F Gutiérrez Nicolás

Adherencia y toxicidad de los inhibidores de la tirosinquinasa en leucemia mieloide crónica

Objective: To analyze adherence and toxicity of tyrosine kinase inhibitor (TKIs) therapy in patients diagnosed with chronic myeloid leukemia (CML). Method: A 18-months retrospective observational study (Ja - nuary 2011-June 2012) which included all patients diagnosed with CML in a secondary hospital (550 beds) and were treated with imatinib, dasatinib or nilotinib. It was collected the following variables: sex, age at diagnosis, years of treatment and side effects. Adherence was evaluated using SMAQ questionnaire and recording dispensations. Results: 25 patients were included and all but two (92.0%) experienced side effects to imatinib,83.3% to dasatinib and 66.7% to nilotinib. The average adherence was 71.3%. There was identified as possible parameters of lack of adherence the female patients (55.6 % vs. 66.7%, p = 0.586), older than 50 (55.6 % vs. 83.3 %, p = 0.125), more than four years of treatment (70.0 % vs. 57.1 %, p = 0.521) and the presence of certain side effects (gastrointestinal disorders and musculoeskeletal pain). Conclusions: Almost one third of patients were considered nonadherent to therapy. Although the sample size did not allow us to establish a statistically significant relation between adherence and the variables analyzed, the clinical relevance of these results show the importance of future studies with larger populations to confirm the trends established in this study.

GRP-078 Guideline For Albumin Use: Effect on Cost Saving

Background Albumin has been widely used in clinical practise. While some of these indications are supported by the results of randomised studies, others are based only on clinical experience and have not been proved in prospective studies. Efforts should be made to define the indications for albumin use, so that patients gain the maximum benefit from its administration. Purpose To evaluate the cost saving obtained by the implementation of a guideline for albumin use in a 737-bed hospital. Materials and Methods Retrospective study that compared albumin use in two periods: July–September 2012 vs. July–September 2011. In June 2012 the guideline for albumin use was distributed to the medical staff. Physicians were requested to complete a form for each albumin order indicating the type and amount of albumin, the clinical service, and the indication for use. Albumin use data and costs were obtained from pharmacy service management system (SAP®) and were tabulated using the Excel® software. Results The total amount of albumin ordered during the study period was 29.360 g (€63,246) vs. 53.195 g (€108,617) for the same period during 2011, which means a reduction of 45%. In terms of cost, the saving obtained amounted to €45,371 (58%). The albumin use by specialty had also changed; a major decrease in use of albumin was observed for Anaesthesiology 4,000 g (75%), General Surgery 3,080 g (65%), Nephrology 4,900 g (64%), Internal Medicine 3,860 g (56%), Haematology 1,410 g (53%) and Digestive 1,400 g (30%). On the other hand, Haemodialysis significantly increased its use of albumin to 2,805 g (65%), although within the approved indication of plasmapheresis. Conclusions An albumin use guideline with restrictions focused on albumin prescriptions had suficient efficacy to reduce consumption and save cost. In our hospital guideline the cost of implementation decreased a 58% (€181,484 per year). No conflict of interest.

