ORIGINAL RESEARCH ARTICLE
`
`Drug Safety 2005; 28 (6): 547-556
`0114-5916/05/0006-0547/$34.95/0
`
` 2005 Adis Data Information BV. All rights reserved.
`
`Drug-Induced Anaphylaxis
`Case/Non-Case Study Based on an Italian
`Pharmacovigilance Database
`
`Roberto Leone,1 Anita Conforti,1 Mauro Venegoni,2 Domenico Motola,3 Ugo Moretti,1
`Ilaria Meneghelli,1 Alfredo Cocci,2 Giulia Sangiorgi Cellini,3 Stefania Scotto,2
`Nicola Montanaro3 and Giampaolo Velo1
`1 Clinical Pharmacology Unit, Reference Centre for Education and Communication within the
`WHO Programme for International Drug Monitoring, University of Verona, Verona, Italy
`2 Lombardy Centre of Pharmacovigilance, Milan, Italy
`3 Department of Pharmacology, Interuniversity Research Centre for Pharmacoepidemiology,
`University of Bologna, Bologna, Italy
`
`Abstract
`
`Objective: To identify the number of cases of anaphylaxis reported in association
`with different classes of drugs and compare it with other reports contained in the
`same database.
`Methods: The data were obtained from a database containing all of the spontane-
`ous reports of adverse drug reactions (ADRs) coming from the Italian regions of
`Emilia Romagna, Lombardy and the Veneto, which are the main contributors to
`the Italian spontaneous surveillance system. The ADRs reported between January
`1990 and December 2003 with a causality assessment of certainly, probably or
`possibly drug related (according to the WHO criteria) were analysed using a case/
`non-case design. The cases were defined as the reactions already coded by the
`WHO preferred terms of ‘anaphylactic shock’ or ‘anaphylactoid reaction’ (this
`last term also included anaphylactic reaction) and those with a time of event onset
`that suggested an allergic reaction and involved at least two of the skin, respirato-
`ry, gastrointestinal, CNS or cardiovascular systems; the non-cases were all of the
`other ADR reports. The frequency of the association between anaphylaxis and the
`suspected drug in comparison with the frequency of anaphylaxis associated to all
`of the other drugs was calculated using the ADR reporting odds ratio (ROR) as a
`measure of disproportionality.
`Results: Our database contained 744 cases (including 307 cases of anaphylactic
`shock with 10 deaths) and 27 512 non-cases. The percentage of anaphylaxis cases
`reported in inpatients was higher than that among outpatients (59.1% vs 40.9%).
`This distribution is significantly different from that of the other ADR reports that
`mainly refer to outpatients. After intravenous drug administrations, anaphylactic
`shock cases were more frequent than anaphylactoid reactions or other ADRs, but
`more than one-third of these reactions were caused by an oral drug. Blood
`substitutes and radiology contrast agents had the highest RORs. Among the
`systemic antibacterial agents, anaphylaxis was disproportionally reported more
`often for penicillins, quinolones, cephalosporins and glycopeptides, but
`diclofenac was the only NSAID with a significant ROR. As a category, vaccines
`
`PGR2020-00009
`Pharmacosmos A/S v. American Regent, Inc.
`Petitioner Ex. 1057 - Page 1
`
`

`

`548
`
`Leone et al.
`
`had a significantly lower ROR, thus indicating that anaphylaxis is reported
`proportionally less than other ADRs.
`Conclusions: Anaphylaxis is a severe ADR that may also occur with commonly
`used drugs. It represents 2.7% of all of the ADRs reported in an Italian spontane-
`ous reporting database.
