In June of 2017 just at the start of the annual American Society of Clinical Oncology (ASCO) meeting in Chicago, Illinois, there were at least 7 new FDA approvals for immuno-oncology agents targeting PD-L1 in cancer. At that time (2017), there were 2030 potential agents targeting 265 different targets across cancer including the modalities of t-cell targeted and other immunomodulators, cell therapy, cancer vaccines, oncolytic viruses, and CD3-targeted bispecific antibodies. Just three years later (2020), prior to the COVID-19 pandemic, this landscape had increased to 4720 potential agents targeting 504 targets across the same spectrum. That represents a 233% growth in these agents. Although only a fraction of these is “approved” (i.e., FDA approved and in use in patients clinically), many these agents are in clinical trials that require patient recruitment using pathology and other testing data. What does this mean for pathologists and laboratory professionals? Depending on the tumor being targeted and the target, there may or may not be a specific laboratory test that needs to be performed which may be routine, like histology parameters or immunohistochemistry, or may require advanced methods, like unique antibodies/clones, specific quantification methods, or molecular testing. The range of testing is not even unique to a specific therapy—for example, pembrolizumab uses staining for PD-L1, MSI, or no testing at all depending on tumor type. For the sub-specialized pathologist that focuses on one or two organs only, mastering the rapid pace and required diagnostic-therapeutic pairings is still a challenge. Imagine what it is like to be a general surgical pathologist in a community setting serving a community cancer center. Moreover, the diagnosis of a specific tumor is often completely disconnected for any biomarkers that may be indicated at the time of collection or several months later depending on therapeutic outcomes. This poses a range of problems in logistics and processing that are still being worked out at the individual system level. Still, the plethora of new treatments for cancer patients is very exciting.
In 2017, the largest group of targets (which was heterogenous) were tumor associated antigens (TAA) which are molecules that are not normally found in the human body produced by tumor cells as the result of changes to cellular processes. Whether it is hybrid proteins, glycosylation, or phosphorylation products, etc., these unique antigens held amazing promise as something we could target and destroy without fear of hurting normal human cells. However, the bulk of these approaches were for tumor vaccines (>90%) in 2017, dropping to 58% in 2020 (and from a total of 265 to only 198). To date, however, only a handful of cancer vaccines have been fully approved including sipuluecel-T for metastatic prostate and T-VEC for advance melanoma. This example category creates a complex set of challenges for pathologists and laboratory professionals. What data is needed about a patient or their tumor before a vaccine can be used? Does it require special studies that are not easily available or are costly? After vaccination, what follow-up tissue or blood studies are needed to follow the patient? Who dictates which tests are required before treatment: industry or medicine? But the more important challenge is: When do we, as the laboratory, pull the trigger to develop and disseminate such information and on-board new tests? Certainly, we are not going to look at Phase I trials and start taking about needs for future diagnostics. But by Phase III (where there is still a high dropout rate before full FDA approval) the number of potential agents and tests may still be daunting. If we wait until approval, now we are behind because our clinical colleagues will start immediately wanting to use the therapy. Tumor vaccines are an interesting category because we assume, for the most part, that there is likely only a diagnostic role needed. But then consider targets like CD-19, PD-L1, PD-1, CD3, Her2, CTLA-4, CD20, MUC1, CD22 and so on which are very familiar to our laboratory family because we often have already a test for these markers.
But is it the correct clone?
Do we have to score or interpret it differently?
When the agent is for cell therapy (the largest growth area of therapy development with 294% growth alone), what role does the transfusion medicine team play in administering or monitoring the patient?
As with the prior example, at what point do we, as a specialty of diagnosticians, dig into the forthcoming clinical trial results to plan? If our colleagues are in academic centers and are part of the clinical trials, they often are aware of and are administering the very tests that determine trial entrance. But if one reads just a few clinical trials of these agents, you may find that the inclusion criteria require a large battery of tests; however, on the other end when it is clinical ready for prime time, only one biomarker may be needed. Such a clustered landscape of information poses frustrating challenges for the clinical team and laboratory team in trying to find the way forward to get patients the life-saving therapies that are quickly arriving.
There is no question that the collision of targeted therapeutics and evolving diagnostics (i.e., precision cancer medicine) has demonstrated phenomenal growth with ever increasing benefits for patients. Affordability and access to these therapeutics aside*, studies continue to be completed and published including combinations therapies and hybrid therapies which show incredible promise. At ASCO 2022, the results of the DESTINY-Breast04 Phase III trial showed that trastuzumab deruxtecan (HER2-directed antibody and topoisomerase inhibitor conjugate) show a 49% reduction in the risk of disease progression or death versus physician’s choice of chemotherapy for patients with HER2-low metastatic breast cancer. That finding should be read a few times to make sure that the impact of this statement is very clear for pathologists and the laboratory. Previously, how we report HER2 (0, 1+, 2+, 3+) was complicated and often required FISH for questionable cases to look directly for HER2 amplification. This new category of patients requires reporting accurately 1+ or 2+ (FISH negative) disease, as it has incredible implications for patients. This news follows the recent new indications for CDK inhibitors in breast cancer related to Ki-67 mitotic score. Just when we thought breast cancer was straightforward, there is more to know and, more importantly, more time and tedium and standardization needed to report it for each patient. And, of course, early triple-negative breast cancer can also be treated with checkpoint inhibitors after PD-L1 testing is performed…but that’s literally old news as the data was release in 2020 at the start of the pandemic.
