not only an ingredient of familiar beverages such as tonic water and bitter lemon, it is currently popular for its flavor and fluorescent properties in newer party beverages described as “Shocktails” 37. In 2004, the FDA was authorized by the Food Allergen Labeling and Consumer Protection Act to require that major food allergens be identified by labeling. Quinine may not be a major food allergen, like peanuts, but labeling of quinine as a potential risk may help to increase awareness of quinine sensitivity, among physicians as well as the public. Increased awareness may improve recognition and diagnosis of adverse reactions to quinine.
A limitation of our review is that it was based on case reports which may not have contained complete data. As the reports typically featured dramatic events, they may not be representative of adverse reactions to quinine and they cannot support an estimate of the frequency of quinine-induced disorders. Long-term outcomes were rarely described but long-term major morbidities may be common. Although case reports are commonly characterized as providing only low-quality scientific evidence, for this review they were sufficient. The issue of article quality is not relevant to the presentation of primary clinical data. Our criteria for determining levels of evidence required complete patient information to establish a definite or probable causal relation of quinine with the adverse reaction. Our conclusions in this review were based only on reports with sufficient data to establish a definite or probable causal association with quinine.
Another limitation is that the methodology of our survey of adverse reactions to other drugs does not provide data of comparable quality to our systematic review. However, once we recognized the extraordinary diversity and severity of adverse reactions to quinine, we wanted to explore the issue of whether other drugs may cause similarly diverse and severe reactions. This survey seemed appropriate to provide preliminary data for future investigations.
The strength of our review is the systematic analysis of all published reports on adverse reactions to quinine. These reports provide a dramatic illustration of the diversity and severity of quinine-induced disorders. All physicians must be aware of these disorders. With accurate diagnosis, recurrence can be prevented simply by avoidance of quinine, in any amount and any form.