`
`Background
`
`Although there is no universal definition of ana-
`phylaxis because of its multi-faceted nature, it can
`be described as an immediate hypersensitivity reac-
`tion with a potentially life-threatening outcome that
`can affect virtually any organ and is due to the union
`of an allergen with the IgE of basophils and mast
`cells. This interaction leads to the rapid release of
`preformed mediators, such as histamine, and other
`mediators, such as leukotrienes, which are responsi-
`ble for the clinical manifestations involving the pul-
`monary, circulatory, cutaneous, neurological and
`gastrointestinal systems.[1-3]
`Anaphylactoid reactions are clinically indistin-
`guishable from anaphylactic reactions and acute
`management is the same for both, but anaphylactoid
`reactions do not necessarily require previous expo-
`sure to an inciting substance.[4] They may be immu-
`nological and not IgE-mediated or non-immunologi-
`cal, in which case the mediators are directly released
`by basophils and mast cells.[5]
`The most common manifestations of anaphylaxis
`are erythema, pruritus, urticaria, angioedema, nau-
`sea, vomiting, diarrhoea, bronchospasm, laryngeal
`oedema, hypotension and cardiovascular collapse
`with shock.[6] There is a direct relationship between
`the time of onset of the symptoms after antigen
`administration and their severity: the more rapid the
`onset, the more severe the episode.[7]
`The incidence of anaphylaxis in the general pop-
`ulation varies and is often under-reported, partially
`because of the absence of a standard definition. As
`only a few epidemiological studies have been pub-
`lished, its exact incidence is difficult to define, but
`seems to be increasing.[8,9] A collaborative study
`conducted in Hungary, India, Spain and Sweden and
`based on definite and probable cases of anaphylaxis,
`estimated that the overall risk of severe anaphylaxis
`
`is 154 per million hospital admissions (0.015%).[10]
`In the US, it is estimated that 1 in every 2700
`inpatients (0.037%) experiences a drug-induced
`anaphylactic reaction[11] and a recent review of the
`current medical literature concerning anaphylaxis in
`the US has estimated that it affects between 1.21%
`and 15.04% of the general population.[5] Another
`general population study of anaphylaxis carried out
`in Olmsted County (MN/US) found an average an-
`nual incidence of 21 per 100 000 person years (95%
`CI 17, 25), with an occurrence rate of 30 per 100 000
`person years (95% CI 25, 35).[12]
`Anaphylactic events seem to be related to both
`age and sex as they are more likely to occur in adults
`and women.[10,13] An analysis of 4 years of English
`hospital admissions for anaphylaxis reported an in-
`creased incidence in females of child-bearing age,
`with a ratio of 1.38 (95% CI 1.27, 1.50), thus sug-
`gesting that endocrine factors may be important in
`the pathogenesis of allergic reactions.[14]
`The most common reported causes of anaphylax-
`is are medications, insect stings, food and latex,
`although any agent capable of stimulating mast cells
`or basophils is a potential cause.[9] When the cause is
`not known, idiopathic anaphylaxis is supposed. The
`incidence of anaphylactic or anaphylactoid reactions
`varies among classes of drugs and it is known that
`reactions to antibacterials, especially penicillins,
`have been observed the most frequently.[6,15] Aller-
`gic reactions to penicillins have been reported as
`occurring during 0.7–8% of treatment courses in
`different studies and anaphylactic reactions occur in
`0.004–0.015%.[16] The incidence of mortality is
`0.0015–0.002% or one death per 50 000–100 000
`treatment courses.[16]
`Other drugs that have been associated with ana-
`phylaxis include aspirin (acetylsalicylic acid) and
`other NSAIDs, anaesthetic medications, radiology
`
` 2005 Adis Data Information BV. All rights reserved.
`
`Drug Safety 2005; 28 (6)
`
`PGR2020-00009
`Pharmacosmos A/S v. American Regent, Inc.
`Petitioner Ex. 1057 - Page 2
`
`

`

`Drug-Induced Anaphylaxis: Data from Spontaneous Reporting
`
`549
`
`Methods
`
`The data were obtained from a database contain-
`ing all of the voluntarily submitted reports of ADRs
`from the Italian regions of Emilia Romagna, Lom-
`bardy and the Veneto. These regions had an estimat-
`ed population of approximately 18 000 000 inhabi-
`tants in January 2000 (about 32% of the Italian
`population) and are the main contributors to the
`Italian spontaneous surveillance system (accounting
`for approximately 54% of all Italian reports). We
`analysed the spontaneous reports collected between
`January 1990 and December 2003 in terms of the
`following information: reporter category, patient’s
`age and sex, reporter’s ADR diagnosis, characteris-
`tics of the underlying diseases, drug exposure (indi-
`cation, duration of treatment and dose), the time of
`onset of the event and its outcome. The reports were
`classified according to the WHO criteria for causali-
`ty assessment[23] and only those with a ‘certain’,
`‘probable’ or ‘possible’ causality assessment were
`included.