4CPS-212 Relationship between ugt1a1*28 and serum bilirubin levels

Background UGT1A1 polymorphisms have been relatated to interindividual variability effectiveness and toxicity in many drugs. In addition, the presence of mutated alleles in this gene has also been identified historically as Gilbert’s disease which exhibits high levels of unconjugated bilirubin. Purpose Analysing the relationship between total serum bilirubin levels and rs8175347 (*28) polymorphism in the UGT1A1 gene. Material and methods Observational, retrospective and unicentre study of 2 years was carried out. Patients with colorectal cancer who have been tested for gDNA determination of UGT1A1 genotype for clinical practice were included. Clinical data were obtained using the application SAP®. A high level of total-bilirrubine was considered as at least a 90% determination, with total-bilirubin higher than 1 mg/dL. Polymorphism *28 of UGT1A1 was established by analysing the genomic DNA of a peripheral blood sample. Genetic characterisation was carried out using the LightClycler® 480 platform and specific allele HybProbe fluorescent probes. The relationship between the UGT1A1 genotype and levels of total serum bilirrubine were determined by univariable statistical analysis. Patients were requested to sign an informed consent form prior to the inclusion. Results One hundred and seventy-three patients were included in the study, with a median age of 62 years (81–27) and 62.3% were males. 46,2% of participants had WT genotype: 21.2% (n=17) of them showed high levels of total serum bilirubin. On the other hand, 53.7% of patients had mutant alleles (*1/*28. *28/28), of which 36.5% (n=34) showed high levels of bilirubin (p=0.003). Conclusion Our results show that the presence of *28 allele in UGT1A1 is associated with high levels of serum bilirubin. With these results, we feel that this finding could provide the clinician with a tool to detect patients with high risk of drug toxicity such as irinotecan or pazopanib, among others. No conflict of interest

5PSQ-131 Mobile applications to check drug interactions: qualitative and quantitative analysis

Background The number of medical apps has increased exponentially in recent years, with more than 2 30 000 available. Because of the lack of regulation, some of these apps may offer inaccurate content or may not reach the minimum quality standards in order to be used by healthcare professionals. Purpose Analyse the availability of drug interaction checker apps for mobile devices and their quality according to the Mobile App Rating Scale (MARS score). Material and methods Cross-sectional study performed in October 2017 to find and classify the best mobile applications to check drug interactions according to MARS score. A search was conducted on two major mobile platforms: Apple’s App Store and Google Play Store. The keyword used to identify the initial sample was ‘drug interaction’. The exclusion criteria were: No drug searcher available or drug searcher only available for a specific drug class. No health and fitness or medicine category. No English language. Pay subscription app. Not updated in the last 36 months. The selected apps were downloaded in a smartphone and in a tablet of both systems in order to be analysed. The app’s quality and reliability was measured by means of MARS. This is an app quality rating tool that provides a measure of different features of health apps. It consists of 19 items clustered in four categories: engagement, functionality, aesthetics and information. Each item is rated in a 1–5 points scale (1-inadequate to 5-excellent). The degree of agreement between the selected apps was not analysed. Data collection and statistical analysis were performed in a Google Drive spreadsheet. Results Of the 139 apps identified, 12 met the inclusion and exclusion criteria. The mean MARS score was 3.01 (1.93–4.28). The mean social score was 4.03. The five apps with best MARS score (0–5) were ‘Medscape’ (4. 28), ‘Drugs.com Medication Guide’ (4.08), ‘Pharmacist Pro-Drug Interaction Checker’ (3.61), ‘Pocket Pharmacist’ (3.55) and ‘Assist UK-Drug Interactions’ (3.26). Conclusion There is a high amount of apps to check drug interactions but only few have enough quality to be used with guarantees by healthcare professionals in their clinical activity. Reference and/or Acknowledgements 1. Stoyanov SR, et al. Mobile app rating scale: a new tool for assessing the quality of health mobile apps. JMIR mHealth and uHealth2015. No conflict of interest

4CPS-147 Determination of tpmt and nudt15 polymorphisms in a patient with common variable immunodeficiency

