Cancer is associated with significant morbimortality globally. Advances in screening, diagnosis, management and survivorship were substantial in the last decades, however, challenges in providing personalized and data-oriented care remain. Artificial intelligence (AI), a branch of computer science used for predictions and automation, has emerged as potential solution to improve the healthcare journey and to promote precision in healthcare. AI applications in oncology include, but are not limited to, optimization of cancer research, improvement of clinical practice (eg., prediction of the association of multiple parameters and outcomes – prognosis and response) and better understanding of tumor molecular biology. In this review, we examine the current state of AI in oncology, including fundamentals, current applications, limitations and future perspectives.

e17587 Background: PARP inhibitors (PARPi) have been approved for maintenance therapy in newly diagnosed ovarian cancer patients with or without homologous recombination deficiency (HRD). In this meta-analysis, the latest clinical data to assess the efficacy and safety of this approach across different subgroups were synthesized. Methods: PubMed, Scopus, and Cochrane databases were searched for randomized controlled trials that compared maintenance PARPi with placebo in newly diagnosed ovarian cancer. The outcomes of interest included progression-free survival (PFS) according to HRD status, overall survival (OS), second PFS (PFS2), and time-to-first subsequent treatment (TFST). Heterogeneity was examined with I2 statistics. A random-effects model was used for outcomes with high heterogeneity. Results: A total of 6 Randomized clinical trials with 3609 patients were included, of whom 2348 (65%) received maintenance PARPi. PFS was significantly longer for HRD-positive patients receiving PARPi [Hazard ratio (HR) 0.44, 95% Confidence interval (CI) 0.37 – 0.52 p<0.001] and for HRD-negative patients [HR 0.71, 95% CI 0.54 – 0.92 p=0.009]. Maintenance PARPi was associated with improved OS in HRD-positive patients [HR 0.59, CI 0.47 -0.73, p<0.001] but not in unselected patients. PFS2 in the HRD-positive population was reported in three trials, and it favored the use of PARPi [HR 0.57, CI 0.43 – 0.76, p<0.001]. Adverse events ≥grade 3 were more common in patients on maintenance PARPi (p<0.001). The occurrence of myelodysplastic syndrome was numerically but not significantly higher in the PARPi group. Conclusions: Maintenance with PARPi in patients with newly diagnosed ovarian cancer significantly prolongs PFS regardless of HRD status. Moreover, PARPi maintenance improves OS and PFS2 in patients with BRCA and/or HRD-positive tumors compared to placebo. [Table: see text]

e12611 Background: Poly (ADP-ribose) polymerase inhibitors (PARPi) are approved for the treatment of germline BRCA mutated (gBRCAm) breast cancer (BC) patients in the adjuvant and metastatic settings. Several studies have explored the role of PARPi in the neoadjuvant setting both with and without chemotherapy. Here, we explore the efficacy and safety of neoadjuvant PARPi in early HER2-negative BC patients. Methods: We searched PubMed, Scopus, and the Cochrane Library databases for randomized (RCT) and non-randomized clinical trials (non-RCT) that reported the proportion of patients with pathological complete response (pCR) after neoadjuvant PARPi and safety. Statistical analysis was performed using R software. Data were pooled using the random effects model. Results: Analysis included 3 RCTs and 4 non-RCTs, with 942 HER-2 negative breast cancer patients, 293 harboring a gBRCA 1/2 mutation. In a pooled analysis, gBRCAm carriers achieved a significantly higher rate of pCR than BRCA wild-type patients (56% versus 37%, p=0.04) with neoadjuvant PARPi. There was no difference in pCR when PARPi were combined with carboplatin (+/- chemotherapy) (40% versus 44%, p=0.69). Among RCTs (n=457) patients received PARPi in combination with either Paclitaxel plus Carboplatin (PCb) (85%) or with Paclitaxel alone (P) (15%). PARPi were associated with a higher pCR which approached, but did not meet statistical significance (51% vs 41%; Odds Ratio [OR]: 1.30, 95%CI 0.97 to 1.74, p=0.08). gBRCAm was associated with higher pCR compared to wild type in RCTs analysis (60% versus 48%; OR: 1.64, 95%CI 1.01 to 2.68, p=0.05). The most common adverse events (AEs) were cytopenias, nausea, fatigue, alopecia, and dizziness. Hematological toxicities were the most frequent serious AEs, including anemia (15%), neutropenia (22%), and thrombocytopenia (3%). The combination of PARPi and carboplatin was associated with higher hematological AEs (78% versus 46%; P=0.002). Conclusions: The addition of PARPi in the neoadjuvant setting improves pCR in early-stage HER2-negative breast cancer and germline BRCA 1/2 mutations, but not for unselected patients. Combination of PARPi with carboplatin does not improve pCR and leads to substantially worse toxicity. Further studies are warranted to identify the best neoadjuvant PARPi schema.[Table: see text]