`The association between drugs and anaphylaxis
`was analysed using a case/non-case design accord-
`ing to the method described by van Puijenbroek et
`al.[24] The cases (defined with the general term of
`anaphylaxis) were the reports describing reactions
`already coded with the WHO preferred terms ‘ana-
`It is known that the incidence of an adverse drug
`phylactic shock’ or ‘anaphylactoid reaction’. The
`reaction (ADR) cannot be estimated on the basis of
`anaphylactoid reactions also included the anaphy-
`spontaneous reports, but the development of quanti-
`lactic reactions, since they share the same clinical
`tative methods for measuring signals does make it
`features and can not be distinguished on clinical
`possible to detect a disproportional number of a
`grounds. Moreover, a case by case analysis was
`reaction when specific knowledge of the composi- made on reports with a time of event onset that
`tion of a database is presented.[21,22]
`suggested an allergic reaction and involved at least
`two of the skin, respiratory, gastrointestinal, CNS or
`cardiovascular systems. Among these, reports with
`allergic symptoms were classified as anaphylactoid
`reactions, whereas those that included shock-like
`symptoms and decreased blood pressure were clas-
`sified as anaphylactic shock. The non-cases were all
`of the other ADR reports included in the database.
`All of the cases were analysed in detail by a
`specially constituted ad hoc panel of experts, includ-
`ing internists, pharmacologists and immunologists,
`
`contrast agents and vaccines; however, it is difficult
`to make a thorough evaluation of the incidence of
`anaphylaxis in relation to individual drugs because
`the reaction is rare and the available data are mainly
`based on case reports or small case series in which
`the denominator is unknown.
`
`One case-control study of the risk of anaphylaxis
`following drug exposure during hospitalisation
`showed an incidence in the range of 5–15 cases per
`100 000 patients for most analgesics and antibacteri-
`als, whereas dextran, parenteral penicillin, pentoxi-
`fylline and streptokinase were associated with an
`incidence of ≥30 cases per 100 000 patients.[17]
`Data concerning drug-related anaphylaxis that is
`obtained from spontaneous reporting systems in
`some countries have shown that dextran, radiology
`contrast agents, antibacterials and NSAIDs are the
`most frequently involved drugs.[18-20] For some
`drugs, this adverse reaction was unknown at the time
`of reporting, thus confirming that pharmacovigi-
`lance remains essential for signal generation.
`
`Various measures have been taken to decrease
`the risk of anaphylaxis on the basis of the results of
`these studies, e.g. the development of non-ionic
`radiology contrast agents, the marketing of dextran
`1 and the worldwide withdrawal of the analgesic
`glafenine.
`
`Objective
`
`The aim of this study, which was based on an
`Italian spontaneous reporting database, was to iden-
`tify the number of cases of anaphylaxis reported in
`association with different drug classes and compare
`the data with the other reports present in the
`database.
`
` 2005 Adis Data Information BV. All rights reserved.
`
`Drug Safety 2005; 28 (6)
`
`PGR2020-00009
`Pharmacosmos A/S v. American Regent, Inc.
`Petitioner Ex. 1057 - Page 3
`
`

`

`550
`
`Leone et al.
`
`whose main task was to check the diagnosis of
`anaphylaxis and the causality relationship.
`The frequency of the association between ana-
`phylaxis and the suspected drug in comparison with
`the frequency of anaphylaxis associated to all the
`other drugs was calculated using the ADR reporting
`odds ratio (ROR) as a measure of disproportional-
`ity.[24] The ROR is identical to the calculation of an
`odds ratio from a case-control study that compares
`each drug in turn to all other drugs. Odds ratios and
`their 95% CIs (adjusted for age, sex and patient
`typology) were calculated by means of logistic re-
`gression using SPSS statistical software. The χ2 test
`or Student’s t-test were used to compare cases and
`non-cases as appropriate. Differences were consid-
`ered significant with p-values of <0.05.
`
`Table I. Main patient characteristics
`
`Patient characteristics Cases
`(n = 744)
`
`Non-cases
`(n = 27 512)
`
`p-Value
`
`Sex [n (%)]
`423 (56.9)
`females
`321 (43.1)
`males
`Mean age (years) ± SD 49.0 ± 21.2
`Patient typology [n (%)]
`outpatients
`inpatients
`χ2 test.
`a
`b Student’s t-test.
`NS = not significant.