Background The lack of activity in thiopurine-methyltrasnferase (TPMT) is related to severe toxicity in the use of thiopurines (azathioprine, 6-mercaptopurine and thioguanine). This lack of activity is often due to genetic polymorphisms located at TPMT, and recently demonstrated in NUDT15 genes. Purpose To describe the case of a patient with inflammatory elbow disease (IBD) and a congenital immunodeficiency who is being prescribed azathioprine and who has been tested for TPMT and NUDT15 genetic polymorphisms determination. Material and methods Male, 21-years-old patient who was diagnosed at the age of eight years with common variable immunodeficiency (CVI) with gastrointestinal manifestations. In 2002 the patient was diagnosed with ulcerative jejunitis and after receiving different first-line treatments (mesalazine and corticosteroid) was not able to obtain a good control of the disease. Thus, his doctor decided to start treatment with azathioprine. Given the risk of severe immunosuppression derived from the CVI and the use of azathioprine the decision was taken to make a genetic polymorphism analysis of TPMT (rs2842934, rs2842934, rs1800460, rs1800584 and rs1142345) and NUDT15 (rs116855232, rs147390019, rs554405994 and rs186364861). DNAg was carried out using the Ramos et al. (2015) method and the characterisation was implemented using PCR and DNA sequencing. Results The patient showed a wild-type genetic profile for the polymorphism analysed. Consequently, he was prescribed with azathioprine with a complete dosage of 100 mg per day (2 mg/kg/d). After 3 months of treatment the patient had maintained a neutrophil normal range (higher than 2000 neutrophils per mm3) and, in addition, achieved a good control of the illness. Therefore the patient continued with the azathioprine at the usual dosage. Conclusion The integration of pharmacy services in the multidisciplinary teams is facilitating the implementation of pharmacogenetics in daily clinical practice in our hospitals. This kind of determinations provides the prescription clinicians with tools to improve the effectiveness and safety of treatments. In this case we have given an example in which the determination of a genetic WT profile for TPMT and NUDT15 has allowed for the full use of azathioprine dosage from the beginning of the treatment, which therefore resulted in an adequate control of the disease. No conflict of interest

5PSQ-081 Determination of genetic polymorphisms in tpmt and nudt-15 in the paediatric onco-haematologic patient. preliminary results

Background Genetic-polymorphisms in thiopurine-methyltransferase (TPMT) and Nudix-hydrolase15 (NUDT15) have been related to higher risk of toxicity associated with administration of 6-mercaptopurine(6-MP). Purpose To describe the implementation of polymorphisms determination in TPMT and NUDT15 in paediatric patients by a simple and economic method. Material and methods A multi-centre, prospective, observational study with a expected duration of 32 months was carried out. Participants were patients younger than 18 years who received treatment with 6-MP. Single nucleotide polymorphisms (SNPs) analysed were: TPMT (rs1800462;rs1800462; rs1800460;rs1142345 and rs1800584) and NUDT15 (rs116855232;rs147390019;rs554405994 and rs186364861). DNAg extraction was carried out using the Ramos et al. method and genotyping was done using PCR and subsequent DNA sequencing. The study was approved by the hospital’s Ethical Committee (CEIC). Legal guardians were requested to sign an informed consent form prior to inclusion. Results During the first 8 months, nine patients were included, with an average age of 3.5 (1–18)and 62.5% of them were females. Six of the included patients (66.6%) were diagnosed with acute lymphoblastic leukaemia, two with non-Hodgkin’s lymphoma (22.2%) and one with acute myeloid leukaemia. Eighty-one genetic-determinations were carried out. None of the patients presented a high-risk genotype for the TPMT gene. One of the children showed a medium-risk genotype *1/*3B,*1/*3C, but after 3 months of treatment with 6-MP he has not shown toxicity. This patient also showed a wild-type genotype for the NUDT15 gene which could explain the absence of toxicity during the treatment. Another patient has shown a heterozygous genotype for the rs116855232 and rs554405994 (NUDT15 gene). This patient has not already received treatment with 6-MP so we cannot evaluate the mutation influence yet. Conclusion Althought we have not found patients with high-risk polymorphisms in TPMT yet, we support the implementation of this screening because the presence of this genotypes is related to severe toxicity and even death-risk in these patients. We also have completed the procedure with the determination of mutations in the NUDT15 gene, increasing the probability of identifying patients with low tolerance to 6-MP. To our knowledge, the present study is the first to evaluate the effect of polymorphisms in both TPMT and NUDT15 in the treatment with 6-MP, so definitive results could cidentify how those polymorphisms affect the toxicity related to 6-MP. References and/or Acknowledgements 1. Ramos-Díaz. 2015May;75(5):1095–8. 2. Kotur, et al. 2015;16(15):1701–12. No conflict of interest

