An Evaluation of “Drug Ineffective” Postmarketing Reports in Drug Safety Surveillance

An Evaluation of “Drug Ineffective” Postmarketing Reports in Drug Safety Surveillance Drugs - Real World Outcomes (2018) 5:91–99 https://doi.org/10.1007/s40801-018-0131-3 OR IGINAL RESEARCH ARTIC L E An Evaluation of ‘‘Drug Ineffective’’ Postmarketing Reports in Drug Safety Surveillance 1,2 1 1 1 • • • • Takashi Misu Cindy M. Kortepeter Monica A. Mun ˜ oz Eileen Wu Gerald J. Dal Pan Published online: 23 February 2018 The Author(s) 2018. This article is an open access publication Abstract reports provided a batch or lot number (39.5 vs. 17.2%) and Introduction The most commonly reported adverse event, were coded with additional PTs beyond ‘‘drug ineffective’’ based on frequency of Medical Dictionary for Regulatory (83.7 vs. 59.2%), the most frequent of which were ‘‘product Activities (MedDRA) preferred terms (PTs), in the US quality issue’’ (23.3%) and ‘‘product substitution issue’’ FDA Adverse Event Reporting System (FAERS) database (18.6%). is ‘‘drug ineffective’’ (DI). This study aimed to describe the Conclusions DI was the most frequently reported adverse DI reports and provide data to support recommendations on event in the FAERS database; however, the yield from how to best evaluate these reports. these reports in terms of usefulness from a pharmacovigi- Methods We characterized all FAERS reports coded with lance perspective was low. Efficient strategies are needed the MedDRA PT ‘‘drug ineffective’’ received between 1 to identify which DI reports are more likely to contain September 2012 and 31 August 2016 using all other useful information. FAERS reports as a comparator. Additionally, we con- ducted a manual evaluation to identify informative data elements in the report narratives. Key Points Results During the study period, 247,513 (6.4% of all FAERS reports) DI reports were entered in FAERS. The most frequently reported adverse event in the Compared with non-DI reports, DI reports were more US FDA Adverse Event Reporting System (FAERS) likely to be reported by consumers (69.8 vs. 48.1%) and database was ‘‘drug ineffective’’ (DI). less likely to report a serious outcome (26.2 vs. 56.3%). Most DI reports in FAERS were reported by Most DI reports (88%) were from the USA. Manual eval- consumers and were non-serious. uation of 552 sample US reports identified 43 reports (7.8%) deemed ‘‘useful’’; a higher proportion of ‘‘useful’’ A minority of DI reports were deemed ‘‘useful’’. Many of these provided a batch or lot number, and the majority were coded with additional preferred Disclaimer The views expressed are those of the authors and do not terms beyond DI. necessarily represent the position of, nor imply endorsement from, the US FDA, the US Government, or the Japanese Pharmaceuticals and Medical Devices Agency. & Cindy M. Kortepeter Cindy.Kortepeter@fda.hhs.gov 1 Introduction Office of Surveillance and Epidemiology, Center for Drug Evaluation and Research, U.S. Food and Drug Adverse event (AE) reporting has been a central part of the Administration, Silver Spring, MD, USA 2 US FDA’s postmarketing drug safety surveillance for Office of Safety II, Pharmaceuticals and Medical Devices nearly 50 years [1]. While population-based databases Agency, Tokyo, Japan 92 T. Misu et al. have become additional sources of postmarketing safety concurrently reported PTs. The primary suspect product’s information, especially in the past two decades, individual application type was determined from a structured field case safety reports (ICSRs) of AEs remain an important provided by the submitting manufacturer. The most fre- source of postmarketing safety information [2, 3]. Each quently reported suspect products, by the product’s active data source has well-characterized strengths and limita- ingredient, were compared between the DI and non-DI tions. A major limitation in postmarketing ICSRs is the report groups. A product’s active ingredient was defined as frequent lack of sufficient detail to allow an independent the ingredient that has a pharmacological effect of a reviewer to make a reasonable assessment of the potential medicinal product reported by drug manufacturers. Addi- relationship between a drug and a reported AE [4]. For tionally, we identified the active ingredients with the example, lack of accurate product identification and rele- highest proportions of DI reports for products with at least vant clinical details limit the inferences that can be made 1000 total reports in the study period. from ICSRs, especially when there are alternative potential Following the high-level characterization of DI and non-DI explanations for the reported AEs. reports, we conducted a manual evaluation to assess the The most commonly reported AE, based on frequency of availability of informative data elements in the report narra- Medical Dictionary for Regulatory Activities (MedDRA) tives and to classify reports by their potential utility in the preferred terms (PTs), in the US FDA Adverse Event assessment of the relationship between a drug and the AE (of Reporting System (FAERS) database is ‘‘drug ineffective’’ drug ineffectiveness). This assessment was made to determine (DI) [5]. These DI reports in FAERS have not been whether the report was ‘‘useful’’ from a pharmacovigilance assessed systematically for quality and inferential value perspective. We defined ‘‘useful’’ as reports containing the from a pharmacovigilance perspective. The objective of necessary information that would prompt a reviewer to con- this study is to describe the DI reports in FAERS and sider further action, which in most cases would be obtaining provide data to support recommendations on how to best additional information. For this study, a ‘‘useful’’ report evaluate these reports. contains criteria 1 and 2 and at least one of the other four criteria, as listed in Table 1. An assessment of causality was not conducted in the determination of a ‘‘useful’’ report. All 2 Methods FAERS reports were stratified by initial FDA received dates (September 2012–August 2013, September 2013–August FAERS is an electronic database that currently contains 2014, September 2014–August 2015, September 2015–Au- over 14 million ICSRs describing AEs or medication gust 2016). Random sampling was conducted from DI errors. Approximately 95% of FAERS reports are submit- reports, reflecting the proportions of reports in the periods. ted to the FDA by drug manufacturers, whereas 5% (called We restricted our manual evaluation to US reports given the ‘‘direct reports’’) are submitted directly to the FDA [5]. limited reporting requirements for reports that manufacturers Any member of the public (e.g., consumers, healthcare receive from outside the USA. A sample size of 552 was professionals [HCPs]) can report an AE to the FDA or the determined for manual evaluation assuming a prevalence rate manufacturer. When the manufacturer receives a report, of DI reports with potential utility of 10% and a precision of they must in turn report the AE to the FDA in accordance 2.5%. The prevalence of usefulness was estimated by piloting with regulatory requirements. Reports in FAERS may a review of 120 reports. contain narrative free text describing an AE, a list of Table 1 Criteria used to classify reports as ‘‘useful’’ for manual products suspected in that event, and information identi- evaluation fying the reporter. In addition to the narrative description Criterion Description of the AE, reporters may supply additional information, such as past medical history, laboratory data, the names of 1 The suspect product of drug ineffective was clearly suspect drug manufacturers, or the event outcome. identifiable We searched FAERS for all reports received by the FDA 2 An informative narrative to support the reported drug between 1 September 2012 and 31 August 2016. The ineffectiveness AND one or more of the following four criteria: retrieved reports were stratified by those coded with and 3 MedDRA preferred term(s) in addition to ‘‘drug without the MedDRA PT ‘‘drug ineffective’’ for compari- ineffective’’ was reported son. We summarized the following report characteristics: 4 Suspect product’s batch or lot number was reported report type (i.e., manufacturer or direct), patient and 5 A beneficial response prior to the administration of the reporter attributes, reporter country, reported outcomes, suspect product was reported primary suspect product’s application type (i.e., new drug 6 Medication switching was reported application [NDA], abbreviated new drug application MedDRA Medical Dictionary for Regulatory Activities [ANDA], biologics license application [BLA]), and Drug Ineffective