`
`304 (40.9)
`440 (59.1)
`
`16 540 (60.1) NSa
`10 972 (39.9)
`49.0 ± 24.4
`
`NSb
`
`17 920 (65.1)
`9592 (34.9)
`
`<0.001a
`
`Results
`
`Two hundred and sixty different drugs were sus-
`pected of causing anaphylaxis. Ninety-two percent
`of cases were attributed to only one suspected drug;
`a similar percentage was observed in non-cases
`(91%). Table II shows the differences between cases
`and non-cases in relation to the drug administration
`As of December 2003, the database contained
`route: anaphylactic shock was more frequently re-
`30 975 ADR reports, of which 2276 (7.3%) were
`ported after intravenous administration than either
`excluded because they were unclassifiable or the
`anaphylactoid reactions or other ADRs; however,
`causality assessment was ‘unlikely’. The primary more than one-third of the cases of anaphylactic
`selection identified a total of 1187 cases of anaphy-
`shock were caused by an oral drug.
`laxis, but the expert panel excluded 443 cases be-
`Table III shows the main six drug categories that
`cause the diagnosis of anaphylaxis was uncertain.
`were associated with an outcome of anaphylaxis,
`The analysis was, therefore, based on 744 cases
`with the corresponding number of other reported
`(including 307 cases of anaphylactic shock with 10
`ADRs and the ROR. Among the systemic an-
`deaths) and 27 512 non-cases. Among the included
`tibacterial agents, anaphylaxis was disproportional-
`reports, the causality assessment was certain in 5%,
`ly more often reported in relation to penicillins,
`probable in 59% and possible in 36%, with no
`quinolones, cephalosporins and glycopeptides with
`significant differences between the cases and non-
`a significant ROR (table IV). The individual an-
`cases.
`tibacterial agents (with a total of ≥30 reports) that
`Table I shows the main characteristics of the
`were suspected of causing anaphylaxis are listed in
`patients. There were no significant differences be-
`table V.
`tween the cases and non-cases in terms of age and
`As shown in table VI, diclofenac was the only
`sex. Patients were also divided, on the basis of the
`NSAID with a significant ROR. There was signifi-
`reporter categories, into outpatients (reports coming
`cant disproportionality between the cases and non-
`from general practitioners and emergency rooms)
`cases for all of the radiology contrast agents (table
`and inpatients (from other hospital wards). The per-
`VII). Propyphenazone (ROR 5.81; 95% CI 3.52,
`centage of anaphylaxis cases was higher among
`9.59) and dipyrone (ROR 2.39; 95% CI 1.22, 4.72)
`hospital patients than outpatients (see table I); this
`were the analgesic drugs that were mainly responsi-
`was particularly evident for anaphylactic shock
`ble for anaphylaxis. There were also reports of ana-
`(69.6% vs 30.4%). This distribution was significant-
`phylaxis that were related to all blood substitutes
`ly different from that of the other ADR reports,
`and perfusion solutions, but the only agents with a
`which mainly referred to outpatients.
`
` 2005 Adis Data Information BV. All rights reserved.
`
`Drug Safety 2005; 28 (6)
`
`PGR2020-00009
`Pharmacosmos A/S v. American Regent, Inc.
`Petitioner Ex. 1057 - Page 4
`
`

`

`Drug-Induced Anaphylaxis: Data from Spontaneous Reporting
`
`551
`
`Discussion
`
`number of reports suitable for analysis were polyge-
`line (ROR 16.62; 95% CI 9.90, 27.90) and human
`albumin (ROR 7.75; 95% CI 2.95, 20.37). As a
`category, vaccines had a significantly lower ROR,
`thus indicating that anaphylaxis is reported propor-
`tionally less than other ADRs.
`
`ous reporting, the association between a drug and a
`reaction may be artificially decreased if another
`drug-specific reaction is widely reported because
`this dilutes the association by increasing the pres-
`ence of the drug in non-case reports.[27]
`In this study, we applied the case/non-case
`method to an Italian spontaneous reporting database
`by comparing the ratio of anaphylaxis reports
`(cases) that were identified by clinical signs and
`This study originated from spontaneous reporting
`chronology with those referring to all other ADRs
`data concerning anaphylaxis and so the results
`(non-cases) and looked for associations with patient
`should be interpreted in this context. The main aim
`characteristics and specific drugs or drug classes.