Outside of therapeutics, diagnostics are evolving quite rapidly with the COVID-19-induced ability to use digital pathology more readily creating a super-highway for artificial intelligence products to be validated for clinical use. PaigeAI has two such products (one for prostate and the second for breast lymph node evaluation released March of 2022) and many others are sure to follow. In parallel, screening, imaging, and surgery have also had advancements that continue to improve patient care and outcomes. So, it seems that everything feels new in cancer but is that the case?
The bulk of tumors diagnosed in the US (and elsewhere) are done with simply H&E staining (up to 75%) with another 20% being further confirmed by a few IHC tests (bringing the total up to 95%). This is not new and, most importantly, is the standard of care for the time being that we use to classify tumors. That classification has dictated, to some degree, the correct NCCN or other cancer protocol that oncologists used to treat patients. At some point, however, sufficient data on the bulk of all tumor types will likely point precision medicine treatments at all cancers. At that point, will a tissue biopsy be necessary with full histology or will a fine needle aspiration with molecular testing dictate the care? The credible assumption is that standard histology and IHC will remain in practice for the foreseeable future because so much billing, accreditation, and compliance is tied closely to them. But we CAN envision a “histology-free” oncopathology approach that matches patients to treatments with a panel of biomarkers. Sounds amazing but also stressful from the point of view of your typical anatomic pathologist.
*But the final thought on this, and perhaps the most important, is cost. Much like the domestic energy market is facing a dwindling pool of customers who agree to pay more and more for “traditional power” while their neighbors pump excessive kilowatts into the grid with their solar panels and windmills enjoying essentially “free power”, progress in cancer screening, detection, and treatment should be dwindling the pool of potential patients and increasing the costs to deliver care to the remainder. However, data and trends suggest that cancer is increasing globally. Why, if we are spending so much money and development on cancer care? Poverty and access. Cancer care is both expensive (in the US) and relatively expensive (in LMICs) with a focus on a small group of patients (0.55% of a population per year develop cancer). Projections of populations who need certain therapeutics are calculated using payer pools and markets that are existing and reliable. That does not include the bulk of LMICs. So, when we consider the cost of the PD-L1 checkpoint inhibitor class per year per patient is upwards of $125,000 USD, how can we even consider that an option for impoverished patients living off $1 USD per day? But if we don’t sort that out and treat these patients, we are assuming that persons who are impoverished are less valuable than persons who can afford expensive care. That evil logic, however, doesn’t hold true because even individuals in the US often become destitute or lose the bulk of their fiscal well-being when they must pay for cancer care—a situation that simply does not occur in countries with socialized medicine and/or universal healthcare.
Cancer care is rapidly evolving and the new tools and therapies available are incredible and miraculous for many patient types who would have faced a death sentence even 10 years ago. But with this amazing progress, we cannot ethically let people with limited resources succumb to these diseases over something so trivial as money. To do so poses harm and sets us up for failure as a species. It is for these reasons that ASCP engages in global health outreach. We are excited to have recently launched the Access To Oncology Medicines (ATOM) program with UICC and more than 2 dozen partners which will rapidly bring high-quality generic cancer therapeutics to low- and middle-income countries. In parallel with the St. Jude/WHO efforts on pediatric cancer globally, we will deliver quality cancer diagnosis and treatment to all patients everywhere.
If you want to learn more about PD-L1 testing and/or overcoming barriers to I-O in persons of color, new education from ASCP is available at no cost at https://www.ascp.org/content/learning/immuno-oncology/.
You can also check out our free educational resources on HER2-low breast cancer and Ki-67 testing in breast cancer at https://www.ascp.org/content/learning/breast-cancer.
Special thanks this month the Kellie Beumer (instructional design) and Melissa Kelly (monitoring and evaluation) from the ASCP medical education grants team for their thoughtful inputs into this piece.
-Dan Milner, MD, MSc, spent 10 years at Harvard where he taught pathology, microbiology, and infectious disease. He began working in Africa in 1997 as a medical student and has built an international reputation as an expert in cerebral malaria. In his current role as Chief Medical officer of ASCP, he leads all PEPFAR activities as well as the Partners for Cancer Diagnosis and Treatment in Africa Initiative.