Reports in Drug Safety Surveillance 93 Data collected during the manual evaluation was aimed Compared with reports without DI, DI was more likely to at determining whether information relevant to the evalu- be reported by consumers (69.8 vs. 48.1%) and less likely ation of drug ineffectiveness was documented in the ICSR to have a serious outcome reported (26.2 vs. 56.3%). Most with respect to the suspected ineffective product. These DI reports (88%) were from the USA. While age distri- included (1) classification of the product as brand (inno- butions were relatively similar between the DI and non-DI vator version) or generic; (2) description of medication report groups (median age 57 years [interquartile range switching; (3) action taken with the product suspected to be {IQR} 43–67] and 58 years [IQR 43–69], respectively), DI ineffective; (4) presence of the suspected product’s batch reports were more often missing the patient age (49.5 vs. or lot number; and (5) concurrently reported PTs. In this 38.7%). In total, 62% of DI reports included additional evaluation, a product was considered ‘‘brand’’ if an inno- coded PTs. The most frequently co-reported PTs in the DI vator’s product name was used or if an active ingredient reports were ‘‘product quality issue’’ (4.2%), ‘‘pain’’ was described as brand in the narrative. A product was (3.8%), and ‘‘fatigue’’ (3.2%). Of the non-DI reports, the classified as ‘‘generic’’ if a non-innovator product was most frequently reported PTs were ‘‘death’’ (4.5%), specified or if an active ingredient was described as generic ‘‘nausea’’ (4.1%), and ‘‘fatigue’’ (3.9%). in the narrative. If a manufacturer name was provided in Table 3 shows the top 20 suspect product’s active the description, the manufacturer (and the suspect product) ingredients with the highest number of reports for the DI was identified as brand or generic using the FDA’s list of and non-DI groups. Adalimumab and etanercept were the Approved Drug Products with Therapeutic Equivalence top two products during the study period in both the DI and Evaluations (commonly known as the Orange Book)[6]. the non-DI groups. Relative to the non-DI reports, many of Reports describing the use of both brand and generic sus- the most frequently reported suspect product’s active pect products were classified as ‘‘multiple’’. Medication ingredients in the DI reports were those used for symp- switching was defined as a switch from one medicinal tomatic management (e.g., pain: naproxen, pregabalin, product to another with the same active ingredient, the gabapentin, buprenorphine, ibuprofen; allergies/asthma: same dosage form, and the same route of administration. loratadine, fexofenadine, fluticasone/salmeterol, albuterol) The switching patterns included (1) a patient taking drug A or where an effect is expected shortly after exposure (e.g., and drug A worked, the patient switched to drug B then DI onabotulinumtoxinA, sildenafil, insulin lispro). Suspect occurred; and (2) a patient experienced DI with drug A, products’ active ingredients with the highest number of DI then switched to drug B and drug B worked. A switch may reports in FAERS were adalimumab (8.0%), etanercept include brand to generic, generic to brand, or generic to (5.7%), naproxen (2.4%), pregabalin (2.3%), and onabo- generic switching. Reports describing ineffectiveness with tulinumtoxinA (2.2%). In all, 100 active ingredients a different batch or lot of the same product were excluded accounted for 80% of all DI reports during the study per- from the medication switching definition. PTs reported in iod. The proportion of DI reports within each active addition to DI were further classified into those describing ingredient was not proportional across suspect products. product quality issues or any AE other than a product For example, DI reports consisted of 41.2% of all reports quality issue. Reports were determined to contain addi- received with onabotulinumtoxinA, but only 6.8% of tional PTs associated with product quality issues if addi- etanercept reports given the large total number of etaner- tional PTs included at least one within the MedDRA high- cept reports. The products with the highest proportions of level group term (HLGT) ‘‘product quality, supply, distri- DI reports are provided in Table 4. bution, manufacturing and quality systems issues’’. All The results of the manual evaluation of 552 reports are included reports were evaluated independently by two presented in Table 5. We could identify the product related reviewers. Differences in evaluation between the two to the DI in the narrative field in 95.3% of the sampled reviewers were discussed as a group for consensus. The reports, of which 75.2% described a brand product. The cases determined to be ‘‘useful’’ were further compared most frequently identified products in the sampled reports with the remainder of the sample. We summarized the included adalimumab (8.5%), etanercept (5.4%), and resulting data using descriptive statistics. naproxen (3.4%). DI associated with medication switching was reported in 6.2%. Ineffectiveness when using a dif- ferent batch or lot of the same product was described in 3 Results 1.3% of DI reports. The suspect product was continued in 14.1% of reports. The suspect product’s batch or lot During the study period, 3.8 million reports were entered number was reported in 17.2%. Of the 552 reports, 59.2% into FAERS, of which 247,513 (6.4%) reports were coded were coded with additional PTs, the most frequent of which with the PT ‘‘drug ineffective’’. The overall characteristics were ‘‘fatigue’’ (4.9%), ‘‘product quality issue’’ (4.5%), of DI and non-DI reports are presented in Table 2. and ‘‘off-label use’’ (3.3%). 94 T. Misu et al. Table 2 Characteristics of Characteristic DI reports (N = 247,513) Non-DI reports (N = 3,625,330) ‘‘drug ineffective’’ and non- drug ineffective reports during N % N % the study period Report type Manufacturer 241,291 97.5 3,482,423 96.1 Direct 6222 2.5 142,907 3.9 Reporter type Consumer 172,834 69.8 1,744,954 48.1 Healthcare provider 69,770 28.2 1,692,858 46.7 Other 444 0.2 115,558 3.2 Missing 4465 1.8 71,960 2 Reporter country USA 217,966 88 2,673,274 73.7 Non-USA 29,547 12 952,056 26.3 Patient age (years) 0–17 6007 2.4 110,215 3 18–64 80,313 32.4 1,341,823 37 C 65 38,612 15.6 771,500 21.3 Missing 122,581 49.5 1,401,792 38.7 Patient gender Female 139,671 56.4 2,037,500 56.2 Male 82,326 33.3 1,261,002 34.8 Unknown/null 25,516 10.3 326,828 9 All outcomes Hospitalization 20,380 8.2 829,646 22.9 Death 4842 2 365,601 10.1 Disability 2832 1.1 67,991 1.9 Life threatening 2150 0.9 85,168 2.3 Required intervention 179 \0.1 10,919 0.3 Congenital anomaly 13 \0.1 14,408 0.4 Other 49,853 20.1 1,184,076 32.7 No serious outcome was reported 182,628 73.8 1,583,848 43.7 Primary suspect product’s application type NDA 144,168 58.3 1,973,700 54.4 BLA 48,946 19.8 822,300 22.7 ANDA 19,704 8 271,748 7.5 Missing 34,695 14 557,582 15.4 Additional PTs other than DI Reported 153,555 62 Not reported 93,958 38 ANDA abbreviated new drug application, BLA biologics license application, DI drug ineffective, FAERS FDA Adverse Event Reporting System, NDA new drug application, PT preferred term We determined that 43 of the 552 reports (7.8%) met our were generic in 51.2% (vs. 7.6% overall sample DI definition of ‘‘useful’’. Among the 43 ‘‘useful’’ reports reports). The most frequently reported products were fen- meeting criteria 1 and 2, a total of 20 reports (46.5%) met tanyl (9.3%), alprazolam (7.0%), and adalimumab (4.7%). an additional two or more criteria, and 12 reports (27.9%) Medication switching was reported in 44.2% (vs. 6.2% met an additional three or more criteria. Characteristics overall sample DI reports). The suspect product was dis- (including the criteria used to classify reports as ‘‘useful’’) continued in 37.2% (vs. 29.7% overall sample DI reports). of the 43 reports determined to be ‘‘useful’’ are in Table 6. A higher proportion of ‘‘useful’’ reports provided a batch or The suspect products identified from the narrative field lot number (39.5 vs. 17.2% overall sample DI reports). The Drug Ineffective Reports in Drug Safety Surveillance 95 Table 3 Most frequently reported suspect product’s active ingredients during the study period (top 20) Non-DI reports (N = 3,625,330) DI reports (N = 247,513) Product active ingredient N % Product active ingredient N % of DI reports % of all reports Adalimumab 200,482 5.5 Adalimumab 19,848 8.0 9.0 Etanercept 193,799 5.3 Etanercept 14,187 5.7 6.8 Calcium chloride, dextrose, magnesium 87,998 2.4 Naproxen sodium 5975 2.4 30.9 chloride, sodium chloride, sodium lactate Lenalidomide 77,064 2.1 