`of spontaneous reporting systems is to detect previ-
`Given the characteristics of spontaneous reporting
`ously unknown adverse drug-related events or those
`and the variability in terminology used by doctors to
`occurring in a quantitatively or qualitatively differ-
`report an ADR, the cases included both anaphylactic
`ent manner from that expected. Quantitative meth-
`and anaphylactoid reactions because although their
`ods have recently been developed to supplement the
`physiopathology differs, their clinical manifesta-
`simple inspection of reports, all of which are based
`tions are often not clearly distinguishable.
`on an assessment of how much the observed report-
`In the period 1990–2003, the reports of anaphy-
`ing frequency of a given ADR deviates from that
`laxis accounted for 2.7% of all of the ADR reports in
`expected within a database.[25]
`the database. This percentage is about four times
`Analysis of databases should take into account
`higher than that found by van Puijenbroek et al.[24] in
`any possible bias that is related to spontaneous re-
`a Dutch study that shared many characteristics with
`porting. One well known problem is under report-
`our own in terms of the total number of reports in the
`ing, but when all drug reactions are similarly under
`spontaneous reporting database, the time of observa-
`reported it is assumed that this would not lead to a
`tion and the case selection criteria. We have no data
`systematic bias in the analysis of large databases.[26]
`concerning under reporting in the two countries, but
`However, uneven under reporting may be a signifi-
`we have no reason to think that it would be specifi-
`cant problem. High reporting of a reaction can be
`cally different for anaphylaxis.
`related to a temporary special attention by doctors
`A number of studies have shown marked varia-
`(e.g. new reactions to a new drug, media claims,
`bility among European countries in the consumption
`specific guidelines, etc.), whereas a well known
`of some classes of drugs, such as antibacterials,[28]
`reaction may not be reported. Moreover, when using
`disproportionality measures in analysis of spontane- NSAIDs[29] and lipid-lowering agents.[30] The an-
`
`Table II. Administration route of suspected drugs
`
`Administration route
`
`anaphylactic shock
`125 (36.2)
`159 (46.1)
`44 (12.7)
`14 (4.1)
`3 (0.9)
`345 (100)
`
`Drugs suspected in cases [n (%)]
`anaphylactoid reactionb
`221 (46.4)
`Oral
`144 (30.3)
`Intravenous
`83 (17.4)
`Other parenteral
`25 (5.3)
`Other routes
`3 (0.6)
`Undefined
`476 (100)
`Total
`χ2 test (cases vs non-cases).
`a
`b WHO Adverse Reaction Terminology (WHO-ART) term that also includes anaphylactic reaction.
`NA = not applied; NS = not significant.
`
`Drugs suspected in
`non-cases [n (%)]
`16 154 (52.9)
`1911 (6.3)
`4490 (14.7)
`1019 (3.3)
`6961 (22.8)
`30 535 (100)
`
`p-Valuea
`
`<0.001
`<0.001
`NS
`NS
`NA
`
` 2005 Adis Data Information BV. All rights reserved.
`
`Drug Safety 2005; 28 (6)
`
`PGR2020-00009
`Pharmacosmos A/S v. American Regent, Inc.
`Petitioner Ex. 1057 - Page 5
`
`

`

`552
`
`Leone et al.
`
`Table III. Drug categories suspected of causing anaphylaxis (cases) and other reported adverse drug reactions (non-cases)
`
`Drug categoriesa
`Antibacterials for systemic use
`NSAIDs and aspirin (acetylsalicylic acid)
`Radiology contrast agents
`Vaccines
`Analgesic drugs
`Blood substitutes and perfusion solutions
`a Categories with ≥30 reports of anaphylaxis.
`b The percentages refer to the total number of cases or non-cases.
`
`Cases [n (%)]b
`255 (34.3)
`104 (14.0)
`100 (13.4)
`50 (6.7)
`40 (5.4)
`33 (4.4)
`
`Non-cases [n (%)]b
`5833 (21.2)
`3131 (11.4)
`576 ( 2.1)
`3315 (12.0)
`806 ( 2.9)
`118 ( 0.4)
`
`Reporting odds ratio (95% CI)
`1.94 (1.66, 2.26)
`1.27 (1.03, 1.56)
`7.26 (5.79, 9.11)
`0.53 (0.39, 0.70)
`1.88 (1.36, 2.61)
`10.78 (7.27, 15.96)
`
`Table IV. Number of cases and non-cases attributed to the different
`classes of antibacterials for systemic use
`
`the type of reaction, such as differences in pharma-
`tibacterial sales data for 1997 show that sales were
`three times higher in Italy than in The Nether-
`cokinetics, the influence of circulating hormones on
`lands[31] and a possible overuse, particularly of in-
`drug metabolism, the greater consumption of drugs
`jectable antibacterials, may partially explain the
`by women and their higher reporting rate to doc-
`higher number of anaphylaxis reports. In the WHO tors.[35]
`database, which collates spontaneous reports from
`The drug categories that were most frequently
`more than 70 countries worldwide, 27% of the re-
`reported as the cause of anaphylaxis were in line
`ports of anaphylactic shock that are associated with with the medical literature (antibacterials, NSAIDs,
`injectable antibacterials come from Italy (Leone R,
`etc.). Blood substitutes (polygeline and human albu-
`personal research: WHO database). This is an ex- min) and radiology contrast agents were the catego-
`tremely high figure as ADR reports from Italy ac-
`ries with the highest ROR, thus indicating that ana-
`count for no more than 1% of the WHO database.
