Abstract Background 3 CDK4/6 inhibitors (CDKi) were approved 2015-7 and are standard of care for metastatic hormone-positive breast cancer, in first line with letrozole or second with fulvestrant. The Phase III randomised controlled trials (RCTs) show similar hazard ratios (HR) for progression-free survival (PFS) of 0.46-0.58. We compared them head-to-head and examined subgroups for differential response. Methods We retrospectively reviewed all patients who received CDKi up to March 2020 at our Oncology centre and the 4 local hospitals we serve. Kaplan-Meier method was used to generate time-to-event curves and Log-rank test was used to compare differences. Results 208 patients had median age 61 and follow-up 20 months. Palbociclib was most frequently used (128 patients) followed by abemaciclib (60) and ribociclib (20). Overall, median PFS was 20.8 months (95% confidence interval, CI, 17.2-28.0) and median OS was not reached. On the Kaplan-Meier curve there is a visible benefit with ribociclib, but due to the small group size, this did not reach statistical significance (p = 0.143). Patients received CDKi with aromatase inhibitor (176) or fulvestrant (32). Median PFS was 22.8 and 9.5 months respectively; HR 0.32 (95% CI 0.17-0.59). Median OS was not reached in either group; HR 0.22 (95% CI 0.09-0.56). Out of the ER strongly-positive patients (160), those with PR 7-8 did better than those with PR 0-4, with median PFS 28.7 and 14.0 months respectively; HR 0.40 (95% CI 0.22-0.72). Median OS was not reached for PR 7-8, and 31.8 months for PR 0-4. 115 patients received 1 (20%), 2 (13%) or more (12%) subsequent therapies. Patients who received everolimus at any point had better outcomes. Median OS was not reached in the everolimus group, and 28.5 months in the rest; HR 0.22 (95% CI 0.11-0.44). As patients suitable for everolimus may represent a distinct population, a comparison of baseline characteristics is presented [Table 1]. Patients with invasive ductal cancer (IDC) or lobular (ILC) had median PFS 17.1 and 24.2 months respectively; HR 0.69 (95% CI 0.43-1.09). Median OS was not reached in either group; HR 0.54 (95% CI 0.26-1.10). Median number of cycles was 17. Treatment was ongoing in 42% of patients at data cut-off. Dose was reduced in 43% and delayed in 74%. Mean dose intensity was 85%. Reasons for stopping treatment were progressive disease (70%), toxicity (20%) or unrecorded (10%). 60% experienced toxicities of grade 3 or above, or which resulted in dose adjustment. The commonest were neutropenia (38%), diarrhoea (11%) and lethargy (10%). Less frequent were mucositis, nausea, vomiting and anorexia (2% each). Of 68 patients with accessible documentation of radiological response, 41 had partial and 27 had complete response. Conclusions Our outcomes are comparable to those achieved in RCTs, despite poorer baseline characteristics [Table 2]. Our data suggest OS benefit with ribociclib. Predictors of good outcome include subsequent everolimus, strong PR+, strong ER+ and ILC. The prognostic benefit of lobular histology has not been definitively shown elsewhere. Table 1.comparing baseline characteristics between the everolimus subgroup and the restSubsequent therapy included everolimus n=35Subsequent therapy excluded everolimus n=61DemographicsMedian age (range)62 (30-83)58 (34-90)PS 0-1 (%)34 (97%)51 (84%)PS 2-3 (%)1 (3%)8 (13%)Tumour characteristicsLeft-sided breast cancer (%)20 (57%)27 (44%)Right-sided breast cancer (%)10 (29%)25 (41%)Bilateral (%)0 (0%)2 (3%)Data missing (%)5 (14%)7 (11%)Grade 1 (%)1 (3%)1 (2%)Grade 2 (%)17 (49%)24 (39%)Grade 3 (%)10 (29%)24 (39%)Data missing (%)7 (20%)12 (20%)Ductal (%)22 (63%)43 (70%)Lobular (%)8 (23%)7 (11%)Data missing (%)5 (14%)11 (18%)PR 7-817 (49%)18 (30%)ER 7-828 (80%)46 (75%)Visceral disease13 (37%)30 (49%)Metastatic at diagnosis (%)5 (14%)10 (16%)Median metastasis-free interval (range)9.5 years (0-37.7 years)4.6 years (0-36.0 years)Subsequent lines of therapy19 (26%)34 (56%)213 (37%)15 (25%)3 or more13 (37%)12 (20%) Table 2.comparing baseline characteristics between the current study and the key RCTsAuditRCTsMedian age (range)6160PS 0-188%99%PS 2-316%1%Ductal62%38%Lobular16%7%Histology missing22%55%PR 0-418%PR- 12% in 1L, 23% in 2LPR 7-848%PR+ 54% in 1L, 74% in 2LVisceral disease40%49%Bone-only18%23%Metastatic at diagnosis21%34%Not metastatic at diagnosis79%66%Median metastasis-free interval6.9 years51% >12m, 9% <=12m Citation Format: Lei Wang, Muhammad Ibrahim, Anthony Neal. Real-world outcomes with palbociclib, abemaciclib and ribociclib; experience of a tertiary oncology centre in the United Kingdom [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P1-18-33.