Pregabalin 5600 2.3 13.0 Natalizumab 58,484 1.6 OnabotulinumtoxinA 5503 2.2 41.2 Rivaroxaban 57,412 1.6 Loratadine 4397 1.8 32.5 Rosiglitazone maleate 54,763 1.5 Sildenafil citrate 3985 1.6 26.0 Teriparatide 53,934 1.5 Gabapentin 3693 1.5 18.2 Interferon beta-1a 52,166 1.4 Fexofenadine hydrochloride 3462 1.4 36.8 Dimethyl fumarate 51,953 1.4 Dalfampridine 3429 1.4 14.1 Ribavirin 51,403 1.4 Buprenorphine 3419 1.4 13.2 Ambrisentan 49,966 1.4 Ibuprofen 3222 1.3 13.8 Levonorgestrel 48,390 1.3 Infliximab 3187 1.3 8.4 Denosumab 38,204 1.1 Polyethylene glycol 3350 2614 1.1 18.4 Pregabalin 37,473 1.0 Fluticasone propionate, 2541 1.0 12.5 salmeterol xinafoate Infliximab 34,602 1.0 Abatacept 2531 1.0 15.8 Peginterferon alfa-2a 33,640 0.9 Apremilast 2454 1.0 10.9 Insulin lispro 31,554 0.9 Levonorgestrel 2352 1.0 4.6 Aspirin 30,110 0.8 Albuterol sulfate 2351 0.9 16.9 Risperidone 27,266 0.8 Insulin lispro 2349 0.9 6.9 DI drug ineffective majority (83.7%) of the reports were coded with additional addition, responsible parties are not required to submit non- PTs beyond ‘‘drug ineffective’’ (vs. 59% overall sample DI serious AE reports from foreign marketing experience [7]. reports); ‘‘product quality issue’’ (23.3%), ‘‘product sub- The suspect products with the highest number of DI stitution issue’’ (18.6%), and ‘‘feeling abnormal’’ (11.6%) reports during the study period are used primarily for the were the three most frequently reported PTs. Of the 43 management of symptomatic conditions (e.g., adalimumab, ‘‘useful’’ reports, 44.2% included PTs related to product etanercept, naproxen, loratadine), suggesting that con- quality issues within the MedDRA HLGT ‘‘product qual- sumers have self-awareness of worsening or no improve- ity, supply, distribution, manufacturing, and quality sys- ment of their own subjective experiences. Similarly, the tems issues’’. suspect product with the highest proportion of DI reports within each active ingredient (docosanol) is used primarily for symptom management. Poitras et al. [8] conducted 4 Discussion similar research using the Canadian Vigilance Database and also identified high proportions of DI reports (ranging DI is the most frequently reported AE in the FAERS from 14.6 to 20% annually using the standardized Med- database. Our analysis demonstrated that, in contrast to all DRA query ‘‘lack of efficacy/effectiveness’’). The most other reports in the FAERS database, the majority of DI frequently reported drug classes in lack-of-efficacy reports cases did not report a serious outcome and were more were tumor necrosis factor-a inhibitors and proton pump likely to be reported by consumers. These reports were inhibitors [8]. primarily from the USA. The US Code of Federal Regu- Our manual evaluation revealed that many of the reports lations requires responsible parties (i.e., applicants, manu- lacked the clinical details needed to distinguish the repor- facturers, packers, or distributors) to submit adverse ted DI from disease progression. Interpretation of DI experiences associated with drugs or therapeutic biologic reports is complicated by the variable efficacy of approved products to the FDA, including experiences associated with products. Evidentiary standards for drug approval are that a failure to produce an expected pharmacologic action. In ‘‘substantial evidence that the drug will have the effect it 96 T. Misu et al. Table 4 Suspect product’s active ingredients with the highest proportion of ‘‘drug ineffective’’ reports in FAERS during the study period (top 25) Product’s active ingredient Total FAERS reports DI reports % DI Docosanol 4282 2124 49.6 OnabotulinumtoxinA 13,362 5503 41.2 Fexofenadine hydrochloride 9400 3462 36.8 Bupivacaine hydrochloride 1085 387 35.7 Diphenhydramine citrate/ibuprofen 1326 472 35.6 Hydroxychloroquine sulfate 2335 829 35.5 Oxybutynin 6573 2201 33.5 Loratadine 13,511 4397 32.5 Suvorexant 4009 1298 32.4 Orlistat 3615 1166 32.3 Naproxen sodium 19,349 5975 30.9 Loratadine/pseudoephedrine sulfate 4309 1250 29.0 Dimethicone/loperamide hydrochloride 2849 819 28.7 Fesoterodine fumarate 2186 628 28.7 Fexofenadine/pseudoephedrine 1467 416 28.4 Miconazole nitrate 1016 288 28.3 Amphetamine aspartate/amphetamine sulfate/dextroamphetamine 2513 705 28.1 saccharate/dextroamphetamine sulfate Acetaminophen/aspirin/caffeine 2422 655 27.0 Omeprazole magnesium 4834 1305 27.0 Amphotericin B 1908 502 26.3 Sildenafil citrate 15,302 3985 26.0 Triamcinolone acetonide 4095 1066 26.0 Solifenacin succinate 5688 1440 25.3 Leflunomide 3933 966 24.6 Loperamide hydrochloride 2788 679 24.4 DI drug ineffective, FAERS FDA Adverse Event Reporting System purports of is presented to have under the conditions of use patient characteristics such as age and disease severity [15]. prescribed, recommended, or suggested in proposed This discrepancy of efficacy expectations between con- labeling thereof’’ [9]. This does not translate to effective- sumers and HCPs may be an additional reason relatively ness in every patient who takes the drug, as evidenced by more DI reports were submitted by consumers. the efficacy of some of the most frequently reported Our ability to attribute ineffectiveness to product quality products reporting DI (e.g., subcutaneous adalimumab issues relies at a minimum on accurate identification of the 40 mg weekly achieved an American College of product in question; however, prior studies have suggested Rheumatology 20 response rate in 53% of patients with that reliance on a product’s name or the reporting manu- rheumatoid arthritis, 74% of patients demonstrated a clin- facturer’s name alone may result in misclassification [4]. ical response from subcutaneous etanercept 0.4 mg/kg Nonetheless, ICSRs of AEs remain an important source of twice weekly for polyarticular juvenile idiopathic arthritis, postmarketing safety information. Safety issues relating to and 81% of patients responded to intradermal onabo- ineffectiveness have been evaluated by the FDA using tulinumtoxinA 50 units for[50% decrease in axillary FAERS data. One example includes an analysis of DI sweat production) [10–12]. In addition, consumer expec- reports for two generic methylphenidate extended-release tations regarding efficacy may differ from that of HCPs (ER) products. These generic ER products were formulated [13, 14]. For example, patients’ high expectations may be to be administered once daily. In this case, the types and shaped by general optimism, attitudes, and advertisements quality of DI reports received by the FDA were determined or other media representations. Although HCPs may gen- to be ‘‘useful’’ because they contained specific details erally anticipate that a medication will be effective, their describing the failure of therapeutic effect during the latter expectations may be refined by considering individual part of the day. The reports also provided specific Drug Ineffective Reports in Drug Safety Surveillance 97 Table 5 Characteristics of ‘‘drug ineffective’’ reports manually evaluated (N = 552) Recorded information Observation N % The suspect product of ‘‘drug ineffective’’ was specified Yes 526 95.3 from narrative field No 26 4.7 Suspect product’s type from narrative field Brand 415 75.2 Generic 42 7.6 Both 13 2.4 Unknown 56 10.2 NA 26 4.7 Most frequently identified products (top 3) Adalimumab 47 8.5 Etanercept 30 5.4 Naproxen 19 3.4 Medication switch reported Yes 34 6.2 No 518 93.8 A beneficial response to the product’s active ingredient Yes 75 13.6 prior to suspect product exposure No 71 12.9 Not reported/unknown 406 73.6 Suspect product was continued Yes 78 14.1 No 164 29.7 Not reported/unknown 310 56.2 Suspect product’s batch or lot number was reported Yes 95 17.2 No 457 82.8 Product application type NDA 330 60.1 BLA 106 19.1 ANDA 42 7.6 Multiple 1 0.2 Missing 73 13.1 PT(s) other than ‘‘drug ineffective’’ reported Yes 327 59.2 No 225 40.8 Concurrently reported PTs (Top 3) Fatigue 27 4.9 Product quality issue 25 4.5 Off label use 18 3.3 DI report was ‘‘useful’’ Yes 43 7.8 No 509 92.2 ANDA abbreviated new drug application, BLA biologics license application, DI drug ineffective, NA not available, NDA new drug application, PTs preferred terms Suspect product’s type was recorded as NA when there was no identifiable suspect product in the narrative field information that identified the manufacturer. These reports change can be useful especially if the patient responded to suggested the generic ER products may not produce the a prior batch or lot, but this must be considered in the same therapeutic effects for patients as the innovator context of disease progression. DI reports without enough product. This prompted an FDA investigation that resulted information