ESMO Real World Data and Digital Oncology最新文献

Through the Clinical Innovation Network, Japan’s regulatory authorities have enhanced the development of registries that utilize real-world data (RWD). The Ministry of Health, Labour and Welfare has issued guidelines, whereas the Pharmaceuticals and Medical Devices Agency has conducted consultations to manage and verify the integrity of these registries, thus improving the framework for the effective use of RWD. The use of cancer registry data as an external control group has been promoted by regulatory bodies and academic institutions. Given the aforementioned background, several high-quality cancer registries, such as the ‘SCRUM-Japan Registry’, ‘MASTER KEY project’, and ‘GALAXY registry’, have been established. The SCRUM-Japan Registry has been instrumental in achieving the world’s first regulatory approval for human epidermal growth factor receptor 2 (HER2)-positive colorectal cancer, demonstrating the value of regulatory-grade registries in managing rare molecular subtypes. However, the broader adoption of registry data for regulatory use in Japan remains limited, primarily owing to the lack of clear standards for using RWD/real-world evidence (RWE) for drug approval. This uncertainty has made pharmaceutical companies hesitant to use such data for regulatory submissions. This review aimed to elucidate the perspectives and related guidelines of the regulatory authorities concerning cancer registries in Japan. In response, the ‘REALISE study’ was initiated to define the ‘relevancy’ and ‘reliability’ of data necessary for new drug approvals and to develop methodologies to ensure data reliability retrospectively. The findings of this study will inform the creation of draft guidelines aimed at broadening the application of RWD/RWE throughout Japan.

Background

The COVID-19 pandemic impacted outpatient clinic services globally. It is unknown how the pandemic affected the follow-up of surgically treated colorectal cancer (CRC) patients. This population-based study aimed to assess the trends in CRC follow-up consultations before and during the COVID-19 pandemic in the Netherlands.

Materials and methods

Nationwide health care activities data between January 2018 and July 2021 were merged with patient-level data from the Netherlands Cancer Registry of stage I-III CRC patients treated with surgical resection. The number of follow-up consultations per patient per year was calculated, and between-group differences were assessed with descriptive statistics. Trends in the number and setting of follow-up consultations were assessed using joinpoint regression analyses. Out-of-hospital follow-up was defined as written, telephone or video consultations.

Results

In total, 42 970 CRC patients were included. The median number of follow-up consultations per year per patient was 2.9 (interquartile range: 2.0-4.7). The median number of follow-up consultations increased with disease stage (P < 0.001) and was higher for patients <60 years of age (P < 0.001). The total number of follow-up consultations did not change during the study period (P = 0.333). The percentage of out-of-hospital follow-up increased from 23% to 80% between January and April 2020 (P < 0.001), and remained between 48% and 59% until the end of the study period.

Conclusions

This population-based study showed a great increased use of out-of-hospital consultations during CRC follow-up, which predominantly corresponded to the severity of the COVID-19 pandemic. Future studies should assess whether the use of out-of-hospital follow-up consultations has persisted after the pandemic.

Background

Avelumab first-line (1L) maintenance is recommended as the standard of care for patients with locally advanced or metastatic urothelial carcinoma (la/mUC) without disease progression following 1L platinum-based chemotherapy (PBC). We report results from READY, a real-world study of avelumab 1L maintenance in an Italian compassionate use program (CUP).

Patients and methods

In this prospective, noninterventional CUP, avelumab was provided on physician’s request to patients with la/mUC without disease progression following four to six cycles of 1L PBC, after approval by the local ethics committees, per Italian compassionate use regulations.

Results

Between January 2021 and March 2022, 414 patients received avelumab 1L maintenance and were assessable for survival/safety analyses; 79.2% were male and median age was 71 years. At data cut-off (30 July 2023), median follow-up was 20.30 months [95% confidence interval (CI) 19.78-20.93 months]. From the start of avelumab treatment, median overall survival (OS) was 26.22 months [95% CI 19.97 months-not estimable (NE); 12-month OS rate, 65.6%] and median progression-free survival was 7.63 months (95% CI 6.02-9.31 months). In patients who had received 1L carboplatin plus gemcitabine (n = 221) or cisplatin plus gemcitabine (n = 184), median OS (95% CI) was 25.10 months (19.97 months-NE) and not reached (16.05 months-NE), respectively. Clinical benefit was observed across other subgroups, including those based on age and best response to PBC. Any-grade treatment-related adverse events occurred in 112 patients (27.1%).

