The upcoming decade of precision medicine for cancer is moving from the translation of specific genetic findings into clinically relevant improvement to the qualitative analyses of the genomic and immune tumor microenvironment, for an integrated treatment strategy in both metastatic and early disease.
Moving Precision Oncology Forward
Over the last two decades, precision therapies have revolutionized the treatment of several cancers, including chronic myeloid leukemia, gastrointestinal stromal tumors, melanoma, and non–small cell lung cancer (NSCLC). This revolution has been facilitated by the rapid evolution of technologies to interrogate genomic changes in cancers directly or noninvasively, and by advances in drug development. For example, in NSCLC, there are seven genomic subtypes associated with one or more approved targeted therapies. This number is expected to grow over the next several years with exciting emerging data for patients with KRASG12C-mutant (1) and HER2-mutant NSCLC (2).
Although precision therapies have led to major clinical improvements in progression-free survival, and in some cases overall survival, in patients with advanced cancer, they are rarely if ever curative and are limited by the development of acquired drug resistance. The mechanistic understanding of drug resistance has led to the development of next-generation precision therapies that are being used either to overcome resistance or as initial treatment to forestall the emergence of resistance. However, a fundamental opportunity and challenge remains to change the treatment paradigm from a one drug/one genetic alteration approach to a combination treatment strategy. In diseases such as human immunodeficiency virus (HIV) infection and Mycobacterium tuberculosis infection, combination therapy is the mainstay of treatment, and current treatment approaches are associated with a high degree of clinical success. However, in both instances, treatment is targeted at foreign organisms. In malignancies such as testicular cancer and non-Hodgkin lymphoma (NHL), combination chemotherapy has led to increases in outcomes and cure rates. In the case of testicular cancer and NHL, treatment toxicities are mainly overlapping (bone marrow suppression) and can be transiently managed with supportive care measures.
In contrast, many precision therapies lack a therapeutic index. Many of the pathways important to survival and proliferation of cancers, including PI3K–AKT and MAPK pathways, are critical for many normal cellular functions. As such, efficient inhibition of these pathways required for clinical efficacy is rarely tolerated for extended periods of time. Thus, to effectively deploy combination precision therapies in the clinic, strategies to maximize therapeutic index need to be developed. Examples of such strategies include the development of inhibitors that are more potent against mutant forms of proteins than their wild-type counterparts. In addition, a large majority of therapeutic strategies to date utilize daily drug dosing with the hypothesis that continuous target inhibition is necessary for a clinical benefit. However, this may not be necessary, especially in a cancer characterized by an oncogenic alteration where transient but potent target inhibition can be effective. The development of alternative treatment strategies, including clinical trials evaluating intermittent combination strategies or alternating combination strategies, will be necessary to determine whether these approaches lead to improvements in clinical outcomes while minimizing toxicities.
Novel drug development approaches are also likely to increase over the next several years and lead to improvements in outcome with precision therapies or allow inhibition of previously “undruggable” targets. One example is antibody–drug conjugates (ADC), which are antibodies linked to a cytotoxic payload (3). The concept is to use the antibody to specifically deliver the payload to be released at the target site or once internalized into the target cells. This approach, in theory, allows specific delivery to tumors while minimizing exposure to normal tissues. The field is rapidly growing to develop linkers where nonspecific cleavage is minimized and strategies to maximize the drug-to-antibody ratio. Effective utilization of this approach requires the identification of an antigen that is highly expressed in tumors compared with normal cells coupled with a high-affinity antibody or antibodies that more effectively bind their target in tumors compared with normal tissues. Many of the initial regulatory approvals focused on hematologic malignancies, but more recent approvals including trastuzumab deruxtecan, enfortumab vedotin, and sacituzumab govitecan have all been in solid tumors. Intriguingly, resistance to one ADC does not imply resistance to another ADC against the exact same target, as demonstrated by the efficacy of trastuzumab deruxtecan in patients with HER2-positive breast cancer previously treated with trastuzumab emtansine (4). Much remains to be understood about the biology of ADCs, mechanism(s) of acquired resistance, and the determinants of sensitivity to ADCs with the same target. Furthermore, new approaches, including use of smaller targeting domains (for example peptides) instead of antibodies to allow greater tumor penetration, and conjugates other than cytotoxic chemotherapies, are likely to further broaden this treatment approach in the future.