to determine the suspect product in the narra- in regulatory actions [16]. tive field or coded solely with the PT ‘‘drug ineffective’’ The low yield of potentially useful DI reports in contrast may have limited utility. Opportunities may exist to use to the high volume of DI reports received annually high- natural language processing or other tools to identify fea- lights the need to develop strategies to efficiently identify tures in the narratives that may be indicative of a report’s which reports are more likely to contain useful information. utility. Additional studies are needed to identify other In the ‘‘useful’’ reports, generic products tend to be characteristics or surveillance algorithms that may be reported as a suspect product more frequently and were useful for prioritizing review of DI reports. often accompanied with the PT ‘‘product quality issue’’. Our study has certain limitations. We did not capture all The information about medication switching or batch or lot potential reports describing drug ineffectiveness because 98 T. Misu et al. Table 6 Characteristics of the 43 ‘‘drug ineffective’’ reports determined to be ‘‘useful’’ Recorded information Observation N % The suspect product of ‘‘drug ineffective’’ was specified Yes 43 100.0 from narrative field No 0 0.0 Suspect product’s type from narrative field Brand 19 44.2 Generic 22 51.2 Unknown 2 4.7 Most frequently identified products (top 3) Fentanyl 4 9.3 Alprazolam 3 7.0 Adalimumab 2 4.7 Medication switching was reported Yes 19 44.2 No 24 55.8 A beneficial response to the product’s active ingredient Yes 20 46.5 prior to suspect product exposure No 8 18.6 Not reported/unknown 15 34.9 Suspect product was continued Yes 9 20.9 No 16 37.2 Not reported/unknown 18 41.9 Suspect product’s batch or lot number was reported Yes 17 39.5 No 26 60.5 Product application type NDA 13 30.2 BLA 3 7.0 ANDA 16 37.2 Missing 11 25.6 PT(s) other than DI reported Yes 36 83.7 No 7 16.3 Concurrently reported PTs (top 3) Product quality issue 10 23.3 Product substitution issue 8 18.6 Feeling abnormal 5 11.6 ANDA abbreviated new drug application, BLA biologics license application, DI drug ineffective, NDA new drug application, PTs preferred terms A ‘‘useful’’ report meets criterion 1 and 2 plus one or more of the other four criteria listed in Table 1 we limited our search to the PT ‘‘drug ineffective’’. We definition of useful was based on the expertise of reviewers selected this PT because it was the most frequently with pharmacovigilance experience, which may limit reported MedDRA PT within the high-level term (HLT) reproducibility. ‘‘therapeutic and nontherapeutic responses’’, and the HLT encompasses concepts broader than ineffectiveness. Other reports describing the concept of ineffectiveness would not 5 Conclusion have been captured by relevant terms included in the HLT because they may be only coded with event-specific PTs. Many DI reports involved non-serious AEs and were For example, a report of ineffectiveness may describe reported by consumers. The most frequently reported increases in blood pressures after switching anti-hyper- products were used primarily for symptomatic manage- tensive drugs but may be coded with the PT ‘‘blood pres- ment. Although DI was the most commonly reported AE sure increased’’ rather than ‘‘drug ineffective’’. While we during the study period, the yield from ‘‘useful’’ reports determined the sample size needed to accurately estimate was low. The ‘‘useful’’ reports were often related to pro- the proportion of DI reports considered ‘‘useful’’, our duct quality issues and mentioned medication switching or resulting small sample of ‘‘useful’’ reports limits the gen- batch or lot information; consequently, these characteristics eralizability of the specific characteristics within the sub- can be important elements in useful report identification. set. Nonetheless, this sample captured findings such as the Additional studies to identify strategies for the efficient increased availability of lot or batch information relative to identification of reports more likely to contain useful the overall population of DI reports. Furthermore, our information are needed. Drug Ineffective Reports in Drug Safety Surveillance 99 Acknowledgements This project was supported in part by an branded and generic antiepileptic drugs. Clin Pharmacol Ther. appointment to the ORISE Research Participation Program at the 2015;97(5):508–17. https://doi.org/10.1002/cpt.81. CDER administered by the Oak Ridge Institute for Science and 5. FDA Adverse Event Reporting System (FAERS) Public Dashboard. Education through an agreement between the US Department of https://www.fda.gov/Drugs/GuidanceComplianceRegulatory Energy and the Center for Drug Evaluation and Research (CDER). Information/Surveillance/AdverseDrugEffects/ucm070093.htm. Dr. Misu conducted this research while he was an ORISE fellow in Accessed 8 Feb 2018. the Office of Surveillance and Epidemiology, CDER, US FDA. 6. Food and Drug Administration. Orange book: approved drug products with therapeutic equivalence evaluations. [Internet]. http://www.accessdata.fda.gov/scripts/cder/ob/. Accessed 8 Feb Compliance with Ethical Standards 7. Postmarketing reporting of adverse drug experiences. 21 CFR Funding No funding was used for the preparation of this manuscript. 314.80. 8. Poitras MF, Sene PD, Beliveau A. Extent of lack of efficacy Conflict of interest Takashi Misu, Cindy M. Kortepeter, Monica A. reporting in the Canada vigilance database. Pharmacoepidemiol Mun ˜ oz, Eileen Wu, and Gerald J. Dal Pan have no conflicts of Drug Saf. 2017;26(S2):589. interest. 9. Federal Food, Drug, and Cosmetic Act, 21 U.S.C. § 355 New drugs (d). Ethical approval This study was granted an exemption for review by 10. Adalimumab (Humira), BLA 125057. FDA Approved Labeling: the FDA Institutional Review Board. December 14, 2017. https://www.accessdata.fda.gov/drugsatfda_ docs/label/2017/125057s403lbl.pdf. Accessed 8 Feb 2018. Open Access This article is distributed under the terms of the 11. Etanercept (Enbrel), BLA 103795. FDA Approved Labeling: Creative Commons Attribution-NonCommercial 4.0 International October 6, 2017. https://www.accessdata.fda.gov/drugsatfda_ License (http://creativecommons.org/licenses/by-nc/4.0/), which per- docs/label/2017/103795s5561lbl.pdf. Accessed 8 Feb 2018. mits any noncommercial use, distribution, and reproduction in any 12. OnabotulinumtoxinA (Botox), BLA 103000. FDA Approved medium, provided you give appropriate credit to the original Labeling: October 2, 2017. https://www.accessdata.fda.gov/ author(s) and the source, provide a link to the Creative Commons drugsatfda_docs/label/2017/103000s5303lbl.pdf. Accessed 8 license, and indicate if changes were made. Feb 2018. 13. Hoffmann TC, Del Mar C. Clinicians’ expectations of the benefits and harms of treatments, screening, and tests: a systematic References review. JAMA Intern Med. 2017;177(3):407–19. https://doi.org/ 10.1001/jamainternmed.2016.8254. 1. Woodcock J, Behrman RE, Dal Pan GJ. Role of postmarketing 14. Hoffmann TC, Del Mar C. Patients’ expectations of the benefits surveillance in contemporary medicine. Annu Rev Med. and harms of treatments, screening, and tests: a systematic 2011;62:1–10. https://doi.org/10.1146/annurev-med-060309- review. JAMA Intern Med. 2015;175(2):274–86. https://doi.org/ 164311. 10.1001/jamainternmed.2014.6016. 2. Ishiguro C, Hall M, Neyarapally GA, Dal Pan G. Post-market 15. Dasgupta B, Combe B, Louw I, Wollenhaupt J, Zerbini CA, drug safety evidence sources: an analysis of FDA drug safety Beaulieu A, et al. Patient and physician expectations of add-on communications. Pharmacoepidemiol Drug Saf. 2012;21(10): treatment with golimumab for rheumatoid arthritis: relationships 1134–6. https://doi.org/10.1002/pds.3317. between expectations and clinical and quality of life outcomes. 3. Lester J, Neyarapally GA, Lipowski E, Graham CF, Hall M, Dal Arthritis Care Res (Hoboken). 2014;66(12):1799–807. https://doi. Pan G. Evaluation of FDA safety-related drug label changes in org/10.1002/acr.22371. 2010. Pharmacoepidemiol Drug Saf. 2013;22(3):302–5. https:// 16. Methylphenidate hydrochloride extended release tablets (generic doi.org/10.1002/pds.3395. Concerta) made by Mallinckrodt and Kudco. http://www.fda.gov/ 4. Bohn J, Kortepeter C, Munoz M, Simms K, Montenegro S, Dal drugs/drugsafety/ucm422568.htm. Accessed 8 Feb 2018. Pan G. Patterns in spontaneous adverse event reporting among http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Drugs - Real World Outcomes Springer Journals

An Evaluation of “Drug Ineffective” Postmarketing Reports in Drug Safety Surveillance

Free
9 pages

Loading next page...
