Search

PURPOSE: Predicting short-term mortality in patients with advanced cancer remains challenging. Whether digitalized clinical text can be used to build models to enhance survival prediction in this population is unclear. MATERIALS AND METHODS: We conducted a single-centered retrospective cohort study in patients with advanced solid tumors. Clinical correspondence authored by oncologists at the first patient encounter was extracted from the electronic medical records. Machine learning (ML) models were trained using narratives from the derivation cohort, before being tested on a temporal validation cohort at the same site. Performance was benchmarked against Eastern Cooperative Oncology Group performance status (PS), comparing ML models alone (comparison 1) or in combination with PS (comparison 2), assessed by areas under receiver operating characteristic curves (AUCs) for predicting vital status at 11 time points from 2 to 52 weeks. RESULTS: ML models were built on the derivation cohort (4,791 patients from 2001 to April 2017) and tested on the validation cohort of 726 patients (May 2017-June 2019). In 441 patients (61%) where clinical narratives were available and PS was documented, ML models outperformed the predictivity of PS (mean AUC improvement, 0.039, P <.001, comparison 1). Inclusion of both clinical text and PS in ML models resulted in further improvement in prediction accuracy over PS with a mean AUC improvement of 0.050 (P <.001, comparison 2); the AUC was > 0.80 at all assessed time points for models incorporating clinical text. Exploratory analysis of oncologist's narratives revealed recurring descriptors correlating with survival, including referral patterns, mobility, physical functions, and concomitant medications. CONCLUSION: Applying ML to oncologists' narratives with or without including patient's PS significantly improved survival prediction to 12 months, suggesting the utility of clinical text in building prognostic support tools. [ABSTRACT FROM AUTHOR]

Introduction: Indigenous Māori women have a 60% higher breast cancer mortality rate compared with European women in New Zealand. We investigated differences in cancer biological characteristics and their impact on breast cancer mortality disparity between Māori and NZ European women. Materials and Methods: Data on 2849 women with primary invasive breast cancers diagnosed between 1999 and 2012 were extracted from the Waikato Breast Cancer Register. Differences in distribution of cancer biological characteristics between Māori and NZ European women were explored adjusting for age and socioeconomic deprivation in logistic regression models. Impacts of socioeconomic deprivation, stage and cancer biological characteristics on breast cancer mortality disparity between Māori and NZ European women were explored in Cox regression models. Results: Compared with NZ European women (n=2304), Māori women (n=429) had significantly higher rates of advanced and higher grade cancers. Māori women also had non-significantly higher rates of ER/PR negative and HER-2 positive breast cancers. Higher odds of advanced stage and higher grade remained significant for Māori after adjusting for age and deprivation. Māori women had almost a 100% higher age and deprivation adjusted breast cancer mortality hazard compared with NZ European women (HR=1.98, 1.55-2.54). Advanced stage and lower proportion of screen detected cancer in Māori explained a greater portion of the excess breast cancer mortality (HR reduction from 1.98 to 1.38), while the additional contribution through biological differences were minimal (HR reduction from 1.38 to 1.35). Conclusions: More advanced cancer stage at diagnosis has the greatest impact while differences in biological characteristics appear to be a minor contributor for inequities in breast cancer mortality between Māori and NZ European women. Strategies aimed at reducing breast cancer mortality in Māori should focus on earlier diagnosis, which will likely have a greater impact on reducing breast cancer mortality inequity between Māori and NZ European women. [ABSTRACT FROM AUTHOR]

Purpose Despite the benefits of adjuvant endocrine therapy for hormone receptor positive breast cancer, many women are non-adherent or discontinue endocrine treatment early. We studied differences in adherence to adjuvant endocrine therapy by ethnicity in a cohort of New Zealand women with breast cancer and its impact on breast cancer outcomes. Methods We analysed data on women ( n = 1149) with newly diagnosed hormone receptor positive, non-metastatic, invasive breast cancer who were treated with adjuvant endocrine therapy in the Waikato during 2005–2011. Linked data from the Waikato Breast Cancer Registry and National Pharmaceutical Database were examined to identify differences by ethnicity in adherence to adjuvant endocrine therapy and the effect of sub-optimal adherence on cancer recurrence and mortality. Results Overall, a high level of adherence of ≥80% was observed among 70.4% of women, which declined from 76.8% to 59.3% from the first to fifth year of treatment. Māori women were significantly more likely to be sub-optimally adherent (<80%) compared with European women (crude rate 37% vs. 28%, p = 0.005, adjusted OR = 1.51, 95% CI 1.04–2.17). Sub-optimal adherence was associated with a significantly higher risk of breast cancer mortality (HR = 1.77, 95% CI 1.05–2.99) and recurrence (HR = 2.14, 95% CI 1.46–3.14). Conclusions Sub-optimal adherence to adjuvant endocrine therapy was a likely contributor for breast cancer mortality inequity between Māori and European women, and highlights the need for future research to identify effective ways to increase adherence in Māori women. [ABSTRACT FROM AUTHOR]