Another exciting evolving area of drug development is identifying bifunctional molecules that connect a ligand that binds to a protein of interest linked to an E3 ubiquitin ligase–recruiting ligand. These molecules, often referred to as PROteolysis TArgeting Chimeras (PROTAC), can lead to selective degradation of a target of interest (5). As most precision therapies are enzymatic inhibitors, this approach can potentially overcome resistance mechanisms to precision therapies and also have a broader therapeutic potential, as target degradation can alter scaffolding functions of proteins, which are not typically affected by an enzymatic inhibitor. However, further work is needed to better understand the therapeutic potential of PROTACs. This strategy still requires the identification of a ligand that specifically binds the target of interest. To date, the most commonly used ligands are ATP-competitive inhibitors. In addition, the approach is not quite as simple as connecting a ligand-binding target with an E3 ligase. Understanding the biology and determinants behind which targets can be effectively degraded and in what therapeutic contexts degradation is preferred over enzymatic inhibition will be important before moving these agents into the clinic. In addition, whether PROTACs can also penetrate the central nervous system (CNS), a feature physicians and patients alike have come to expect from modern precision therapies, remains to be determined.
Another advance in precision therapies is the identification and development of allosteric inhibitors. These agents can bind nonoverlapping regions with ATP-competitive inhibitors and as such are not subject to the same resistance mechanisms. Furthermore, they can potentially cobind simultaneously with an ATP-competitive inhibitor, thus providing maximal target inhibition and decreasing the likelihood of developing an on-target mechanism of resistance. Clinical and preclinical examples to date include ABL and EGFR inhibitors (6). As allosteric sites are less likely to be conserved than ATP-binding sites among closely related kinases, this approach may also allow the development of highly specific agents.
In addition to the many advances in the treatment of cancer driven by the development of targeted agents, the recent development of immune-modulatory molecules has revolutionized the systemic treatment of cancer, including advanced or metastatic NSCLC, without actionable molecular alterations. Indeed, in this setting, until 2015, the first line of therapy consisted of platinum-doublet chemotherapy. Modest improvements in outcomes were made 15 years ago with the addition of the antiangiogenic agent bevacizumab and histology-driven pemetrexed for patients with nonsquamous NSCLC, but the potential for a durable benefit has been elusive.
In recent years, immunotherapy has become a standard component of first-line treatment for patients with cancers including advanced NSCLC, improving survival and offering hope for long-term disease control. However, as only a subset of patients derives clinical benefit from current immunotherapy strategies, it is critical to better understand the determinants driving response, resistance, and adverse effects. In this regard, patients with NSCLC should be ideally qualified for immunotherapy, based on dedicated specific predictive factors. Decisions regarding which immunotherapy to use or whether a combination approach is warranted should ideally be driven by a rational, mechanistic insight to maximize disease control, reduce side effects, and minimize costs. To achieve further progress, we must address the current limitations in achieving a durable benefit from immunotherapy for most patients. First, biomarkers must be refined to indicate how the treatment should be selected and managed for individual patients. Second, in patients benefitting from such a strategy, the emergence of resistance must be further and more deeply studied to improve our understanding in each cancer type, ideally before studying large-scale hypothetical or sometimes opportunistic combinations. Finally, moving forward, treatment approaches need to be optimized and tailored for individual patients, including innovative immunotherapy approaches, with a focus on extending the breadth, depth, and durability of benefit to more patients.
Currently, PD-L1 expression on tumor cells of patients with advanced NSCLC is the only predictive factor validated in prospective clinical trials. However, recently tumor mutational burden (TMB) has emerged as an exploratory biomarker, although a homogenous definition and rigorous methodologic assessment are still lacking.
The complexity of the tumor–immune microenvironment indicates that a single biomarker–based strategy cannot identify patients who should or should not receive immunotherapy treatment. Definition of predictive models/signatures considering the different components affecting tumor–host interactions will be required. In addition to TMB, identification of different specific genetic alterations, clonality of mutations, gene-expression signatures through RNA sequencing, neoantigen assessment, human leukocyte antigen (HLA) diversity, and other immune checkpoint molecules, as well as tumor microenvironment composition, including characterization of tumor-infiltrating lymphocytes (TIL), are under investigation as part of a multilevel biomarker definition. Such complex and composite biomarkers for response to immune checkpoint inhibitors (ICI) will have profound implications in the area of precision immuno-oncology (7), in particular to identify mechanisms of immune evasion and resistance.