4CPS-146 Ustekinumab and adalimumab for psoriasis patients who are no-responders to etanercept: a comparative effectiveness study

Background Adalimumab and ustekinumab have demonstrated high effectiveness in the treatment of moderate-severe psoriasis in randomised controlled trials. There is, however, limited data available on the comparative effectiveness of ustekinumab and adalimumab in psoriasis patients unsuccessfully treated with a first biologic line with etanercept. Purpose To evaluate the comparative effectiveness of adalimumab and ustekinumab in patients previously treated with etanercept using PASI 90 score. Material and methods A single-centre, retrospective, observational, comparative study was carried out from 1 November 2011 to 31 November 2015. Participants were patients with moderate-severe psoriasis that, after unsuccessful etanercept therapy, were treated with adalimumab or ustekinumab. An unblinded revision of each patient’s clinical history was carried out to assess clinical data. The primary analysis compared the percentages of patients in each treatment group who achieved ≥90% improvement from baseline PASI score (PASI 90) at week 12. Secondary endpoints included percentages of patients with PASI 90 at week 96. Statistical analysis was performed with the SPSS 22.0 software. Results Thirty-four psoriasis patients were included in the study: 15 (44.1%) patients received adalimumab and 19 (55.9%) received ustekinumab as a second-line therapy. The median age in adalimumab and ustekinumab group were 58 (SD 6.7) and 50 years (SD 17.3) (p=0.08). After 12 weeks of study treatment, 68.4% of ustekinumab-treated patients (13/19) achieved a PASI 90 response against 46.6% (7/15) in the adalimumab group (p=0.2). At week 96, more patients had a PASI 90 in the ustekinumab group compared with the adalimumab group, but the difference was not statistically significant (68.4% versus 46.6%; p=0,2). Conclusion Previously studies have shown that adalimumab and ustekinumab are effective after anti-TNF inhibitors’ therapy. However, to our knowledge, the present study is the first to evaluate the comparative effectiveness measured as PASI 90 of ustekinumab and adalimumab in psoriasis patients that failed with etanercetp. Our results suggests that there is no significant difference in the efficacy of ustekinumab between ustekinumab and adalidumab in the percentage of patients achieving PASI90. Of course, these results need to be evaluated with randomised and prospective clinical trials. References and/or Acknowledgements 1. J Dermatolog Treat2015;26(3):217–22. 2. J Eur Acad Dermatol Venereol2011;25(9):1037–40. No conflict of interest

OHP-005 Carbapenem restriction and its long term effectivness in an intensive care unit

This abstract was published in error and withdrawn at the author’s request.

