Q: You have exhausted surgical, radiation, and standard chemotherapy options for a patient with an advanced epithelial malignancy. Yet the patient has a strong will to live and to advance science during palliative care. No active clinical trial is available. How would you evaluate pre-clinical data to try to identify an investigational drug for “expanded-compassionate” use?
A: Thank you for the question. This is a challenging situation. It is subjective and complex.
My answer – “it depends”. Let me give you 2 scenarios.
Scenario A: A 30-year-old surgeon, recently married with the love of his life who has just graduated from a neurosurgery residency, is expecting his first child and diagnosed with Stage IV non-small cell lung cancer. Has exhausted radiation, surgery, and standard chemotherapy. Has a strong will to live.
Scenario B: An 85-year-old man, who has a wife in the nursing home and a son in California and another son in Australia has been diagnosed with Stage IV non-small lung cancer. Has exhausted radiation, surgery, and standard chemotherapy. He lives alone and has a declining PS with multiple co-morbid conditions.
In the first case, yes, as an experimental investigator I would go to any lengths to try to provide investigational drug for expanded-compassionate use.
In the second case, I would likely arrange a family meeting to discuss the reality and lean towards providing best supportive palliative care.
Breakthroughs in oncology or any other field infers that successful outcome in the face of what was considered impossible. We all live in the future. If we do not take risky experiments, breakthroughs cannot be made. It could be argued that there is a 1 in 100 chance that the investigational agent for expanded compassionate use works. I would say that the chance is ONE and not zero. For his mother, father, wife and future child it is HOPE. It is with this hope that we live in.
How do I evaluate the pre-clinical and/ or clinical data for justifying compassionate use?
I refer you to the webpage of Personalized Cancer Therapy at MD Anderson. Here there are quite a few resources to define actionability of a gene. I am highlighting the main aspects from that site as below:
Actionable Gene: A gene is deemed actionable if 1) there are clinically available drugs that directly or indirectly target tumors with genomic alterations in the gene of interest with minimally preclinical evidence of their use in tumors with alterations in the gene of interest, and/or 2) there are clinical trials specifically selecting for patients with tumors harboring genomic alterations in the gene of interest.
Actionable Variant: A variant is deemed actionable if all of the following criteria are met: 1) the variant occurs in a gene deemed actionable, 2) the alteration type (mutations, amplification, etc) for that gene is deemed actionable, and 3) there is either published literature or data from the MD Anderson Zayed al Nahyan Institute for Personalized Cancer Therapy (IPCT) functional genomics platform that the alteration is likely to be tumor promoting (e.g. activating mutation in an oncogene or inactivating mutation in a tumor suppressor), or there is evidence that the variant confers sensitivity or resistance to a clinically available therapy.
Precision oncology decision support level of evidence classification: level of evidence for drug effectiveness in a specific tumor type harboring a specific biomarker*
1A | Drug is FDA-approved for the same tumor type harboring a specific biomarker.
1B | An adequately powered, prospective study with biomarker selection/stratification, or a meta-analysis/overview demonstrates a biomarker, predicts tumor response to a drug or that the drug is clinically effective in a biomarker-selected cohort in the same tumor type.
2A | Large-scale study demonstrates a biomarker is associated with tumor response to the drug in the same tumor type. This could be a prospective trial where biomarker study is the secondary objective or an adequately powered retrospective cohort study or a case-control study.
2B | Clinical data that the biomarker predicts tumor response to drug in a different tumor type.
3A | Single or few unusual responder(s), or case studies, show a biomarker is associated with response to drug, supported by scientific rationale.
3B | Preclinical data (in vitro or in vivo models or functional genomics) demonstrates that a biomarker predicts response of cells to drug treatment.
Bottomline, is that clinical studies provide highest level of evidence and pre-clinical studies lowest level of evidence (but at least provide some evidence for hope in the face of nothing).