Improving the understanding of immunotherapy resistance represents one of the significant hurdles to overcome. Immune escape mechanisms are observed from preinvasive lesions to established metastatic disease, combining defects in antigen presentation machinery, loss of heterozygosity at the HLA region, neoantigen silencing, activation of immune checkpoints, altered Th1/Th2 cytokine ratios, defects in IFNγ signaling, and immune microenvironment evolution.
Resistance to ICIs should be classified into the two broad categories of primary and acquired resistance, the latter referring to patients who presented an initial response to ICI therapy. Recently, different definitions have been proposed to better characterize such clinical heterogeneous scenarios (8). Unlike primary resistance, the rates of acquired resistance have not been routinely reported but might significantly vary across tumor types (9), revealing the existence of distinct determinants of initial response, long-term durability, and acquired resistance. Consequently, a definition of de novo versus acquired resistance should ideally be built for each cancer type, allowing for the conduct of relevant dedicated clinical trials.
Current efforts to broaden immunotherapy strategies have been focused on different combinations of existing treatments or developing novel agents that stimulate both innate and adaptive immune responses. There is a rapidly growing appreciation for the critical role of next-generation immune-modulating agents targeting CD27, CD40, TIM3, LAG3, TIGIT, GITR, ICOS, STING, and STAT3, which, combined with anti–PD-1/PD-L1, might generate greater antitumor activity than each respective monotherapy by increasing the systemic level of tumor-specific CD8+ T cells. Cytokine-directed therapies are also under evaluation, such as IL15, IL10, IL2, or type I IFN agonists, as well as TGFβ, IL1β, or IL6 antagonists. However, none of these strategies have reached a clinical application yet.
In addition, the emerging rationale for combining epigenetic drugs with immunotherapy to modulate the tumor microenvironment is being evaluated. Epigenetic regulation of T cells, antigen-presenting cells, and tumor antigen processing pathways is directly involved in immunotherapy responses and may be crucial targets in attempting to improve current options (10).
Chimeric antigen receptor (CAR) T-cell therapy has gathered significant excitement due to its success in hematologic malignancies. In contrast, the clinical experience with CAR T-cell therapy for solid tumors has been less encouraging, with only a few patients achieving responses due to limitations including the lack of targetable antigens, heterogeneous antigen expression, inadequate T-cell trafficking and survival, and the immunosuppressive tumor microenvironment. Much excitement in cancer immunotherapy is focused on TIL therapy, which has shown durable complete responses in a subset of patients with metastatic melanoma. Ex vivo expansion of TILs can release T cells from a suppressive microenvironment and reactivate them to target the tumor. By this method, billions of activated T cells can be produced and infused back into a patient. In contrast to CAR T, TIL product is composed of polyclonal cells capable of targeting multiple tumor antigens.
New immune modulations targeting natural killer (NK) cells and innate lymphoid actors are also being considered as a potential therapeutic strategy. NK cells can control tumor growth and metastatic spread, and based on their strong cytolytic activity against tumor cells, different approaches have been developed for harnessing their function, from the adoptive transfer of autologous NK cell–enriched cell populations to CAR-engineered NK cells (11).
Improving Therapeutic Approaches in Early-Stage Cancer
On the basis of the recent growing successes of molecularly targeted therapy and immunotherapy in cancers such as advanced or metastatic NSCLC, it is logical to apply these modalities in the neoadjuvant or adjuvant setting, where the effect of chemotherapy may marginally improve the survival of patients with resected lung cancer. The major proof-of-concept study for this idea is the ADAURA trial that confirmed the use of adjuvant osimertinib in resected EGFR mutation–positive lung cancer (12). In this randomized phase III study, a significant improvement in the primary endpoint of disease-free survival was reported in patients with stage II–IIIA disease, as well as an increase in the percentage of patients without CNS disease at 2 years. Although the overall survival data are currently immature, given the strongly positive results, we expect general adaptation of adjuvant osimertinib for EGFR mutation–positive early-stage lung cancer in the coming decade.