 
/lp/springer_journal/an-evaluation-of-drug-ineffective-postmarketing-reports-in-drug-safety-kFQEO7m2YD
Publisher
Springer Journals
Copyright
Copyright © 2018 by The Author(s)
Subject
Medicine & Public Health; Pharmacotherapy; Pharmacology/Toxicology; Internal Medicine
ISSN
2199-1154
eISSN
2198-9788
D.O.I.
10.1007/s40801-018-0131-3
Publisher site
See Article on Publisher Site

Abstract

Drugs - Real World Outcomes (2018) 5:91–99 https://doi.org/10.1007/s40801-018-0131-3 OR IGINAL RESEARCH ARTIC L E An Evaluation of ‘‘Drug Ineffective’’ Postmarketing Reports in Drug Safety Surveillance 1,2 1 1 1 • • • • Takashi Misu Cindy M. Kortepeter Monica A. Mun ˜ oz Eileen Wu Gerald J. Dal Pan Published online: 23 February 2018 The Author(s) 2018. This article is an open access publication Abstract reports provided a batch or lot number (39.5 vs. 17.2%) and Introduction The most commonly reported adverse event, were coded with additional PTs beyond ‘‘drug ineffective’’ based on frequency of Medical Dictionary for Regulatory (83.7 vs. 59.2%), the most frequent of which were ‘‘product Activities (MedDRA) preferred terms (PTs), in the US quality issue’’ (23.3%) and ‘‘product substitution issue’’ FDA Adverse Event Reporting System (FAERS) database (18.6%). is ‘‘drug ineffective’’ (DI). This study aimed to describe the Conclusions DI was the most frequently reported adverse DI reports and provide data to support recommendations on event in the FAERS database; however, the yield from how to best evaluate these reports. these reports in terms of usefulness from a pharmacovigi- Methods We characterized all FAERS reports coded with lance perspective was low. Efficient strategies are needed the MedDRA PT ‘‘drug ineffective’’ received between 1 to identify which DI reports are more likely to contain September 2012 and 31 August 2016 using all other useful information. FAERS reports as a comparator. Additionally, we con- ducted a manual evaluation to identify informative data elements in the report narratives. Key Points Results During the study period, 247,513 (6.4% of all FAERS reports) DI reports were entered in FAERS. The most frequently reported adverse event in the Compared with non-DI reports, DI reports were more US FDA Adverse Event Reporting System (FAERS) likely to be reported by consumers (69.8 vs. 48.1%) and database was ‘‘drug ineffective’’ (DI). less likely to report a serious outcome (26.2 vs. 56.3%). Most DI reports in FAERS were reported by Most DI reports (88%) were from the USA. Manual eval- consumers and were non-serious. uation of 552 sample US reports identified 43 reports (7.8%) deemed ‘‘useful’’; a higher proportion of ‘‘useful’’ A minority of DI reports were deemed ‘‘useful’’. Many of these provided a batch or lot number, and the majority were coded with additional preferred Disclaimer The views expressed are those of the authors and do not terms beyond DI. necessarily represent the position of, nor imply endorsement from, the US FDA, the US Government, or the Japanese Pharmaceuticals and Medical Devices Agency. & Cindy M. Kortepeter Cindy.Kortepeter@fda.hhs.gov 1 Introduction Office of Surveillance and Epidemiology, Center for Drug Evaluation and Research, U.S. Food and Drug Adverse event (AE) reporting has been a central part of the Administration, Silver Spring, MD, USA 2 US FDA’s postmarketing drug safety surveillance for Office of Safety II, Pharmaceuticals and Medical Devices nearly 50 years [1]. While population-based databases Agency, Tokyo, Japan 92 T. Misu et al. have become additional sources of postmarketing safety concurrently reported PTs. The primary suspect product’s information, especially in the past two decades, individual application type was determined from a structured field case safety reports (ICSRs) of AEs remain an important provided by the submitting manufacturer. The most fre- source of postmarketing safety information [2, 3]. Each quently reported suspect products, by the product’s active data source has well-characterized strengths and limita- ingredient, were compared between the DI and non-DI tions. A major limitation in postmarketing ICSRs is the report groups. A product’s active ingredient was defined as frequent lack of sufficient detail to allow an independent the ingredient that has a pharmacological effect of a reviewer to make a reasonable assessment of the potential medicinal product reported by drug manufacturers. Addi- relationship between a drug and a reported AE [4]. For tionally, we identified the active ingredients with the example, lack of accurate product identification and rele- highest proportions of DI reports for products with at least vant clinical details limit the inferences that can be made 1000 total reports in the study period. from ICSRs, especially when there are alternative potential Following the high-level characterization of DI and non-DI explanations for the reported AEs. reports, we conducted a manual evaluation to assess the The most commonly reported AE, based on frequency of availability of informative data elements in the report narra- Medical Dictionary for Regulatory Activities (MedDRA) tives and to classify reports by their potential utility in the preferred terms (PTs), in the US FDA Adverse Event assessment of the relationship between a drug and the AE (of Reporting System (FAERS) database is ‘‘drug ineffective’’ drug ineffectiveness). This assessment was made to determine (DI) [5]. These DI reports in FAERS have not been whether the report was ‘‘useful’’ from a pharmacovigilance assessed systematically for quality and inferential value perspective. We defined ‘‘useful’’ as reports containing the from a pharmacovigilance perspective. The objective of necessary information that would prompt a reviewer to con- this study is to describe the DI reports in FAERS and sider further action, which in most cases would be obtaining provide data to support recommendations on how to best additional information. For this study, a ‘‘useful’’ report evaluate these reports. contains criteria 1 and 2 and at least one of the other four criteria, as listed in Table 1. An assessment of causality was not conducted in the determination of a ‘‘useful’’ report. All 2 Methods FAERS reports were stratified by initial FDA received dates (September 2012–August 2013, September 2013–August FAERS is an electronic database that currently contains 2014, September 2014–August 2015, September 2015–Au- over 14 million ICSRs describing AEs or medication gust 2016). Random sampling was conducted from DI errors. Approximately 95% of FAERS reports are submit- reports, reflecting the proportions of reports in the periods. ted to the FDA by drug manufacturers, whereas 5% (called We restricted our manual evaluation to US reports given the ‘‘direct reports’’) are submitted directly to the FDA [5]. limited reporting requirements for reports that manufacturers Any member of the public (e.g., consumers, healthcare receive from outside the USA. A sample size of 552 was professionals [HCPs]) can report an AE to the FDA or the determined for manual evaluation assuming a prevalence rate manufacturer. When the manufacturer receives a report, of DI reports with potential utility of 10% and a precision of they must in turn report the AE to the FDA in accordance 2.5%. The prevalence of usefulness was estimated by piloting with regulatory requirements. Reports in FAERS may a review of 120 reports. contain narrative free text describing an AE, a list of Table 1 Criteria used to classify reports as ‘‘useful’’ for manual products suspected in that event, and information identi- evaluation fying the reporter. In addition to the narrative description Criterion Description of the AE, reporters may supply additional information, such as past medical history, laboratory data, the names of 1 The suspect product of drug ineffective was clearly suspect drug manufacturers, or the event outcome. identifiable We searched FAERS for all reports received by the FDA 2 An informative narrative to support the reported drug between 1 September 2012 and 31 August 2016. The ineffectiveness AND one or more of the following four criteria: retrieved reports were stratified by those coded with and 3 MedDRA preferred term(s) in addition to ‘‘drug without the MedDRA PT ‘‘drug ineffective’’ for compari- ineffective’’ was reported son. We summarized the following report characteristics: 4 Suspect product’s batch or lot number was reported report type (i.e., manufacturer or direct), patient and 5 A beneficial response prior to the administration of the reporter attributes, reporter country, reported outcomes, suspect product was reported primary suspect product’s application type (i.e., new drug 6 Medication switching