e21159 Background: Epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKI) are the standard therapy for EGFR-mutated non-small cell lung cancer (NSCLC) patients. Unfortunately, patients eventually develop resistance to EGFR-TKI and disease progression. Re-exposure after a drug-free interval may be an alternative to overcome tumor resistance. We performed a systematic review and meta-analysis to assess EGFR-TKI retreatment’s efficacy in advanced NSCLC. Methods: We systematically searched PubMed, Embase, and Cochrane databases for clinical trials and observational cohort studies evaluating EGFR-TKI retreatment in advanced NSCLC patients. We aimed to assess the objective response rate (ORR), disease control rate (DCR), and survival outcomes. Subgroup analyses were performed according to the type of EGFRmutation and the TKI drug used in retreatment. We further stratified studies to assess the efficacy of rechallenging with the same drug used initially or a different one. Heterogeneity was assessed using the Cochran Q test, and I2 statistics and random effects models were fitted. Results: We included 16 studies (7 prospective clinical trials, and 9 retrospective cohorts) with 806 patients. Most of them had adenocarcinoma (70.8%), were females (58.1%), and were non-smokers (74.5%). The most frequently TKI given as the initial treatment was gefitinib (58.1%), whereas in the rechallenge it was erlotinib (36.6%) followed by gefitinib (31.6%). In a pooled analysis of patients who were retreated with TKI, the median PFS was 4.1 months (95%CI 3.0 - 4.4), and OS was 12.6 months (95%CI 10.2 - 12.6). ORR was 16% (95%CI 10 - 22%) and DCR was documented in 63% (95%CI 0.54 - 0.72%). Patients harboring a sensitive EGFR mutation had a significantly higher DCR, compared to patients with EGFRT790M mutation, and those with unknown mutational status, (DCR: 70%, 62%, and 48%, respectively, p = 0.02). Patients rechallenged with gefitinib, erlotinib, or afatinib had a similar DCR of 60%, while patients re-exposed to osimertinib had a greater DCR of 70% (95%CI 59 - 80%), (p < 0.01). Regarding ORR, no significant difference was observed amongst the groups defined by the type of EGFR mutation (p = 0.74) or type of TKI used (p = 0.05). Rechallenge using the same versus a different TKI resulted in similar ORR and DCR. In a subgroup analysis of 102 patients who had disease control with the first TKI, 62% (95%CI 45 – 79%) achieved disease control with TKI rechallenge. Conclusions: Our meta-analysis suggests that a subgroup of advanced EGFR-mutated NSCLC patients who failed TKI treatment benefit from rechallenge with an EGFR-TKI after a TKI-free interval. Re-exposure with either the same or a different TKI was shown to be equally effective. Patients treated with osimertinib and those with EGFR-sensitive mutations have better responses to the treatment.

Background: Alzheimer’s disease (AD) has several risk factors. APOE4 is the main one, and it has been suggested that there may be a synergy between it and BCHE-K as a risk factor. Objective: To investigate the association between APOE4 and BCHE-K as a risk factor for AD. Methods: We searched PubMed, Web of Science, Embase, and Scopus on August 8, 2021 for studies that analyzed the association of APOE4 and BCHE-K with AD. The random effect model was performed in meta-analysis according to age group. A chi-square was performed with the meta-analysis data to verify if the effect found is not associated only with the E4 allele. Results: Twenty-one studies with 6,853 subjects (3,528 AD and 3,325 Controls) were included in the meta-analysis. The quality of the evidence is moderate. There is a positive E4-K association for subjects with AD as shown by the odds ratio of 3.43. The chi-square meta test, which measures the probability that the E4-K association is due to chance, has an odds ratio of 6.155, indicating that the E4-K association is not a random event. The odds ratio of an E4-K association in subjects with AD increases to OR 4.46 for the 65- to 75-year-old group and OR 4.15 for subjects older than 75 years. The probability that the E4-K association is due to chance is ruled out by chi-square meta test values of OR 8.638 and OR 9.558. Conclusion: The synergy between APOE4 and BCHE-K is a risk factor for late-onset AD.