Conclusions

In READY, avelumab 1L maintenance showed clinical benefit in patients in Italy with la/mUC without progression following PBC, including across clinical subgroups, further supporting its use as the standard of care in this setting.

Background

Head and neck sarcomas (HNS) are rare and diverse cancers with distinct biology, unique treatment constraints and poor survival outcomes. Furthermore, HNS are understudied in Asians, and prospective clinical trials are untenable. To better understand HNS and improve treatment, real-world studies in Asians with accurate histological typing are thus needed.

Materials and methods

A retrospective cohort study of patients with histologically confirmed sarcoma diagnosis in the head and neck region between 1985 and 2023 was carried out at the National Cancer Centre Singapore. Multivariate Cox regression was used to analyse risk factors for overall survival (OS), and parametric time-to-event modelling was used to develop a prognostic calculator.

Results

A total of 275 patients were analysed. The 5-year OS was 43.2% (95% confidence interval 36.2% to 51.6%). Among demographic risk factors, a high incidence of radiotherapy-associated sarcomas in the population at 11.3% placed the population at higher risk for aggressive disease (decreased treatment response and poorer prognosis). With interventions, microscopically negative (R0) surgical resection margins were significantly associated with improved OS. Parametric time-to-event simulations suggested microscopically positive (R1) resections to also be beneficial for OS in locally advanced tumours and nonaggressive sarcoma histology, and improved greatly alongside high-dose radiotherapy.

Conclusion

We present the largest Asian HNS cohort, with diverse subtypes and disease extent. Our analysis highlights poor outcomes from a higher incidence of radiotherapy-associated disease, showing the challenging landscape of HNS in Asia. Through our prognostic calculator, we demonstrate how meaningfully curated real-world data in a rare disease entity can be used for the prediction of OS in individual patients with specific treatment approaches.

As artificial intelligence (AI) advances, oncologists stand at the forefront of a transformative era in healthcare. AI, which empowers machines to learn from data, make decisions, and carry out tasks typically requiring human intelligence, is revolutionizing our clinical landscape. It promises streamlined workflows, enhanced diagnostic accuracy, and personalized treatments tailored to each patient’s unique profile. In the vast sea of patient data, AI serves as a guiding compass, ensuring no detail is overlooked, amplifying clinical acumen, and refining treatment decisions. However, to ensure AI’s benefits reach patients effectively, it is imperative that oncologists actively guide its development and application. This overview aims to equip oncologists with the tools to critically appraise and influence the trajectory of AI in oncology, ensuring its integration leads to meaningful advances in patient care.

Background

Real-world data (RWD) provide essential insights into the effectiveness and safety of breast cancer treatments, particularly in diverse patient populations, where traditional clinical trials may have limitations. Integrating RWD into breast cancer research enhances the understanding of treatment outcomes and supports clinical decision-making, complementing the findings from controlled clinical studies.

Design

This article reviews the integration of RWD into breast cancer research, highlighting the benefits and challenges. Various sources of RWD, including electronic health records (EHRs), insurance claims, and patient registries, are examined, with a focus on their application in studies of triple-negative breast cancer. The article also explores the role of artificial intelligence (AI) in managing RWD, particularly through technologies like natural language processing (NLP) and predictive analytics, which enhance data collection, storage, and analysis.

Results

RWD has demonstrated significant value in informing clinical decision-making and improving patient outcomes in breast cancer treatment. The integration of AI into the management of RWD has provided deeper insights into patient outcomes and supported personalized treatment strategies. Specific studies leveraging RWD have shown improved understanding of breast cancer subtypes, such as triple-negative breast cancer, and enhanced the effectiveness of treatment protocols.