A future challenge is how we may implement adjuvant therapy for patients with uncommon driver oncogenes such as ALK or RET translocations. Incidence of these translocations is about 3% to 5% in patients with advanced NSCLC; thus, it will be a tremendous effort to confirm the role of targeted adjuvant therapy in the even smaller population with resectable disease. To date, there are two ongoing randomized phase III studies in patients with resected ALK-positive lung cancer. The Adjuvant Lung Cancer Enrichment Marker Identification and Sequencing Trial (ALCHEMIST) study randomizes patients with resected stage IB–IIIA ALK-positive lung cancer to either crizotinib or placebo, and the ALINA study plans to enroll patients for either 2 years of alectinib or four cycles of chemotherapy postsurgery.
The only randomized phase III study on RET-positive lung cancer is LIBRETTO-432, which compares selpercatinib with placebo in patients with stage IB–IIIA disease having completed the therapies with curative intent. Conceptually, these studies are promising; however, limited by the rarity of these mutations in early-stage disease, it will take years before there will be a definitive change in the treatment paradigm. However, we remain optimistic that the results will become available within this coming decade.
Another novel adjuvant approach is to monitor and treat minimal residual disease (MRD). One reason for the low reported efficacy of adjuvant chemotherapy is that the treatments are being offered in a nonselective manner to all patients, although some patients may not need it. Some patients are cured by the surgery alone, and only patients with detectable MRD may eventually recur. It is now feasible to identify patients with MRD by monitoring the plasma circulating tumor DNA (ctDNA) such that adjuvant therapy can be offered selectively (13). It is known that patients with residual plasma ctDNA postsurgery are associated with a higher risk of recurrence and shorter survival. The TRACERx study offers the bespoke approach to ctDNA detection. Abbosh and colleagues (14) used multiplex-PCR next-generation sequencing to track subclonal mutations in patients longitudinally and identify relapses. The median lead time of this approach to radiologic recurrence was 70 days; early intervention before radiologic recurrence may potentially increase the chance of long-term survival. In addition, the MERMAID study aims to screen plasma samples from more than 2,000 patients at week 3 to 4 postsurgery and identify patients with MRD according to their genomic definition. Eligible patients are randomized to receive four cycles of chemotherapy with durvalumab followed by ten cycles of durvalumab or chemotherapy plus placebo, followed by placebo, with the primary endpoint of disease-free survival.
Moving forward, for patients without a driver oncogene, the role of neoadjuvant and adjuvant immunotherapy needs to be intensively studied (15). The first report on neoadjuvant immunotherapy provides promising evidence of 45% major pathologic response (MPR) among 20 patients who received two cycles of nivolumab before curative surgery. Other single-arm studies, including LCMC3 on neoadjuvant atezolizumab and NEOSTAR on neoadjuvant nivolumab ± ipilimumab, also reported promising MPR at 18% and 23%, respectively. Different randomized neoadjuvant studies, such as CHECKMATE-816, KEYNOTE-671, AEGEAN, and IMpower-030, are ongoing. The design of these studies is similar, and most adopt the combination of chemotherapy and immunotherapy as an experimental arm. Only CHECKMATE-816 investigates the immunotherapy combination of nivolumab and ipilimumab, and this study, being more mature than others, may expect early data release in the near future. Similarly, a total of four randomized adjuvant immunotherapy studies are ongoing; namely, the PEARLS, BR-31, IMpower-010, and ANVIL studies investigate the role of adjuvant pembrolizumab, durvalumab, atezolizumab, and nivolumab, respectively. Each of these studies will enroll nearly a thousand patients with resected lung cancer, and we anticipate data maturation in a few years.
Other interesting anti-immune/inflammation drugs are also under investigation in the adjuvant setting, such as canakinumab, a selective IL1β inhibitor shown to reduce lung cancer incidence and mortality in a cardiology study. On the basis of the fact that IL1β may promote tumor invasion and metastasis, it is reasonable to hypothesize that its inhibition may reduce cancer recurrence after surgery, which is being assessed in the CANOPY-A study.