CP-225 UGT1A1*28 polymorphism and irinotecan effectiveness

Background UGT1A*28 polymorphisms have been associated with an increase in SN-38, the active metabolite of irinotecan. Thus some authors also related the presence of this mutated allele with an increase in the effectiveness of irinotecan treatment. Purpose To evaluate the influence of the UGT1A1*28 polymorphism on the effectiveness of irinotecan. Material and methods This was a prospective, observational, 4 year single centre study (November 2012–May 2016). All adult colorectal cancer patients treated with the FOLFIRI protocol (irinotecan 180 mg/m2/fluorouracil 400 mg/m2/fluorouracil 2400 mg/m2/leucovorin 200–400 mg/m2) were included. Inclusion criteria were: ECOG 0–1, haemoglobin > 10 g/dL, leucocytes >3000/mm3 and platelets >100 000/mm3). Effectiveness was evaluated as progression free survival (PFS) and overall survival (OS). The rs8175347 UGT1A1 polymorphism was established by analysing the genomic DNA of a peripheral blood sample. Genetic characterisation was carried out using LightClycler 480 platform and specific allele HybProbe fluorescent probes. The study was approved by the hospital’s ethical committee (CEIC) and classified as EPA-SP by the Spanish Agency for Drugs and Health Products (AEMPS) with GNC-QUI-2013-01 code. Patients were requested to sign an informed consent form prior to inclusion. Results The study included 34 patients, average age 60 (27–81) years, of which 77.7% were men. 90.6% of patients were treated with anti-VEGFR or anti-VEGFA, and irinotecan was prescribed as secondline treatment. 44.4% of patients showed UGT1A1 wild-type (WT) alleles, while 41.2% and 14.7% had heterozygous and mutated homozygous alleles, respectively. After 4 years of follow-up, median PFS and OS were 7.0 and 23.0 months for patients with any mutated allele in the UGT1A1 gene, while for patients with the WT genotype, values were 8.0 (p=0.4590) and 15.0 (p=0.6128) months, respectively. Moreover, median PFS and OS were 4.0 (p=0.648) and 44.0 (p=0.1628) months for patients with the *28/*28 genotype and 7.0 (p=0.650) and 23.0 months (p=0.8354) for heterozygous patients. Conclusion Our results show that the rs8175347 polymorphism in UGT1A1 does not influence the effectiveness of irinotecan. Prospective randomised studies with a large number of patients are required to establish if this polymorphism influences the effectiveness of irinotecan therapy. No conflict of interest

PKP-005 Determination of genetic polymorphisms affecting metabolism of thiopurines

Background Thiopurines are widely used in different treatments. However, exposure to this drug can cause toxicity in the patient due to polymorphisms in the gene coding NUDT15, an enzyme involved in the metabolic activity of thiopurines. Carriers of defective alleles of NUDT15 accumulate large amounts of active metabolites which can cause serious damage to DNA. Purpose To develop a methodology to identify mutations that abrogate or reduce the activity of NUDT15. In this way a more personalised therapy to the patient can be applied with associated low cost. Material and methods Genetic variants of NUDT15 that cause a decrease or absence of total enzyme activity were identified: rs116855232 (called *6), rs147390019 (*7), rs186364861 (*8) and rs554405994 (*9). Determination of polymorphisms was performed by PCR and subsequent sequencing. For this purpose, pairs of primers that flank the region of interest were designed. Given the proximity of the polymorphisms on the chromosome, each primer pair amplified a region containing two of the four polymorphisms, so only two pairs of amplification primers were designed: F:GCA TCA CTA TGA GTT TAT TAG TAG C/R:CAC CAG ATG GTT CAG ATC TTC for *6 and *7 and F:ACG CAT TAC GCA CCG C/R:GCT CAC CCG AAC TCC AGA T for *8 and *9 polymorphisms. A primer was also designed within each amplified region for sequencing: CAC TAT GAG TTT ATT AGT AGC AAG (*6 and *7) and CGC TAT GAC GGC CAG (*8 and *9) . Primer design was performed with the GeneRunner programme and with the online analysis tool Primer-Blast, to confirm the specificity of pairs of primers. Reading the chromatograms was carried out with the programme Mega. DNA extraction was performed from a drop of blood deposited on paper Whatman 9031.1 Results Regions encompassing the four polymorphisms were amplified using only two primer pairs. The yield of the reaction was optimum, allowing its subsequent sequencing. The direct costs associated with determination of four markers was €27 (€6.75 for each polymorphism). Conclusion Genotyping NUDT15 allowed individualised treatment doses, as a carrier for any of these mutations. Future studies will determine the initial dose of the drug. We wished to show a simple and economical method, accessible to any laboratory with basic equipment in molecular biology, which allows detection of mutations in patients that adversely affect the metabolism of thiopurines. References and/or acknowledgements 1. Ramos-Díaz R, et al. Cancer Chemother Pharmacol2015;75:1095–8. No conflict of interest