Conclusion

Despite the benefits, challenges remain in integrating RWD and AI into clinical practice, particularly regarding transparency, interpretability, and ethical considerations. Addressing these challenges requires robust data governance frameworks, interdisciplinary collaboration, and investment in advanced analytical tools. The potential for RWD and AI to transform breast cancer treatment and improve patient care is significant, underscoring the need for ongoing research and collaboration.

Background

Cyclin-dependent kinase 4/6 inhibitors (CDK4/6is) are widely used to treat hormone receptor-positive (HR+)/ human epidermal growth factor receptor 2-negative (HER2−) metastatic breast cancer (MBC). This study aimed to capture the real-world efficacy and tolerability of CDK 4/6is.

Patients and methods

Data were retrospectively collected from five centres in South West England between April 2017 and November 2022.

Results

Six hundred and sixty-six patients were included (median age 66 years; interquartile range 23-92 years). Five hundred and forty-four (82.7%) were treated with CDK4/6i as first-line therapy and 122 (18.3%) as second-line therapy. Median follow-up time was 28 months (range 0-76 months). Five hundred and thirty-seven received palbociclib (80.6%), 85 patients received abemaciclib (12.8%) and 44 received ribociclib (6.6%). Palbociclib and ribociclib most frequently caused neutropenia (38.2% and 26.4%, respectively) whilst abemaciclib caused diarrhoea (61.2%). Rates of dose reduction (DR) (between 53.8% and 59.2%) and time to first DR were similar for all agents (2-3 cycles). For first-line therapy, median progression-free survival (PFS) was 31 months (25-35 months) for palbociclib, 16 months [9 months-not reached (NR)] for abemaciclib and 44 months (21-NR) for ribociclib. Median overall survival (OS) was 47 months (41 months-NR) for palbociclib and was not reached for abemaciclib or ribociclib. Low patient numbers precluded analysis of second-line therapy. On multivariate analysis, visceral metastases and Eastern Cooperative Oncology Group performance status were associated with shorter PFS and OS, whilst DR was associated with longer PFS and OS.

Conclusion

These data demonstrate that CDK4/6is are an effective and safe treatment for metastatic HR+/HER2− breast cancer. Efficacy was in line with trial data and other real-world data. DR was associated with improved PFS and OS, suggesting that trials of CDK4/6is at a lower starting dose are warranted.

The performance of the augmented pathologist that works in synergy with artificial intelligence (AI) is generally accepted as the most accurate in comparison to AI standing alone and the general pathologist standing alone. Human–machine interactions triggered by the synergic daily work give rise to trust-related concerns and potential biases that need to be addressed. The long-term use of AI requires actions to prevent deskilling of the pathology workforce, and to ensuring appropriate education of future generations. Establishment of clear guidelines for the verification and validation of AI tools is crucial for the maintenance of high-quality cancer pathology.

Background

Modern electronic medical records (EMRs) contain a valuable amount of data. These data can be unlocked for research by manual data collection, which is highly labor intensive. Therefore, we explored whether automated text mining (TM) could be used to extract the performance status (PS) and progression-free survival (PFS) in a cohort of 328 non-small-cell lung cancer patients.

Materials and methods

Unstructured Dutch text data were derived from different EMR fields containing mainly information recorded during outpatient visits. A rule-based TM approach using regular expressions was used to extract PS and PFS in the R programming language. For PS, quantitative evaluation metrics, such as the weighted F1-score, were used to determine the accuracy of the TM-extracted data. For PFS, the median PFS was compared between the two approaches using the Kaplan–Meier method. In addition, the C-index was determined.

Results

A PS was obtained for 196 patients (60%) using the TM approach. In 189 (96%) patients, the TM-curated PS matched the manually curated PS. The weighted F1-score was 96.5%. The median PFS was 7.42 months for the manually curated data (n = 328) and 8.00 months for the TM-curated data (n = 301). The C-index was 0.916.

Conclusions

The developed TM approach is able to extract PS and PFS from the EMR with a very good performance. Therefore, this approach increases the efficiency of reliable data collection from EMRs, facilitating the use of real-world data (RWD) in clinical research.