This article is more than 5 years old. Information may no longer be current.
Progress has been slow to realize potential of molecular imaging in precision oncology
Click Here to Manage Email Alerts
I enjoyed reading the review by Mankoff and colleagues, and I must say that I could not agree more with the authors’ views.
However, it is important to provide some perspective on the rate of progress and what will be needed to reach some goals in the field.
Ten years ago, Gary Kelloff, MD — a colleague of mine at NCI — and I wrote a review titled “Imaging and Oncologic Drug Development” that discussed most of the same issues.
It is clear, as Mankoff and colleagues point out, there is opportunity for impact on clinical care, but the process has been very slow for various reasons they enumerate.
Areas of progress
Some progress clearly has been made with development of 18F-fluoroestradiol (FES)-PET to visualize ER expression in breast cancer that could possibly help guide therapy for patients with advanced disease. The example of FDG-PET use in gastrointestinal stromal tumors to monitor imatinib effects is well known; it is a poster child for molecular imaging.
The idea of molecular imaging becoming more clinically useful when tissue is difficult to get or inaccessible is an excellent point. At present, however, there is competition from liquid biopsy and, of course, circulating tumor cells can be used to evaluate expression of biomarkers like ER, or pharmacodynamic (PD) effects, such as pERK or Ki67. Liquid biopsies have their limitations — including that some patients with advanced disease have too few cells to analyze — and the lack of spatial correlation with biomarker expression is clearly a limitation.
Another point that imaging can help guide therapy is valid, although imaging may not detect microscopic invasive disease that ultimately may limit the efficacy of targeted radiotherapy.
Ten years of hurdles
In our Journal of Clinical Oncology review 10 years ago, Kelloff and I pointed out there were more than 200 molecular targets for which therapeutics were under development.
At the time in the care of patients with cancer, it was clear that we did not visualize molecular targets, we did not visualize the drugs at their targets within a patient’s tumor, and we did not typically visualize the effects of the drugs on the tumors as mediated through signaling or PD biomarkers.
Little progress has been made in 10 years with isotopic labeling of probes for clinical imaging, and even less with other technologies, such as optical imaging. The idea of phase 0 trials that would use tracer amounts of imaging was starting to get off the ground 10 years ago, and there was momentum for collaboration between NCI, CMS, FDA, academia and industry.
As Mankoff and colleagues point out, it has been difficult in oncology for clinicians to adopt molecular imaging to guide therapy. The reluctance is largely evidence based, and the bar will only get higher as other competing technologies — such as liquid biopsy — gain wider usage. It is clear ASCO guidelines have been encouraging limited usage of PET and this, as well as cost, probably affects some of the innovative directions.
The lack of availability of relevant molecular imaging probes was a reason for slow progress 10 years ago, as well as today. This is a huge issue that requires investment, incentives, infrastructure and collaboration among stakeholders.
It should be pointed out that molecular imaging can identify tumor heterogeneity like no other available technique, and it also has potential to predict unexpected toxicity.
I am reminded of the cardiac binding of trastuzumab (Herceptin, Genentech) that was shown by imaging in The New England Journal of Medicine years ago as potentially a way to anticipate toxicities. The point about the use of imaging to detect and characterize tumor heterogeneity was not emphasized enough in the JAMA Oncology article, but is very important to remember.
Another hurdle involves whether preclinical molecular imaging can be translated and whether it can be clinically useful.
EGFR or folate-receptor imaging has shown promise in preclinical models, but will it ever be helpful in the clinic? Is it clear that molecular imaging must measure molecular targets to impact precision oncology? I would argue the answer is no, as exemplified by efforts led by Martin G. Pomper, MD, PhD, at Johns Hopkins Medicine, with labeling classical chemotherapeutic drugs. Where the drugs go and whether they reach the tumor is clinically relevant irrespective of “molecular targets.”
Of course, no one disputes the need to image molecular targets of therapeutics and to demonstrate clinical utility. Metabolism imaging continues to gain sophistication (eg, glutamine imaging) and this holds some promise.
Need for integration
It is easy to agree with what needs to be done in terms of collaboration. There must be incentives for drug companies to label all new drugs and translate to clinical imaging. This could be like a companion diagnostic and could be used to monitor therapy.
NCI, FDA and CMS can impact to motivate these directions. The whole process must be shown to be clinically useful, and better than other emerging technologies, such as liquid biopsy. In other words, there needs to be integration.
The point here is that clinical trials need to look at both technologies simultaneously to ultimately impact the practice of precision oncology. Eventually, adoption of emerging technologies — such as molecular imaging — must be incorporated into clinical practice guidelines that would help with widespread usage by clinical oncologists.
Reference:
El-Deiry WS, et al. J Clin Oncol. 2006;24:3261-3273.
Slamon DJ, et al. N Engl J Med. 2001;doi:10.1056/NEJM200103153441101.
For more information:
Wafik S. El-Deiry, MD, PhD, FACP, is deputy director for translational research at Fox Chase Cancer Center. He also is a HemOnc Today Editorial Board member. He can be reached at wafik.eldeiry@gmail.com.
Disclosure: El-Deiry reports no relevant financial disclosures.