Grégoire Gessain

M.D., Ph.D

Spotlight Medical

Pathologist and Chief Medical Officer at Spotlight Medical

PLCO (Learn more about this study)

PLCOI-1753

Dec 4, 2024

An artificial intelligence tool for therapeutic optimization in early invasive breast cancer patients.

Breast cancer is the most prevalent cancer worldwide, with 2.3 million new cases in 2020 and 690,000 fatalities (Sung et al., 2021). Predominantly affecting women, nearly 1 in 8 women will develop breast cancer during their lifetime. The hormone receptor-positive (HR+), human epidermal growth factor receptor 2-negative (HER2-) breast cancer category accounts for 74% of all breast cancers (Howlader et al., 2014). Patients are stratified into different stages based on the TNM classification (Giuliano et al., 2018), to help oncologists select treatments adapted to the severity of their patients’ condition. One of the oncologist's goals is to prevent cancer recurrence while minimizing treatment-induced toxicity for their patients. For decades, oncologists have relied on a standard of care based on a combination of surgery, radiotherapy, hormone therapy and chemotherapy. Under such treatment, the majority of HR+/HER2- early invasive breast cancer patients will never relapse 5 years after the initial diagnosis (Early Breast Cancer Trialists’ Collaborative Group, 2024). However, these patients undergo a high amount of treatment-related toxicities that could be avoided thanks to a more personalized risk stratification. Therefore, there is a need to develop new stratification tools to help oncologists identify the patients who will relapse and those who will not, to better adapt their treatment.

At Spotlight, we had access to a cohort of hundreds of patients with early invasive HR+/HER2- breast cancer, treated in France in 2012. We designed a new type of interpretable multimodal artificial intelligence algorithm based on the combined analysis of routine pathological slides and clinical data to predict relapse. So far, this technology managed to stratify patients into two distinct groups: a low-risk group with a low number of relapses and a high-risk group with a high number of relapses at 10 years.

Primary objective : the main goal is to confirm our preliminary results with external validation cohort of early-invasive breast cancer patients from the PLCO database.

Sung, Hyuna, et al. "Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries." CA: a cancer journal for clinicians 71.3 (2021): 209-249.
Howlader et al. US incidence of breast cancer subtypes defined by joint hormone receptor and HER2 status, J Natl Cancer Inst. 2014, April 28;106(5).
Giuliano, Armando E., Stephen B. Edge, and Gabriel N. Hortobagyi. "of the AJCC cancer staging manual: breast cancer." Annals of surgical oncology 25 (2018): 1783-1785.
Early Breast Cancer Trialists’ Collaborative Group. "Reductions in recurrence in women with early breast cancer entering clinical trials between 1990 and 2009: a pooled analysis of 155 746 women in 151 trials". The Lancet, Volume 404, Issue 10461, 1407-1418.

For each patient, we are looking for:
- the routine clinical and pathological data as indicated below
- at least one representative digital whole slide image (WSI) of the breast cancer, stained by H&E (at surgery, no biopsy)

Inclusion criteria:
- Women (regardless of menopausal status)
- Definitive surgery of the primary breast tumor
- ER+: ≥10% as defined by immunohistochemistry (IHC) according to American Society of Clinical Oncology (ASCO)/College of American Pathologists (CAP) guidelines for hormone receptor testing (Hammond et al., 2010).
- HER2-: score IHC 0, 1+ or 2+ (if IHC 2+ then FISH negative) according to ASCO/CAP guidelines for HER2 testing (Wolff et al., 2018).
- Early stage resected invasive breast cancer without distant metastases.
- One available representative H&E digital slide of the breast tumor (at surgery, no biopsy).

Exclusion criteria:
- Patients who did not give consent to participate in a research study.
- Metastatic (including contralateral axillary lymph nodes).
- Patients with inflammatory breast cancer.
- Patients with a history of cancer (except non-melanoma skin cancer or cervix CIS).

List of routine clinicopathological data:
- Sex
- Age at diagnosis or at surgery
- Ethnic origin
- Weight
- Height
- History of cancer
- Menopausal status
- Surgery (type and date)
- Resection status (surgical margin)
- Radiotherapy
- Chemotherapy (and regimen: adjuvant, neo-adjuvant)
- Hormonotheray
- Histological grade (SBR-EE): I, II, III
- Estrogen receptor (ER) IHC %
- Progesteron receptor (PR) IHC %
- Ki67 IHC %
- Tumor size (mm)
- Number of positive lymph nodes
- Loco-regional relapses (and date)
- Distant metastatic relapses (and date)
- New cancer (type and date)
- Vital status (dead or alive)
- Date and cause of death
- Date of last follow-up

Hammond et al. “American Society of Clinical Oncology/College ofAmerican Pathologists Guideline Recommendations for Immunohistochemical Testing of Estrogen and Progesterone Receptors in Breast Cancer.” Journal of Clinical Oncology 28.16 2010.
Wolff et al. “Human Epidermal Growth Factor Receptor 2 Testing in Breast Cancer: American Society of Clinical Oncology/College of American Pathologists Clinical Practice Guideline Focused Update.” Journal of Clinical Oncology, 36:20 2018.