Background: Indigenous Māori women experience a 60% higher breast cancer mortality rate compared with European women in New Zealand. We explored the impact of differences in rates of screen detected breast cancer on inequities in cancer stage at diagnosis and survival between Māori and NZ European women. Methods: All primary breast cancers diagnosed in screening age women (as defined by the New Zealand National Breast Cancer Screening Programme) during 1999-2012 in the Waikato area (n = 1846) were identified from the Waikato Breast Cancer Register and the National Screening Database. Stage at diagnosis and survival were compared for screen detected (n = 1106) and non-screen detected (n = 740) breast cancer by ethnicity and socioeconomic status. Results: Indigenous Māori women were significantly more likely to be diagnosed with more advanced cancer compared with NZ European women (OR = 1.51), and approximately a half of this difference was explained by lower rate of screen detected cancer for Māori women. For non-screen detected cancer, Māori had significantly lower 10-year breast cancer survival compared with NZ European (46.5% vs. 73.2%) as did most deprived compared with most affluent socioeconomic quintiles (64.8% vs. 81.1%). No significant survival differences were observed for screen detected cancer by ethnicity or socioeconomic deprivation. Conclusions: The lower rate of screen detected breast cancer appears to be a key contributor towards the higher rate of advanced cancer at diagnosis and lower breast cancer survival for Māori compared with NZ European women. Among women with screen-detected breast cancer, Māori women do just as well as NZ European women, demonstrating the success of breast screening for Māori women who are able to access screening. Increasing breast cancer screening rates has the potential to improve survival for Māori women and reduce breast cancer survival inequity between Māori and NZ European women. [ABSTRACT FROM AUTHOR]

Background Indigenous and/or minority ethnic women are known to experience longer delays for treatment of breast cancer, which has been shown to contribute to ethnic inequities in breast cancer mortality. We examined factors associated with delay in adjuvant chemotherapy and radiotherapy for breast cancer, and its impact on the mortality inequity between Indigenous Maori and European women in New Zealand. Methods All women with newly diagnosed invasive non-metastatic breast cancer diagnosed during 1999-2012, who underwent adjuvant chemotherapy (n = 922) or radiation therapy (n = 996) as first adjuvant therapy after surgery were identified from the Waikato breast cancer register. Factors associated with delay in adjuvant chemotherapy (60-day threshold) and radiation therapy (90-day threshold) were analysed in univariate and multivariate models. Association between delay in adjuvant therapy and breast cancer mortality were explored in Cox regression models. Results Overall, 32.4% and 32.3% women experienced delays longer than thresholds for chemotherapy and radiotherapy, respectively. Higher proportions of Māori compared with NZ European women experienced delays longer than thresholds for adjuvant radiation therapy (39.8% vs. 30.6%, p = 0.045) and chemotherapy (37.3% vs. 30.5%, p = 0.103). Rural compared with urban residency, requiring a surgical re-excision and treatment in public compared with private hospitals were associated with significantly longer delays (p < 0.05) for adjuvant therapy in the multivariate model. Breast cancer mortality was significantly higher for women with a delay in initiating first adjuvant therapy (hazard ratio [HR] =1.45, 95% confidence interval [CI] 1.05-2.01). Mortality risks were higher for women with delays in chemotherapy (HR = 1.34, 95% CI 0.89-2.01) or radiation therapy (HR = 1.28, 95% CI 0.68-2.40), although these were statistically non-significant. Conclusions Indigenous Māori women appeared to experience longer delays for adjuvant breast cancer treatment, which may be contributing towards higher breast cancer mortality in Māori compared with NZ European women. Measures to reduce delay in adjuvant therapy may reduce ethnic inequities and improve breast cancer outcomes for all women with breast cancer in New Zealand. [ABSTRACT FROM AUTHOR]