`phylaxis is probably their most frequent serious
`The distribution of drugs among cases and non-
`adverse event. Polygeline, a polymerised gelatin
`cases by an administration route clearly shows the
`used as a plasma volume expander, is known to
`association between anaphylactic shock and intrave-
`cause anaphylaxis with an incidence ranging from
`nous drug administration, thus confirming that “ana-
`0.78% to 26%.[36,37] Histamine release is also a well
`phylactic reactions are usually more dramatic when
`known effect of human albumin,[38] although to a
`the drug is given by injection than when it is given
`lesser extent than polygeline, but the incidence of
`orally”.[32] Intravenous drug administration is almost
`allergic reactions due to albumin seems to be
`exclusively used in hospital settings, which may
`low.[39,40]
`explain the greater frequency of anaphylaxis in inpa-
`It is well known that low-osmolality, non-ionic
`tients in comparison with other ADRs. On the con-
`radiology contrast agents cause fewer life-threaten-
`trary, ADRs other than anaphylaxis are more fre-
`quently reported in outpatients and are associated
`with orally administered drugs (obviously because
`this is the most common route of drug administra-
`tion).
`We did not find any differences in the distribu-
`tion of age and sex among the cases and non-cases.
`Females were more likely to develop anaphylaxis as
`well as the other ADRs. A common finding of
`several studies that were based on spontaneous re-
`porting and other surveillance systems was that wo-
`men experience more ADRs than men.[9,33,34] This is
`probably related to multiple factors that depend on
`
`Drug categories
`
`Cases
`
`Penicillins
`Quinolones
`Cephalosporins
`Macrolides
`Glycopeptides
`Aminoglycosides
`Other β-lactams
`Tetracyclines
`Other antibacterials
`
`88
`72
`50
`14
`7
`5
`3
`1
`3
`
`Non-cases Reporting odds
`ratio (95% CI)
`1.64 (1.30, 2.05)
`2.17 (1.69, 2.79)
`2.36 (1.76, 3.17)
`0.58 (0.34, 0.99)
`2.46 (1.14, 5.30)
`2.16 (0.87, 5.33)
`1.85 (0.58, 5.92)
`0.41 (0.06, 2.91)
`0.68 (0.22, 2.12)
`
`2085
`1292
`815
`883
`106
`86
`60
`91
`164
`
` 2005 Adis Data Information BV. All rights reserved.
`
`Drug Safety 2005; 28 (6)
`
`PGR2020-00009
`Pharmacosmos A/S v. American Regent, Inc.
`Petitioner Ex. 1057 - Page 6
`
`

`

`Drug-Induced Anaphylaxis: Data from Spontaneous Reporting
`
`553
`
`Table V. Individual antibacterial agents suspected of causing ana-
`phylaxis
`
`Druga
`Reporting odds ratio (95% CI)
`11.97 (6.53, 21.94)
`Cinoxacin
`4.74 (2.15, 10.45)
`Piperacillin
`3.75 (2.43, 5.79)
`Ceftriaxone
`3.43 (1.48, 7.95)
`Teicoplanin
`3.01 (1.98, 4.60)
`Levofloxacin
`a Agents with ≥30 reports of anaphylaxis.
`
`Among the glycopeptide antibacterials, we found
`a higher ROR for teicoplanin even though it is
`generally reported to be better tolerated than vanco-
`mycin (ROR 0.90; 95% CI 0.12, 6.56).[60] There are
`no published case reports of anaphylaxis and the
`first case of hypersensitivity was published in May
`2004.[61] However, the WHO database contains 29
`reports of anaphylactic shock and 25 of anaphylac-
`toid reactions that are associated with teicoplanin,
`including those that come from Italy (Leone R,
`personal research: WHO database).