was reported application [NDA], abbreviated new drug application MedDRA Medical Dictionary for Regulatory Activities [ANDA], biologics license application [BLA]), and Drug Ineffective Reports in Drug Safety Surveillance 93 Data collected during the manual evaluation was aimed Compared with reports without DI, DI was more likely to at determining whether information relevant to the evalu- be reported by consumers (69.8 vs. 48.1%) and less likely ation of drug ineffectiveness was documented in the ICSR to have a serious outcome reported (26.2 vs. 56.3%). Most with respect to the suspected ineffective product. These DI reports (88%) were from the USA. While age distri- included (1) classification of the product as brand (inno- butions were relatively similar between the DI and non-DI vator version) or generic; (2) description of medication report groups (median age 57 years [interquartile range switching; (3) action taken with the product suspected to be {IQR} 43–67] and 58 years [IQR 43–69], respectively), DI ineffective; (4) presence of the suspected product’s batch reports were more often missing the patient age (49.5 vs. or lot number; and (5) concurrently reported PTs. In this 38.7%). In total, 62% of DI reports included additional evaluation, a product was considered ‘‘brand’’ if an inno- coded PTs. The most frequently co-reported PTs in the DI vator’s product name was used or if an active ingredient reports were ‘‘product quality issue’’ (4.2%), ‘‘pain’’ was described as brand in the narrative. A product was (3.8%), and ‘‘fatigue’’ (3.2%). Of the non-DI reports, the classified as ‘‘generic’’ if a non-innovator product was most frequently reported PTs were ‘‘death’’ (4.5%), specified or if an active ingredient was described as generic ‘‘nausea’’ (4.1%), and ‘‘fatigue’’ (3.9%). in the narrative. If a manufacturer name was provided in Table 3 shows the top 20 suspect product’s active the description, the manufacturer (and the suspect product) ingredients with the highest number of reports for the DI was identified as brand or generic using the FDA’s list of and non-DI groups. Adalimumab and etanercept were the Approved Drug Products with Therapeutic Equivalence top two products during the study period in both the DI and Evaluations (commonly known as the Orange Book)[6]. the non-DI groups. Relative to the non-DI reports, many of Reports describing the use of both brand and generic sus- the most frequently reported suspect product’s active pect products were classified as ‘‘multiple’’. Medication ingredients in the DI reports were those used for symp- switching was defined as a switch from one medicinal tomatic management (e.g., pain: naproxen, pregabalin, product to another with the same active ingredient, the gabapentin, buprenorphine, ibuprofen; allergies/asthma: same dosage form, and the same route of administration. loratadine, fexofenadine, fluticasone/salmeterol, albuterol) The switching patterns included (1) a patient taking drug A or where an effect is expected shortly after exposure (e.g., and drug A worked, the patient switched to drug B then DI onabotulinumtoxinA, sildenafil, insulin lispro). Suspect occurred; and (2) a patient experienced DI with drug A, products’ active ingredients with the highest number of DI then switched to drug B and drug B worked. A switch may reports in FAERS were adalimumab (8.0%), etanercept include brand to generic, generic to brand, or generic to (5.7%), naproxen (2.4%), pregabalin (2.3%), and onabo- generic switching. Reports describing ineffectiveness with tulinumtoxinA (2.2%). In all, 100 active ingredients a different batch or lot of the same product were excluded accounted for 80% of all DI reports during the study per- from the medication switching definition. PTs reported in iod. The proportion of DI reports within each active addition to DI were further classified into those describing ingredient was not proportional across suspect products. product quality issues or any AE other than a product For example, DI reports consisted of 41.2% of all reports quality issue. Reports were determined to contain addi- received with onabotulinumtoxinA, but only 6.8% of tional PTs associated with product quality issues if addi- etanercept reports given the large total number of etaner- tional PTs included at least one within the MedDRA high- cept reports. The products with the highest proportions of level group term (HLGT) ‘‘product quality, supply, distri- DI reports are provided in Table 4. bution, manufacturing and quality systems issues’’. All The results of the manual evaluation of 552 reports are included reports were evaluated independently by two presented in Table 5. We could identify the product related reviewers. Differences in evaluation between the two to the DI in the narrative field in 95.3% of the sampled reviewers were discussed as a group for consensus. The reports, of which 75.2% described a brand product. The cases determined to be ‘‘useful’’ were further compared most frequently identified products in the sampled reports with the remainder of the sample. We summarized the included adalimumab (8.5%), etanercept (5.4%), and resulting data using descriptive statistics. naproxen (3.4%). DI associated with medication switching was reported in 6.2%. Ineffectiveness when using a dif- ferent batch or lot of the same product was described in 3 Results 1.3% of DI reports. The suspect product was continued in 14.1% of reports. The suspect product’s batch or lot During the study period, 3.8 million reports were entered number was reported in 17.2%. Of the 552 reports, 59.2% into FAERS, of which 247,513 (6.4%) reports were coded were coded with additional PTs, the most frequent of which with the PT ‘‘drug ineffective’’. The overall characteristics were ‘‘fatigue’’ (4.9%), ‘‘product quality issue’’ (4.5%), of DI and non-DI reports are presented in Table 2. and ‘‘off-label use’’ (3.3%). 94 T. Misu et al. Table 2 Characteristics of Characteristic DI reports (N = 247,513) Non-DI reports (N = 3,625,330) ‘‘drug ineffective’’ and non- drug ineffective reports during N % N % the study period Report type Manufacturer 241,291 97.5 3,482,423 96.1 Direct 6222 2.5 142,907 3.9 Reporter type Consumer 172,834 69.8 1,744,954 48.1 Healthcare provider 69,770 28.2 1,692,858 46.7 Other 444 0.2 115,558 3.2 Missing 4465 1.8 71,960 2 Reporter country USA 217,966 88 2,673,274 73.7 Non-USA 29,547 12 952,056 26.3 Patient age (years) 0–17 6007 2.4 110,215 3 18–64 80,313 32.4 1,341,823 37 C 65 38,612 15.6 771,500 21.3 Missing 122,581 49.5 1,401,792 38.7 Patient gender Female 139,671 56.4 2,037,500 56.2 Male 82,326 33.3 1,261,002 34.8 Unknown/null 25,516 10.3 326,828 9 All outcomes Hospitalization 20,380 8.2 829,646 22.9 Death 4842 2 365,601 10.1 Disability 2832 1.1 67,991 1.9 Life threatening 2150 0.9 85,168 2.3 Required intervention 179 \0.1 10,919 0.3 Congenital anomaly 13 \0.1 14,408 0.4 Other 49,853 20.1 1,184,076 32.7 No serious outcome was reported 182,628 73.8 1,583,848 43.7 Primary suspect product’s application type NDA 144,168 58.3 1,973,700 54.4 BLA 48,946 19.8 822,300 22.7 ANDA 19,704 8 271,748 7.5 Missing 34,695 14 557,582 15.4 Additional PTs other than DI Reported 153,555 62 Not reported 93,958 38 ANDA abbreviated new drug application, BLA biologics license application, DI drug ineffective, FAERS FDA Adverse Event Reporting System, NDA new drug application, PT preferred term We determined that 43 of the 552 reports (7.8%) met our were generic in 51.2% (vs. 7.6% overall sample DI definition of ‘‘useful’’. Among the 43 ‘‘useful’’ reports reports). The most frequently reported products were fen- meeting criteria 1 and 2, a total of 20 reports (46.5%) met tanyl (9.3%), alprazolam (7.0%), and adalimumab (4.7%). an additional two or more criteria, and 12 reports (27.9%) Medication switching was reported in 44.2% (vs. 6.2% met an additional three or more criteria. Characteristics overall sample DI reports). The suspect product was dis- (including the criteria used to classify reports as ‘‘useful’’) continued in 37.2% (vs. 29.7% overall sample DI reports). of the 43 reports determined to be ‘‘useful’’ are in Table 6. A higher proportion of ‘‘useful’’ reports provided a batch or The suspect products identified from the narrative field lot number (39.5 vs. 17.2% overall sample DI reports). The Drug Ineffective Reports in Drug Safety Surveillance 95 Table 3 Most frequently reported suspect product’s active ingredients during the study period (top 20) Non-DI reports (N = 3,625,330) DI reports (N = 247,513) Product active ingredient N % Product active ingredient N % of DI reports % of all reports Adalimumab 200,482 5.5 Adalimumab 19,848 8.0 9.0 Etanercept 193,799 5.3 Etanercept 14,187 