CHICAGO—Clinicians can extend the lives of patients whose solid tumors are refractory to standard therapies by looking deep into the tumor’s genome in search of its Achille’s heel. Matching a mutation that defines the tumor’s vulnerability with a specific Phase I agent is particularly important for patients in the last 12 weeks of life. So said Daniel D. Von Hoff, MD, Physician in Chief and Senior Investigator at the Translational Genomics Research Institute in Phoenix, presenting the 2010 David A. Karnofsky Memorial Award Lecture here at the ASCO Annual Meeting, which he titled, “The Last 12 Weeks.”
“Phase I trials can be more therapeutic for the individual patient sitting in front of us right now. Do whatever it takes to maximize a patient’s chances of having a benefit from a specific Phase I agent. Let’s not just try the next new agent to come along.”
Over the last 35 years, Dr. Von Hoff has taken 317 new anticancer agents into clinical trials with four different Phase I clinical trial teams. These Phase I trials include patients who have not responded to multiple standard therapies, and have a life expectancy of 12 weeks, a Karnofsky status of at least 70%, and acceptable hepatic and renal function.
“People trust us in this time of their life for the best help we can possibly provide with a great sense of urgency,” Dr. Von Hoff said. “[A Phase I trial] may be the last thing our patients have an opportunity to try.”
Clonal Assay Success
Only three of the first 26 (8%) new agents he took into Phase I trials had evidence of clinical antitumor activity. To improve a patient’s chance of response, he began to utilize the human tumor cloning assay (HTCA).
“There are quite a few drugs we now use every day because a specific patient went on a specific Phase I agent suggested by the HTCA. When using the HTCA to select patients, we saw responses, dramatic responses, durable responses, in almost every Phase I trial.”
The agents first identified in this way include docetaxel, temsirolimus, irinotecan, and gefitinib.
Dr. Von Hoff provided a dramatic example of a 43-year-old man with abdominal mesothelioma whose disease had progressed on a handful of agents and who had been referred to Dr. Von Hoff in 1998. “We tested this patient’s tumor in the HTCA and found it to be resistant to many agents, but it was sensitive to the new anticancer agent LY2313514. We had been looking at this agent in the cloning assay to see where we might take it and what patients to put on trial,” he said.
The investigational agent was active in 2 of 3 mesothelioma patients tested as well as in 9 of 36 non-small cell lung cancer (NSCLC) patients.
The referred patient was sensitive to LY2313514, which is now known as pemetrexed, and the man had more than a four-year response to the drug. “This was the first patient we tested with the continual reassessment method of one patient per dose level,” he said.
Pemetrexed in combination with cisplatin has since proven to extend survival of mesothelioma patients beyond 12 months. Pemetrexed is also approved for non-small-cell lung cancer and for maintenance of non-squamous histology patients. “Where you see an initial response in a Phase I trial can be critically important to the development of the agent,” noted Dr. Von Hoff. “There is a good correlation between the first response in a Phase I trial and what the new agent is initially approved for.”
The cloning assay helped provide responses for patients in other Phase I trials and paid off in multiple areas of early drug development, including gemcitabine, irinotecan, topotecan, docetaxel, mitoxantrone, and gefitinib, among others. Then funding for the cloning assays dried up.
“Unfortunately, with less direct guidance on what new agent to use, the number of responses seen in our Phase I trials dropped off dramatically. It was obvious we needed new tools to measure sensitivity or targets in a greater percentage of patients,” Dr. Von Hoff said.
Measuring Tumor Sensitivity
He began to “work as hard as possible to find an Achille’s heel in a patient’s tumor before we put the patient into a specific Phase I trial. We call this theme ‘context of vulnerability.”
There are clinical contexts of vulnerability—for example, a non-smoking woman with adenocarcinoma of the lung, and genomic contexts of vulnerability, such as epidermal growth factor receptor mutations.
In his Phase I clinic, Dr. Von Hoff uses translational genomics to improve a patient’s chances of benefiting for a specific drug. A pilot trial used molecular profiling in two simple formats, immunohistochemistry (IHC) and microarray analysis, in 112 patients referred for Phase I studies.
The “surprising results,” he said, showed that 74% of the tumors had at least one potential target identified by IHC and 99% identified by microarray analysis.
