William G. Nelson, MD, PhD, DSc Photo by Joe Rubino

ON HIS 1990 STORM FRONT ALBUM, Billy Joel sang, “I don’t know why I go to extremes,” lyrics he reportedly wrote as an apology to his then-wife for his mercurial personality. In cancer treatment, the reasons for extreme responses—such as the unexpected disappearance of cancer after a treatment—likewise often remain a mystery to researchers.

Typically, oncologists discuss expected treatment outcomes with patients based on clinical trial data. Using statistics from studies, doctors may describe, for example, a response rate, which is the percentage of treated patients who can expect to have a substantial reduction in tumor size. Oncologists may also mention median overall survival rates, which refer to how long at least half the patients live after they receive treatment or enroll in a clinical trial.

The continued dramatic increases in new anticancer drugs, along with the ready availability of cancer gene sequencing tests, may help to illuminate what went right for patients who have these uncommon but striking responses and to personalize care beyond the averages. In an often-cited example of a clinical trial showing the merit of this approach, the drug everolimus did not provide particular benefit for most of the 45 people with advanced bladder cancer who took it. However, one person in the study had a durable complete response. Subsequent gene sequencing of that person’s tumor tissue revealed a likely explanation for the benefit: The cancer contained acquired defects in the TSC1 and NF2 genes. These genetic mutations unleashed the cancer growth-promoting actions of an enzyme called mTOR, which was also the target of everolimus. As a result, this person appeared to be sensitive to mTOR inhibition by everolimus.

There have now been several attempts to systematically investigate and catalog extreme responders. In a 2020 National Cancer Institute-sponsored study, researchers analyzed tumor specimens from 111 people with different types of cancer who had exceptional responses to various treatments. Researchers analyzed tumor tissue for gene defects, immune cell infiltrates and other microscopic features. They linked almost one-quarter of the extreme responses to specific gene defects, including those affecting DNA damage repair and growth signaling, and to immune system engagement. The mechanistic basis for the other pronounced responses in the rest of the cancer cases remained an enigma. Nonetheless, the study findings provided insights into cancer cell vulnerabilities, which may pay off in the form of new therapeutic approaches in the future.

Rare exceptional responses pose significant challenges when it comes to Food and Drug Administration approvals for new anticancer drugs. Clinical trial designs tend to be oriented toward assessing common responses to ensure they will control cancer for most people. Biopharmaceutical companies often have limited interest in producing agents that work only rarely.

However, more detailed molecular analysis of tumors, along with artificial intelligence that can detect patterns in large datasets, may help unravel the secrets of exceptional responses. These insights could fuel an evolution that takes us from precision to ultraprecision medicine.

William G. Nelson, MD, PhD, DSc, is the director of the Johns Hopkins Kimmel Cancer Center in Baltimore.