Alpana (name changed) was looking to start a family, like any newly married woman in her early thirties. When she was diagnosed with hormone receptor sensitive early-stage breast cancer, the world came tumbling down on her and her family. A typical treatment plan would have included one year of chemotherapy and five years of hormone therapy. Alpana would not have been able to conceive during this time. If chemotherapy and hormone therapy are given to her as part of her usual treatment, the doctor offered the option of fertility preservation to her. However, she would have to wait another 5-6 years to conceive as a result. Alpana had to make a difficult decision because she was already in her 30s. As a ray of optimism, the doctor informed her that the tumor was small and that an AI-based Indian prognostic test was available to determine whether she needed chemotherapy and to plan the next course of treatment.
Alpana decided to have an AI-based prognostic test performed on her tumor sample, and the tumor was classed as ‘low-risk’ for cancer recurrence. To Alpana, what did this mean? It meant she could forego chemotherapy and focus on conceiving and giving birth for a year before beginning five years of hormone therapy. Alpana achieved her goal and gave birth to a healthy baby. She appreciates being a doting mother to her child five years after being diagnosed. Her hormonal therapy is working well for her right now.
Artificial intelligence (AI) is now widely used in all aspects of our life, including health care and diagnosis. Artificial intelligence-based solutions have greatly aided in the automation of patient data management, medical records, disease diagnosis accuracy, treatment planning, and optimization. In the field of cancer, AI and machine learning approaches are widely used in pathology and radiology to confirm malignancy and to better understand individual tumors so that patients’ responses to various treatments (called prognosis) such as endocrine therapy, chemotherapy, and immunotherapy can be predicted.
Breast cancer is becoming more common in both developed and developing countries. In 2020, 2.3 million women worldwide were diagnosed with breast cancer. According to the 2020 cancer statistics, breast cancer is the most common type of cancer in India, with over 57 percent of cases detected in advanced stages. In India, the number of women diagnosed with breast cancer under the age of 40 has nearly doubled in the last two decades, and 30% of breast cancer patients are under the age of 50. The main causes of increased mortality in breast cancer patients in India are advanced stage diagnosis and diagnosis at a young age.
Preliminary and in-depth information on a patient’s tumor via ‘prognostic testing’ has a huge impact on determining the best treatment strategy, which in turn improves ‘quality of life’ while lowering costs. For malignancies such as breast, prostate, and colon cancer, prognostic tests have been established. Prognostic tests for hormone receptor sensitive breast cancer patients like Alpana assist clinicians make informed decisions about the efficacy of chemotherapy for each patient.
Prognostic tests are often developed using sophisticated mathematical equations or AI-based techniques based on large amounts of patient data. The mathematical methodologies used and the patient data used determine the applicability of these tests. Prognostic tests based on patient data and logistic regression are often used in the United States and Europe. Alpana’s prognostic test was created by a Bengaluru-based business and is based on a support vector based artificial intelligence method. This test has been validated internationally and compared to Western tests. It has been in use for the past five years in India, Sri Lanka, the United Arab Emirates, Turkey, and other countries, with outstanding early data. AI approaches like Support Vector Machines (SVM) offer an advantage over other statistical/mathematical methods like logistic regression in that they can better simulate a complicated disease like cancer and so improve diagnostic accuracy, resulting in optimal treatment planning.
Furthermore, ‘deep learning’ based algorithms (another application of AI) outperform any human assistance in identifying intricate patterns portrayed on the stained tumor slide that are invisible to the naked eye. Once developed, deep learning-based prognostic testing will improve accuracy. One caveat: AI/ML-based approaches are very dependent on the quality of data used to develop them, thus it’s important to employ the best data possible.
As a result, the introduction of high-accuracy AI-based prognostic testing in breast cancer is a game changer. They assist in the customization of treatment for each unique patient, resulting in a decrease in the usage of chemotherapy, enhanced patient quality of life, and cost savings for the patient or payer.