Prospective Trial of IHC/Microarray Approach
“Even patients with a history of extensive prior treatment have tumors that can harbor targets—for example, estrogen receptor, for which we have standard treatments. We needed a prospective trial of the IHC/microarray approach to see how good it was for patients with advanced refractory disease who were ready for Phase I trials,” he said.
Dr. Von Hoff led just such a prospective trial in nine centers across the United States, using treatments suggested by molecular profiling, in hopes of favorably changing the clinical course for an individual patient with advanced refractory cancer who otherwise would be getting a Phase I agent.
The primary objective of the study was to compare progression-free survival, using a therapy suggested by molecular profiling with that of the patient’s most recent therapy, in effect using the patient as his or her own control. Of 66 patients treated, about one quarter responded positively, suggesting that the molecular profiling hypothesis was worth pursuing.
“Some patients had impressive shrinkage of their tumor, and about half had some tumor shrinkage. Some treatments were simple, such as exemestane for a breast cancer patient with unsuspected ER positivity,” Dr. Von Hoff said. “We can help some of these patients by using standard agents. It puts a system in place on which to add new technologies.”
Gene Sequencing, Comparative Genomic Hybridization, Complete Sequencing
A third study has used specific gene sequencing, comparative genomic hybridization, and complete sequencing of a patient’s cancer genome.
“Who says you can’t see responses on the first patient in a Phase I trial?” Dr. Von Hoff recalled a patient with metastatic basal cell carcinoma who had the PTCH1 mutation and was treated with the Hedgehog pathway antagonist GDC-449. The patient had a dramatic response with only mild side effects, he said.
This experience led him to consider a complete sequencing approach. “Complete sequencing of a cancer patient’s genome is hard. Our team realizes the massive amounts of information that needs processing, the time and effort. Also, how will you tell what abnormality drives the tumor? But you have to start somewhere. It’s not too early for the person with advanced refractory disease sitting in front of you now.”
He reported on the third person in the world, a 63-year-old man who has the rare pancreatic ampullary adenocarcinoma, whose tumor has been completely sequenced. “It took three to four weeks and 200 people hours to generate the data and three weeks of analysis including 400 hours of computing time.”
He and his research teams found a Kras mutation and a PTEN deletion. The patient will be treated with a combination of PI3 kinase inhibitor and mTOR inhibitors.
Dr. Von Hoff noted that the context of vulnerability in a patient’s tumor may be in the stroma of a patient’s tumor, not just the tumor cells.
When he profiled the 112 tumors in the pilot study, he also looked at the stroma and found the protein SPARC (secreted protein acidic and rich in cysteine) that came up in pancreatic cancer, melanoma, and other tumor types. Nab paclitaxel interacts with SPARC and is approved for treatment of women with breast cancer. Phase I/II trials of nab paclitaxel and gemcitabine in pancreatic cancer have shown improvement in survival to more than 12 months, and Phase III trials are now ongoing.
“There are many other targets in the stroma, including hyaluronan, Hedgehog pathway, and multiple others,” he said.
Finally, Dr. Von Hoff asked, “What do you do when in spite of all the guile and profiling you can muster against a patient’s tumor we have nothing we think will help? But actually we do have something that helps, and that is trying to help.”
“The practice of oncology and the area of Phase I trials is tough,” he continued. He likened the one patient in front of a clinician in a sea of patients to the story of the boy and the starfish. When an old man chastised the boy for trying to save hundreds of starfish that had washed up on shore, the boy tossed one last one back into the sea and said, “It made a difference for that one starfish.”
OT Tweets During the Lecture
OT Editor Serena Stockwell (@OncologyTimes) “live-tweeted” Dr. Von Hoff's Karnofsky talk on June 5:
* Karnofsky Awd winner (highest awd) is Daniel Von Hoff, talking on “The Last 12 Weeks”
* Von Hoff: Talk is dedicated to all the people treated in Phase 1 trials
* Von Hoff: Want 2 find ‘context of vulnerability' of tumor to maximize finding target in what may be pt's last 12 wks of life
* Von Hoff: Who says you can't see results in 1st pt treated in a Phase 1 trial? Definitely can, gave real-pt examples
* Von Hoff: pt's rare cancer and normal cells totally sequenced–just last Wed
* Von Hoff: the cancer was pancreatic ampullary adenocarcinoma
* Von Hoff: showed picture of himself many yrs ago w/ BANGS—so funny—and said my wife loved me even then
* Dan Von Hoff talk incredibly moving. He had LONG standing ovation at end!!