`
`ing adverse reactions than higher-osmolality mol-
`ecules.[41-43] As a consequence, they are widely used
`in Italy despite their higher cost. Some cases of
`Our analysis indicates that diclofenac is the
`anaphylaxis related to non-ionic radiology contrast NSAID that is most frequently associated with ana-
`agents have been reported,[44-48] although the mecha-
`phylaxis and it is worth noting that this result is
`nisms involved are still a matter of debate.[49-52] We
`similar to that found by van Puijenbroek et al.[24] in
`found a high ROR for all of these agents, with no
`The Netherlands. There are some published cases of
`particular differences between the individual com-
`diclofenac-induced anaphylaxis after oral adminis-
`pounds. This finding suggests that the risk of ana-
`tration and patch testing.[62-66] In a case-cohort study
`phylaxis with non-ionic radiology contrast agents
`of 934 hospital admissions for anaphylaxis in The
`should not be disregarded and emphasises the need Netherlands, diclofenac was one of the most com-
`to train radiologists to recognise, prevent and treat mon drug-related causes; the authors estimated an
`the anaphylaxis induced by contrast media.[53]
`incidence of diclofenac-induced anaphylaxis of 1 in
`10 000 to 1 in 20 000 prescriptions.[67] It has been
`suggested that diclofenac is associated with selec-
`tive hypersensitivity without cross-reactivity with
`other NSAIDs[68] and in an experimental setting, the
`drug induces a direct T cell-dependent popliteal
`lymph node reaction and has intrinsic adjuvant ac-
`tivity that selectively induces the interleukin-4 me-
`diated production of IgG1 and IgE.[69]
`
`In this study, antibacterials were responsible for
`the highest number of reports of anaphylaxis and
`three of the four classes showing significant
`disproportionality
`(penicillins, quinolones and
`cephalosporins) are among the most widely pre-
`scribed antibacterials in Italy. In particular, ceftriax-
`one (a third-generation injectable cephalosporin)
`and the fluoroquinolone levofloxacin were the third
`and fourth antibacterial, respectively, used in Italy in
`2003, ranked by cost.[54] It is estimated that anaphy-
`laxis occurs in 0.46–1.2 per 100 000 patients treated
`with fluoroquinolones and several case reports and
`case series have been associated with individual
`quinolones.[55,56] Our results indicate a higher dis-
`proportionality of anaphylaxis for levofloxacin and
`cinoxacin.
`
`Cinoxacin product information states that the
`most frequently reported adverse events in the
`postmarketing surveillance of cinoxacin have been
`rash and anaphylactic reactions.[57] Several case re-
`ports have also been published on anaphylactic reac-
`tions associated with the use of cinoxacin.[58,59]
`
`Table VI. Number of cases and non-cases attributed to NSAIDs
`and aspirin (acetylsalicylic acid)
`
`Druga
`
`Cases
`
`Non-cases
`
`353
`30
`Diclofenac
`339
`15
`Ketoprofen
`566
`15
`Aspirin
`495
`8
`Nimesulide
`141
`7
`Naproxen
`150
`6
`Ketorolac
`128
`4
`Ibuprofen
`234
`3
`Rofecoxib
`254
`2
`Celecoxib
`167
`1
`Piroxicam
`a Only the drugs with ≥100 reports.
`
`Reporting odds ratio
`(95% CI)
`3.23 (2.21, 4.73)
`1.65 (0.98, 2.78)
`0.98 (0.58, 1.64)
`0.59 (0.29, 1.20)
`1.84 (0.86, 3.95)
`1.48 (0.65, 3.36)
`1.16 (0.43, 3.14)
`0.47 (0.15, 1.48)
`0.29 (0.07, 1.17)
`0.22 (0.03, 1.58)
`
` 2005 Adis Data Information BV. All rights reserved.
`
`Drug Safety 2005; 28 (6)
`
`PGR2020-00009
`Pharmacosmos A/S v. American Regent, Inc.
`Petitioner Ex. 1057 - Page 7
`
`

`

`554
`
`Leone et al.