The significant improvement in molecular diagnostics and therapeutics has dramatically changed the natural history of cancer. Using precision therapeutics, including targeted therapies and immune checkpoint blockade, has changed our approach to the efficacy benefit as an endpoint for cancer treatments. Thanks to these precision oncology developments, we are able to consider the rate of long-term survival (2, 3, 4, or 5 years) as the most important practice-changing endpoint for all stakeholders, including patients, caregivers, clinicians, researchers, academics, industry, and regulators, worldwide. Converting undruggable targets to druggable targets and further characterizing the molecular subtype of each solid and hematologic cancer are the most important challenges to broaden the use of targeted therapies and reduce the use of old-fashioned chemotherapy agents, which are hindered by limited efficacy and significant impact on quality of life. In addition, for patients with advanced disease, improving our understanding of and molecular therapeutics for targeting drug resistance, occurring mainly due to clonal evolution and selective pressure, are the key points to improve the care and the cure of our patients, considering long-term efficacy as the new standard in daily clinical practice across different cancer types. To shift this knowledge from the metastatic to early setting, with the improvement of screening programs and increased understanding of MRD and ctDNA, will be key to definitively change the progression of malignant disease, and thus our ability to further help our patients.
S. Peters received education grants, provided consultation, attended advisory boards, and/or provided lectures for AbbVie, Amgen, AstraZeneca, Bayer, Biocartis, Bioinvent, Blueprint Medicines, Boehringer Ingelheim, Bristol-Myers Squibb, Clovis, Daiichi-Sankyo, Debiopharm, Eli Lilly, F. Hoffmann-La Roche, Foundation Medicine, Illumina, Janssen, Merck Sharp and Dohme, Merck Serono, Merrimack, Novartis, PharmaMar, Pfizer, Regeneron, Sanofi, Seattle Genetics, Takeda, and Vaccibody, from whom she has received honoraria (all fees to institution). T. Mok reports personal fees from Abbvie, Inc, ACEA Pharma, Alpha Biopharma Co. Ltd, Amgen, Amoy Diagnostics Co. Ltd., BeiGene, Berry Oncology, Boehringer Ingelheim, Blueprint Medicines Corporation, CStone Pharmaceuticals, Curio Science, Daiichi Sankyo, Eisai, Fishawack Facilitate Ltd., Gritstone Oncology Inc., Guardant Health, Hengrui Therapeutics, Ignyta Inc., IQVIA, Incyte Corporation, Inivata, InMed Medical Communication, Janssen, Lilly, Loxo-Oncology, Lunit Inc., MD Health Brazil, Medscape/WebMD, Mirati Therapeutics Inc., MoreHealth, OrigiMed, PeerVoice, Physicians' Education Resource, P. Permanyer SL, Puma Technology Inc, PrIME Oncology, Qiming Development HK Ltd., Research to Practice, Sanofi-Aventis R&D, Takeda, Touch Medical Media, Virtus Medical Group, Yuhan Corporation; grants and personal fees from AstraZeneca, Bristol-Myers Squibb, G1 Therapeutics, Merck Sharp & Dohme, Merck Serono, Novartis, Pfizer, Roche; other from geneDecode, AstraZeneca PLC, Hutchison Chi-Med, Sanomics Ltd., Aurora Tele-Oncology Ltd.; grants from Clovis Oncology, SFJ Pharmaceuticals, and grants from XCovery outside the submitted work. A. Passaro reports personal fees from AstraZeneca, Agilent/Dako, Boehringer Ingelheim, Bristol-Myers Squibb, Eli Lilly, Janssen, Merck Sharp & Dohme, Pfizer, and Roche Genentech outside the submitted work. P.A. Jänne reports grants and personal fees from AstraZeneca, Boehringer Ingelheim, Eli Lilly, Daiichi Sankyo, Takeda Oncology; personal fees from Pfizer, Roche/Genentech, Chugai, Ignyta, LOXO Oncology, SFJ Pharma, Voronoi, Biocartis, Novartis, Sanofi, Mirati Therapeutics, Transcenta, Silicon Therapeutics, Syndax, Nuvalent, Bayer, Esai; grants from Revolution Medicines, PUMA, and grants from Astellas during the conduct of the study; in addition, P.A. Jänne is an inventor on a DFCI-owned patent for EGFR mutations issued and licensed to Lab Corp. No other disclosures were reported.