5.7 6.8 Calcium chloride, dextrose, magnesium 87,998 2.4 Naproxen sodium 5975 2.4 30.9 chloride, sodium chloride, sodium lactate Lenalidomide 77,064 2.1 Pregabalin 5600 2.3 13.0 Natalizumab 58,484 1.6 OnabotulinumtoxinA 5503 2.2 41.2 Rivaroxaban 57,412 1.6 Loratadine 4397 1.8 32.5 Rosiglitazone maleate 54,763 1.5 Sildenafil citrate 3985 1.6 26.0 Teriparatide 53,934 1.5 Gabapentin 3693 1.5 18.2 Interferon beta-1a 52,166 1.4 Fexofenadine hydrochloride 3462 1.4 36.8 Dimethyl fumarate 51,953 1.4 Dalfampridine 3429 1.4 14.1 Ribavirin 51,403 1.4 Buprenorphine 3419 1.4 13.2 Ambrisentan 49,966 1.4 Ibuprofen 3222 1.3 13.8 Levonorgestrel 48,390 1.3 Infliximab 3187 1.3 8.4 Denosumab 38,204 1.1 Polyethylene glycol 3350 2614 1.1 18.4 Pregabalin 37,473 1.0 Fluticasone propionate, 2541 1.0 12.5 salmeterol xinafoate Infliximab 34,602 1.0 Abatacept 2531 1.0 15.8 Peginterferon alfa-2a 33,640 0.9 Apremilast 2454 1.0 10.9 Insulin lispro 31,554 0.9 Levonorgestrel 2352 1.0 4.6 Aspirin 30,110 0.8 Albuterol sulfate 2351 0.9 16.9 Risperidone 27,266 0.8 Insulin lispro 2349 0.9 6.9 DI drug ineffective majority (83.7%) of the reports were coded with additional addition, responsible parties are not required to submit non- PTs beyond ‘‘drug ineffective’’ (vs. 59% overall sample DI serious AE reports from foreign marketing experience [7]. reports); ‘‘product quality issue’’ (23.3%), ‘‘product sub- The suspect products with the highest number of DI stitution issue’’ (18.6%), and ‘‘feeling abnormal’’ (11.6%) reports during the study period are used primarily for the were the three most frequently reported PTs. Of the 43 management of symptomatic conditions (e.g., adalimumab, ‘‘useful’’ reports, 44.2% included PTs related to product etanercept, naproxen, loratadine), suggesting that con- quality issues within the MedDRA HLGT ‘‘product qual- sumers have self-awareness of worsening or no improve- ity, supply, distribution, manufacturing, and quality sys- ment of their own subjective experiences. Similarly, the tems issues’’. suspect product with the highest proportion of DI reports within each active ingredient (docosanol) is used primarily for symptom management. Poitras et al. [8] conducted 4 Discussion similar research using the Canadian Vigilance Database and also identified high proportions of DI reports (ranging DI is the most frequently reported AE in the FAERS from 14.6 to 20% annually using the standardized Med- database. Our analysis demonstrated that, in contrast to all DRA query ‘‘lack of efficacy/effectiveness’’). The most other reports in the FAERS database, the majority of DI frequently reported drug classes in lack-of-efficacy reports cases did not report a serious outcome and were more were tumor necrosis factor-a inhibitors and proton pump likely to be reported by consumers. These reports were inhibitors [8]. primarily from the USA. The US Code of Federal Regu- Our manual evaluation revealed that many of the reports lations requires responsible parties (i.e., applicants, manu- lacked the clinical details needed to distinguish the repor- facturers, packers, or distributors) to submit adverse ted DI from disease progression. Interpretation of DI experiences associated with drugs or therapeutic biologic reports is complicated by the variable efficacy of approved products to the FDA, including experiences associated with products. Evidentiary standards for drug approval are that a failure to produce an expected pharmacologic action. In ‘‘substantial evidence that the drug will have the effect it 96 T. Misu et al. Table 4 Suspect product’s active ingredients with the highest proportion of ‘‘drug ineffective’’ reports in FAERS during the study period (top 25) Product’s active ingredient Total FAERS reports DI reports % DI Docosanol 4282 2124 49.6 OnabotulinumtoxinA 13,362 5503 41.2 Fexofenadine hydrochloride 9400 3462 36.8 Bupivacaine hydrochloride 1085 387 35.7 Diphenhydramine citrate/ibuprofen 1326 472 35.6 Hydroxychloroquine sulfate 2335 829 35.5 Oxybutynin 6573 2201 33.5 Loratadine 13,511 4397 32.5 Suvorexant 4009 1298 32.4 Orlistat 3615 1166 32.3 Naproxen sodium 19,349 5975 30.9 Loratadine/pseudoephedrine sulfate 4309 1250 29.0 Dimethicone/loperamide hydrochloride 2849 819 28.7 Fesoterodine fumarate 2186 628 28.7 Fexofenadine/pseudoephedrine 1467 416 28.4 Miconazole nitrate 1016 288 28.3 Amphetamine aspartate/amphetamine sulfate/dextroamphetamine 2513 705 28.1 saccharate/dextroamphetamine sulfate Acetaminophen/aspirin/caffeine 2422 655 27.0 Omeprazole magnesium 4834 1305 27.0 Amphotericin B 1908 502 26.3 Sildenafil citrate 15,302 3985 26.0 Triamcinolone acetonide 4095 1066 26.0 Solifenacin succinate 5688 1440 25.3 Leflunomide 3933 966 24.6 Loperamide hydrochloride 2788 679 24.4 DI drug ineffective, FAERS FDA Adverse Event Reporting System purports of is presented to have under the conditions of use patient characteristics such as age and disease severity [15]. prescribed, recommended, or suggested in proposed This discrepancy of efficacy expectations between con- labeling thereof’’ [9]. This does not translate to effective- sumers and HCPs may be an additional reason relatively ness in every patient who takes the drug, as evidenced by more DI reports were submitted by consumers. the efficacy of some of the most frequently reported Our ability to attribute ineffectiveness to product quality products reporting DI (e.g., subcutaneous adalimumab issues relies at a minimum on accurate identification of the 40 mg weekly achieved an American College of product in question; however, prior studies have suggested Rheumatology 20 response rate in 53% of patients with that reliance on a product’s name or the reporting manu- rheumatoid arthritis, 74% of patients demonstrated a clin- facturer’s name alone may result in misclassification [4]. ical response from subcutaneous etanercept 0.4 mg/kg Nonetheless, ICSRs of AEs remain an important source of twice weekly for polyarticular juvenile idiopathic arthritis, postmarketing safety information. Safety issues relating to and 81% of patients responded to intradermal onabo- ineffectiveness have been evaluated by the FDA using tulinumtoxinA 50 units for[50% decrease in axillary FAERS data. One example includes an analysis of DI sweat production) [10–12]. In addition, consumer expec- reports for two generic methylphenidate extended-release tations regarding efficacy may differ from that of HCPs (ER) products. These generic ER products were formulated [13, 14]. For example, patients’ high expectations may be to be administered once daily. In this case, the types and shaped by general optimism, attitudes, and advertisements quality of DI reports received by the FDA were determined or other media representations. Although HCPs may gen- to be ‘‘useful’’ because they contained specific details erally anticipate that a medication will be effective, their describing the failure of therapeutic effect during the latter expectations may be refined by considering individual part of the day. The reports also provided specific Drug Ineffective Reports in Drug Safety Surveillance 97 Table 5 Characteristics of ‘‘drug ineffective’’ reports manually evaluated (N = 552) Recorded information Observation N % The suspect product of ‘‘drug ineffective’’ was specified Yes 526 95.3 from narrative field No 26 4.7 Suspect product’s type from narrative field Brand 415 75.2 Generic 42 7.6 Both 13 2.4 Unknown 56 10.2 NA 26 4.7 Most frequently identified products (top 3) Adalimumab 47 8.5 Etanercept 30 5.4 Naproxen 19 3.4 Medication switch reported Yes 34 6.2 No 518 93.8 A beneficial response to the product’s active ingredient Yes 75 13.6 prior to suspect product exposure No 71 12.9 Not reported/unknown 406 73.6 Suspect product was continued Yes 78 14.1 No 164 29.7 Not reported/unknown 310 56.2 Suspect product’s batch or lot number was reported Yes 95 17.2 No 457 82.8 Product application type NDA 330 60.1 BLA 106 19.1 ANDA 42 7.6 Multiple 1 0.2 Missing 73 13.1 PT(s) other than ‘‘drug ineffective’’ reported Yes 327 59.2 No 225 40.8 Concurrently reported PTs (Top 3) Fatigue 27 4.9 Product quality issue 25 4.5 Off label use 18 3.3 DI report was ‘‘useful’’ Yes 43 7.8 No 509 92.2 ANDA abbreviated new drug application, BLA biologics license application, DI drug ineffective, NA not available, NDA new drug application, PTs preferred terms Suspect product’s type was recorded as NA when there was no identifiable suspect product in the narrative field information that identified the manufacturer. These reports change can be useful especially if the patient responded to suggested the generic ER products may not produce the a prior batch or lot, but this must be considered in the same therapeutic effects for patients as the innovator