`
`Table VII. Number of cases and non-cases attributed to radiology
`contrast agents
`
`Drugsa
`
`Cases
`
`Non-cases
`
`157
`29
`Iopromide
`150
`24
`Iomeprol
`99
`15
`Iopamidol
`46
`5
`Iodixanol
`30
`4
`Ioversol
`a Only the agents with ≥30 reports.
`
`Reporting odds ratio
`(95% CI)
`7.07 (4.72, 10.57)
`6.08 (3.93, 9.41)
`5.70 (3.29, 9.86)
`4.04 (1.60, 10.20)
`4.95 (1.74, 14.09)
`
`agent capable of activating mast cells or basophils
`can potentially cause anaphylactic or anaphylactoid
`reactions. Levofloxacin, cinoxacin, ceftriaxone, di-
`clofenac, propyphenazone and dipyrone are widely
`used in Italy, but showed a significant dispropor-
`tionality for anaphylaxis in comparison with other
`ADRs.
`The most important point to bear in mind regard-
`ing anaphylaxis is that it is preferable to prevent it
`occurring rather than having to treat a reaction.
`Unfortunately, these ADRs are characterised by
`their unpredictable nature and a predisposition to
`anaphylaxis is not easily recognisable or testable. In
`clinical practice prevention methods are to collect
`background information about the history of previ-
`ous anaphylactic episodes and replace the offending
`agent with another that is not cross-reactive. Other
`useful measures could be the proper recognition of
`anaphylaxis when it occurs and employment of a
`20–30-minute observation period after the adminis-
`tration of parenteral medications.
`At the same time, much work remains to be done
`to describe these reactions accurately and we believe
`that morbidity and mortality due to anaphylaxis can
`only be minimised by improving our knowledge of
`its epidemiology and risk factors, and ensuring that
`drugs are used rationally and safely.
`
`The results concerning analgesic-related ana-
`phylaxis suggest caution when using propy-
`phenazone or dipyrone. The estimated incidence of
`propyphenazone-induced anaphylaxis
`is 0.2%,
`whereas the incidence of dipyrone-related anaphy-
`laxis is 0.02%.[70] In Italy, propyphenazone is also
`available over the counter in association with a
`peripheral cough suppressant and the suppository
`formulation is widely used in children. The adminis-
`tration of dipyrone is controversial because of life-
`threatening adverse events, such as agranulocytosis
`and aplastic anaemia, and its use is banned in vari-
`ous countries. Cases of anaphylaxis following
`dipyrone therapy have mainly been reported in
`Spain, Germany and Russia.[71-73]
`Over the 14 years of our pharmacovigilance sys-
`tem, 3365 adverse reactions related to vaccines have
`been reported, including 50 cases of anaphylaxis.
`Unlike in the case of other drugs, we found that the
`ROR for vaccines was <1, which may reflect a bias
`related to the tendency of Italian doctors to report
`We are very grateful to the Pharmaceutical Departments
`of Emilia Romagna, Lombardy and the Veneto and their local
`any (serious and non-serious) adverse vaccine reac-
`Health Districts for collecting the adverse reaction forms.
`tion (as also required by a recent law), thus leading
`We would also like to thank the Uppsala Monitoring
`to a high proportion of non-cases. However, a low Centre for allowing us to consult the WHO database. No
`rate of anaphylaxis after the vaccination of children
`sources of funding were used to assist in the preparation of
`and adolescents has also been observed in a study
`this study. The authors have no conflicts of interest that are
`directly relevant to the content of this study.
`population, with an estimated risk of 0.65 cases per
`million vaccine doses (95% CI 0.21, 1.53).[74]
`
`Acknowledgements
`
`Conclusion
`
`This analysis of 744 reports of anaphylaxis in an
`Italian spontaneous reporting database using case/
`non-case methodology confirms the drug categories
`with the highest risk and shows that the possibility
`of anaphylaxis should not be disregarded, since any
`
`References
`1. Bochner BS, Lichtenstein LM. Anaphylaxis. N Engl J Med
`1991; 324: 1785-90
`2. Ewan PW. Anaphylaxis. BMJ 1998; 316: 1442-5
`3. O’Hollaren MT. Anaphylaxis: new clues to clinical patterns and
`optimum treatment. Medscape Allergy & Clinical Immunolo-
`gy 2002; 2 (2) [online]. Available from URL: http://www.med-
`scape.com/viewarticle/439721 [Accessed 2005 Apr 21]
`4. Anderson JA. Allergic reactions to drugs and biological agents.
`