Abstract P2-19-09: A multi-centre prospective phase - 2 surgical study evaluating health related quality of life after immediate latissimus dorsi (LD) breast reconstruction: Duration of effects over 2 years

Abstract

Abstract Introduction: International decision-making bodies recommend integrating Patient Reported Outcome Measures (PROMs) alongside clinical outcomes. Methodology evaluating PROMs after types of breast reconstruction (BRR) has been poor in a systematic review of all studies since 19781. Few studies investigate the clinical meaning and magnitude of HRQL changes following types of immediate latissimus dorsi (LD) BRR (either implant-assisted (LDI) or tissue only autologous (ALD)) in relation to effective covariates: clinico-pathology, complications and adjuvant treatments. Twelve month data, showed poor role functioning and pain after ALD, with chemotherapy affecting all HRQL domains and no effects of radiotherapy (PMRT)2. Here we examined annual changes in each HRQL domain of a general cancer and breast cancer specific PROM and relationships with effective covariates over 36-months. Methods: An ethics approved multi centre prospective cohort study commenced in 2007. EORTC QLQ-C30, BR-23 and FACT-B were completed pre-operatively and at 3, 6, 12, 24 and 36 months post-operatively. Generalised estimating equations were used up to 24 months. The effect sizes (mean difference divided by the standard deviation) were assessed for all HRQL domains up to 36 months. Results: 142 patients were assessed (80 ALD, 62 LDI). The ALD group had a higher BMI (mean 27.6 versus 24.9 in LDI, p = 0.0025, effect size 3.04), more neo-adjuvant chemotherapy (12.5% versus 1.7% in LDI, p = 0.041) and symptomatic presentations (60% versus 45% in LDI, p = 0.047). Invasive tumours were significantly larger in the ALD (median 20, IQ range 14-30) compared to the LDI groups (median 15, IQ range 4.5-20.8) (p = 0.0031, effect size 2.99), with higher grade tumours (83% grades 2 and 3 for ALD, 70% for LDI) (p = 0.029), and lympho-vascular invasion (41% in ALD versus 21% in LDI, p = 0.019). The effects of explanatory variables on HRQL after 24 months were noticeably weaker compared to 12 months1. There were no significant differences between types of BRR regarding HRQL outcomes. PMRT adversely affected social functioning (p = 0.010, effect size = 2.6) and arm symptoms (p = 0.049, effect size = 1.97). Early complications (<3 months) significantly impaired role functioning (p = 0.017, effect size = 2.39), social functioning (p = 0.027, effect size = 2.21), pain (p = 0.035, effect size = 2.11) and physical wellbeing (p = 0.035, effect size = 2.11). Younger age adversely affected psychosocial functioning with decreased role functioning (p = 0.055, effect size = 1.92) and social functioning (p = 0.025, effect size = 2.24). Conclusion: Evaluation of HRQL effects particularly in functional domains (role and social) are important beyond 12 months up to 3 years, particularly following PMRT, early surgical complications and in younger women. The effects of PMRT may be a surrogate for more aggressive tumours affecting social functioning and arm symptoms. Disease-specific PROMS have a complimentary role alongside surgery-specific PROMS and require future analyses in cohort studies. 1. Winters ZE, et al. Annals of Surgery 2010; 252(6): 929-42. 2. Winters ZE, et al. BJS 2013; 100(2): 240-51. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P2-19-09.