context of disease progression. DI reports without enough product. This prompted an FDA investigation that resulted information to determine the suspect product in the narra- in regulatory actions [16]. tive field or coded solely with the PT ‘‘drug ineffective’’ The low yield of potentially useful DI reports in contrast may have limited utility. Opportunities may exist to use to the high volume of DI reports received annually high- natural language processing or other tools to identify fea- lights the need to develop strategies to efficiently identify tures in the narratives that may be indicative of a report’s which reports are more likely to contain useful information. utility. Additional studies are needed to identify other In the ‘‘useful’’ reports, generic products tend to be characteristics or surveillance algorithms that may be reported as a suspect product more frequently and were useful for prioritizing review of DI reports. often accompanied with the PT ‘‘product quality issue’’. Our study has certain limitations. We did not capture all The information about medication switching or batch or lot potential reports describing drug ineffectiveness because 98 T. Misu et al. Table 6 Characteristics of the 43 ‘‘drug ineffective’’ reports determined to be ‘‘useful’’ Recorded information Observation N % The suspect product of ‘‘drug ineffective’’ was specified Yes 43 100.0 from narrative field No 0 0.0 Suspect product’s type from narrative field Brand 19 44.2 Generic 22 51.2 Unknown 2 4.7 Most frequently identified products (top 3) Fentanyl 4 9.3 Alprazolam 3 7.0 Adalimumab 2 4.7 Medication switching was reported Yes 19 44.2 No 24 55.8 A beneficial response to the product’s active ingredient Yes 20 46.5 prior to suspect product exposure No 8 18.6 Not reported/unknown 15 34.9 Suspect product was continued Yes 9 20.9 No 16 37.2 Not reported/unknown 18 41.9 Suspect product’s batch or lot number was reported Yes 17 39.5 No 26 60.5 Product application type NDA 13 30.2 BLA 3 7.0 ANDA 16 37.2 Missing 11 25.6 PT(s) other than DI reported Yes 36 83.7 No 7 16.3 Concurrently reported PTs (top 3) Product quality issue 10 23.3 Product substitution issue 8 18.6 Feeling abnormal 5 11.6 ANDA abbreviated new drug application, BLA biologics license application, DI drug ineffective, NDA new drug application, PTs preferred terms A ‘‘useful’’ report meets criterion 1 and 2 plus one or more of the other four criteria listed in Table 1 we limited our search to the PT ‘‘drug ineffective’’. We definition of useful was based on the expertise of reviewers selected this PT because it was the most frequently with pharmacovigilance experience, which may limit reported MedDRA PT within the high-level term (HLT) reproducibility. ‘‘therapeutic and nontherapeutic responses’’, and the HLT encompasses concepts broader than ineffectiveness. Other reports describing the concept of ineffectiveness would not 5 Conclusion have been captured by relevant terms included in the HLT because they may be only coded with event-specific PTs. Many DI reports involved non-serious AEs and were For example, a report of ineffectiveness may describe reported by consumers. The most frequently reported increases in blood pressures after switching anti-hyper- products were used primarily for symptomatic manage- tensive drugs but may be coded with the PT ‘‘blood pres- ment. Although DI was the most commonly reported AE sure increased’’ rather than ‘‘drug ineffective’’. While we during the study period, the yield from ‘‘useful’’ reports determined the sample size needed to accurately estimate was low. The ‘‘useful’’ reports were often related to pro- the proportion of DI reports considered ‘‘useful’’, our duct quality issues and mentioned medication switching or resulting small sample of ‘‘useful’’ reports limits the gen- batch or lot information; consequently, these characteristics eralizability of the specific characteristics within the sub- can be important elements in useful report identification. set. Nonetheless, this sample captured findings such as the Additional studies to identify strategies for the efficient increased availability of lot or batch information relative to identification of reports more likely to contain useful the overall population of DI reports. Furthermore, our information are needed. Drug Ineffective Reports in Drug Safety Surveillance 99 Acknowledgements This project was supported in part by an branded and generic antiepileptic drugs. Clin Pharmacol Ther. appointment to the ORISE Research Participation Program at the 2015;97(5):508–17. https://doi.org/10.1002/cpt.81. CDER administered by the Oak Ridge Institute for Science and 5. FDA Adverse Event Reporting System (FAERS) Public Dashboard. Education through an agreement between the US Department of https://www.fda.gov/Drugs/GuidanceComplianceRegulatory Energy and the Center for Drug Evaluation and Research (CDER). Information/Surveillance/AdverseDrugEffects/ucm070093.htm. Dr. Misu conducted this research while he was an ORISE fellow in Accessed 8 Feb 2018. the Office of Surveillance and Epidemiology, CDER, US FDA. 6. Food and Drug Administration. Orange book: approved drug products with therapeutic equivalence evaluations. [Internet]. http://www.accessdata.fda.gov/scripts/cder/ob/. Accessed 8 Feb Compliance with Ethical Standards 7. Postmarketing reporting of adverse drug experiences. 21 CFR Funding No funding was used for the preparation of this manuscript. 314.80. 8. Poitras MF, Sene PD, Beliveau A. Extent of lack of efficacy Conflict of interest Takashi Misu, Cindy M. Kortepeter, Monica A. reporting in the Canada vigilance database. Pharmacoepidemiol Mun ˜ oz, Eileen Wu, and Gerald J. Dal Pan have no conflicts of Drug Saf. 2017;26(S2):589. interest. 9. Federal Food, Drug, and Cosmetic Act, 21 U.S.C. § 355 New drugs (d). Ethical approval This study was granted an exemption for review by 10. Adalimumab (Humira), BLA 125057. FDA Approved Labeling: the FDA Institutional Review Board. December 14, 2017. https://www.accessdata.fda.gov/drugsatfda_ docs/label/2017/125057s403lbl.pdf. Accessed 8 Feb 2018. Open Access This article is distributed under the terms of the 11. Etanercept (Enbrel), BLA 103795. FDA Approved Labeling: Creative Commons Attribution-NonCommercial 4.0 International October 6, 2017. https://www.accessdata.fda.gov/drugsatfda_ License (http://creativecommons.org/licenses/by-nc/4.0/), which per- docs/label/2017/103795s5561lbl.pdf. Accessed 8 Feb 2018. mits any noncommercial use, distribution, and reproduction in any 12. OnabotulinumtoxinA (Botox), BLA 103000. FDA Approved medium, provided you give appropriate credit to the original Labeling: October 2, 2017. https://www.accessdata.fda.gov/ author(s) and the source, provide a link to the Creative Commons drugsatfda_docs/label/2017/103000s5303lbl.pdf. Accessed 8 license, and indicate if changes were made. Feb 2018. 13. Hoffmann TC, Del Mar C. Clinicians’ expectations of the benefits and harms of treatments, screening, and tests: a systematic References review. JAMA Intern Med. 2017;177(3):407–19. https://doi.org/ 10.1001/jamainternmed.2016.8254. 1. Woodcock J, Behrman RE, Dal Pan GJ. Role of postmarketing 14. Hoffmann TC, Del Mar C. Patients’ expectations of the benefits surveillance in contemporary medicine. Annu Rev Med. and harms of treatments, screening, and tests: a systematic 2011;62:1–10. https://doi.org/10.1146/annurev-med-060309- review. JAMA Intern Med. 2015;175(2):274–86. https://doi.org/ 164311. 10.1001/jamainternmed.2014.6016. 2. Ishiguro C, Hall M, Neyarapally GA, Dal Pan G. Post-market 15. Dasgupta B, Combe B, Louw I, Wollenhaupt J, Zerbini CA, drug safety evidence sources: an analysis of FDA drug safety Beaulieu A, et al. Patient and physician expectations of add-on communications. Pharmacoepidemiol Drug Saf. 2012;21(10): treatment with golimumab for rheumatoid arthritis: relationships 1134–6. https://doi.org/10.1002/pds.3317. between expectations and clinical and quality of life outcomes. 3. Lester J, Neyarapally GA, Lipowski E, Graham CF, Hall M, Dal Arthritis Care Res (Hoboken). 2014;66(12):1799–807. https://doi. Pan G. Evaluation of FDA safety-related drug label changes in org/10.1002/acr.22371. 2010. Pharmacoepidemiol Drug Saf. 2013;22(3):302–5. https:// 16. Methylphenidate hydrochloride extended release tablets (generic doi.org/10.1002/pds.3395. Concerta) made by Mallinckrodt and Kudco. http://www.fda.gov/ 4. Bohn J, Kortepeter C, Munoz M, Simms K, Montenegro S, Dal drugs/drugsafety/ucm422568.htm. Accessed 8 Feb 2018. Pan G. Patterns in spontaneous adverse event reporting among

Journal

Drugs - Real World OutcomesSpringer Journals

Published: Feb 23, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off