Sperling Medical Group

reading & research

Artificial Intelligence in Medicine: Predicting Which Patients Benefit from ADT

Advanced prostate cancer (PCa) threatens to shorten a man’s life. Advanced means PCa is no longer contained in the gland but is still within the prostate bed and groin lymph nodes. In such cases, prostatectomy is generally not done. The rationale is to avoid putting a patient through surgery with no hope of cure, yet risking damaged quality of life due to side effects. Thus, radiation therapy is the usual standard of care. In addition, the patient may be put on a course of androgen deprivation therapy (ADT) with intent to boost the effectiveness of radiation.

ADT is also called chemical castration because it stops the male hormone production and/or blocks cells from taking hormones up. ADT is not curative, but it stops PCa in its tracks by depriving it of the hormonal fuel it needs. The benefit is delaying the spread of PCa to remote sites, including skeleton and other organs. This spread is called distant metastasis (DM). However, ADT damages quality of life. When the supply of androgens is cut off, side effects include loss of sex drive, breast tenderness, hot flashes, diminished bone density and muscle tone, fatigue, and moodiness.

Most men with advanced PCa accept the tradeoff of enduring such side effects in exchange for adding months or years before PCa outsmarts ADT and moves on to DM. The problem is, ADT does not benefit all patients, which raises the question, why put a patient through poor quality of life if you knew in advance it would not benefit him? Until now, however, there was no good way to predict who would gain time before DM, and who wouldn’t. Here’s where Artificial Intelligence can help.

New AI predictive model

A clinical team presented data on a new Artificial Intelligence (AI) model that can identify which patients are unlikely to gain more time from a course of ADT. The model was developed based on data from a nonprofit research group, NRG Oncology, that collects patient information on radiation therapy with or without ADT. By analyzing the characteristics of those whose ADT delayed DM, and those who did not, AI developed a predictive model linked with ADT benefit.

NRG Oncology conducted a study of 1594 men with advanced PCa who were randomly assigned to radiation plus or minus ADT for four months. When the cases were submitted to analysis according to the AI model, those who received ADT and were AI model positive benefited by a significant reduction in DM compared with those who received radiation alone. On the other hand, those who were negative had no benefit from the addition of ADT. The authors concluded that “an AI-based predictive model was able to identify prostate cancer patients, with predominately intermediate-risk disease, who are likely to benefit from short-term ADT.”[i]

The results were presented by Duke University Medical Oncologist Andrew Armstrong, MD at the 2023 annual meeting of the American Society of Clinical Oncology (ASCO, Chicago, IL). Armstrong shared that “… about 29% of our high-risk patients tested AI biomarker negative, and they could thus be spared the long-term [adverse] effects of ADT for 2 to 3 years… Conversely, [in] another group of what we call NCCN clinical intermediate-risk patients, about 43% were AI-biomarker positive and may benefit from longer extensions of their ADT to reduce distant metastases.”

Thus, with the assistance of AI, doctors and advanced PCa patients can decide in advance if adding ADT to radiotherapy is a worthwhile trade-off. You can find more blogs on AI in medicine at our companion website, Sperling Medical Group.

NOTE: This content is solely for purposes of information and does not substitute for diagnostic or medical advice. Talk to your doctor if you are experiencing pelvic pain, or have any other health concerns or questions of a personal medical nature.

[i] Spratt DE, Tang S, Sun Y, Huang HC et al. Artificial Intelligence Predictive Model for Hormone Therapy Use in Prostate Cancer. Res Sq [Preprint]. 2023 Apr 21:rs.3.rs-2790858.