Precision Oncology: 2022 in Review

Since its establishment as a field at the turn of the 21st century, precision oncology and the use of genomically matched therapies have continued on a steady trajectory of growth. This trajectory has been documented by precision oncology knowledge bases that curate an expanding list of clinically actionable biomarkers and corresponding matched therapies. These knowledge bases thus serve as decision support tools for cancer care providers to interpret and act on genomic data from tumor sequencing.

Precision oncology knowledge bases use standardized methodology to annotate mutations found in cancer for their evidence-based oncogenic and therapeutic potential. In tandem with multiple organizations such as the Association for Molecular Pathology (AMP), the American Society of Clinical Oncology (ASCO), the European Society of Medical Oncology (ESMO), and the College of American Pathologists (CAP), academic knowledge bases such as CIViC, PODS, and OncoKB and commercial knowledge bases such as JAX-CKB address the significance of clinically actionable alterations by assigning levels of evidence to these biomarkers. Each level corresponds to the cumulative data that support whether an alteration is predictive of drug response in a specific cancer subtype.

OncoKB (1), an FDA-recognized precision oncology knowledge base, designates a tumor type–specific molecular event as level 1 when associated with a standard-of-care therapy that is FDA approved and as level 2 when listed by the National Comprehensive Cancer Network (NCCN) guidelines. Tumor type–specific alterations considered predictive of clinical benefit in phase III or compelling phase I/II trials receive a level 3 designation. Lastly, level 4 alterations are validated in experimental preclinical model systems as biomarkers predictive of response to targeted therapies.

Based on OncoKB as of October 28, 2022, six treatments were FDA approved in unique biomarker-selected indications and nine biomarker- and indication-specific treatments were listed in the NCCN guidelines in 2022. Additionally, compelling clinical evidence for nine precision oncology therapies, two of which received FDA breakthrough designation, led to their inclusion as level 3 investigational agents in OncoKB (Table 1). We review the major advances in precision oncology over the last year centering around three primary pillars: advances in (i) rational drug design, (ii) study design, and (iii) novel biomarkers.

KRAS, the most frequently mutated oncogene across all cancers, had long been considered an “undruggable” target due to the extremely high affinity of the enzyme for GTP/GDP and its absence of obvious molecular pockets for allosteric inhibition. This changed when seminal mechanistic studies spearheaded by Ostrem and colleagues (2) resulted in the development of direct KRAS^G12C inhibitors. These drugs, sotorasib and adagrasib, eventually received FDA accelerated approval and breakthrough designation, respectively, for patients with KRAS^G12C-mutant non–small cell lung cancer (NSCLC). In 2022, updated data from the phase II KRYSTAL1 trial were reported at the ASCO Gastrointestinal Cancers Symposium. A 41% objective response rate (ORR; n = 27) was observed with adagrasib monotherapy across KRAS^G12C-mutant gastrointestinal (GI) cancers. With this cumulative evidence of clinical activity of adagrasib in KRAS^G12C-mutant pancreatic ductal adenocarcinoma (PDAC) and biliary tract, gastroesophageal junction, and small bowel cancers, KRAS^G12C received OncoKB level 3 designation for these cancers (3).

Successful KRAS^G12C targeting led to the development of a pipeline of selective inhibitors targeting other RAS mutations, and these agents can be divided into two classes. The first class of agents targets specific RAS mutations. Novel KRAS G12D inhibitors, for example, such as RMC-9805 and MRTX1133, have been developed this year. RMC-9805 demonstrated proof-of-concept activity in vitro and in vivo, and MRTX1133 showed enhanced activity when used in combination with agents targeting either EGFR or the PI3K pathway (4). The second class of agents targets multiple activating RAS mutations. RMC-6236 is one such pan-RAS inhibitor that targets RAS in its GTP-bound form through the formation of ternary complexes. This RAS-ON inhibitor binds the intracellular chaperone cyclophilin-A that then binds GTP-bound RAS and interferes with the ability of RAS to activate downstream kinases. RMC-6236 demonstrated robust activity in multiple RAS-mutant xenograft models from lung, colorectal, and pancreas lineages. The first-in-human trial opened in 2022 and is currently enrolling patients whose cancers harbor various RAS mutations (e.g., KRAS G12A/D/R/S/V), making these KRAS-mutant alleles level 4 in OncoKB (5).

TP53 is the most frequently mutated gene in cancer, and mutant p53 had been considered undruggable until recently. A tumor suppressor and transcription factor, p53 responds to cellular stress by activating downstream antitumor responses such as DNA repair and apoptosis. Heterozygous TP53 mutations in the DNA-binding domain (DBD) create dominant-negative mutant proteins that are unable to function as transcription factors. Instead, mutant p53 binds and blocks the activity of wild-type p53. The rare TP53 Y220C mutation confers decreased DBD thermal stability, impairing the transcriptional activity of the mutant protein. This year, multiple groups including Guiley and Shokat (6) and the makers of PC14586 (7) presented data on the development of p53 Y220C–targeting drugs that leverage a druggable cavity in the DBD while restoring mutant DBD thermostability to wild-type levels. As presented at the 2022 ASCO Annual Meeting, patients with solid tumors harboring TP53 Y220C mutations were treated on a basket trial with the targeted agent PC14586, and confirmed partial responses were observed in six of 33 evaluable patients. Based on these data, TP53 Y220C is now considered a level 3 alteration by OncoKB.

We continue to see generational improvements in the quality of small-molecule therapies available to patients with oncogene-driven cancers. Next-generation tyrosine kinase inhibitors (TKI) have been designed to increase the selective activity against specific mutant alleles and on-target resistance alterations, address central nervous system (CNS) penetration, and increase tolerability by avoiding off-target adverse effects.

Over a decade of prior drug development efforts gave rise to multiple PI3K inhibitors. Despite promising preclinical activity, these agents failed in the clinic due to unacceptable toxicities including hyperglycemia attributed to the physiologic glucose regulatory activity of PI3K. Although the prior approval of the alpha isoform–selective PI3K inhibitor alpelisib represented a major advance in precision oncology, newer mutation-specific inhibitors (e.g., those that target PIK3CA H1047X) have also been developed. These inhibitors include RLY-2608, ST-478, and LOXO-783, which are currently being tested in phase I/II trials (8).

The sequential use of early-generation targeted inhibitors in molecularly defined cancers often results in the emergence of resistance mutations. Thus, beyond de novo mutation coverage, the number of drugs with resistance mutation coverage is increasing. Futibatinib, for example, is an irreversible FGFR1–4 inhibitor with activity against the FGFR2 V565I/L gatekeeper mutants (9). In October 2022, futibatinib received accelerated approval from the FDA for the treatment of FGFR2 fusion–positive intrahepatic cholangiocarcinoma based on an ORR of 42% and a median duration of response of 9.7 months. Additionally, pemigatinib, an FGFR1–3 inhibitor with gatekeeper mutation coverage, was approved in August this year for the treatment of FGFR1 fusion–positive myeloid/lymphoid neoplasms, making FGFR1 fusions OncoKB level 1.

Additional examples of resistance mutation–targeted agents include the ROS1/TRK fusion–targeted repotrectinib, the ALK inhibitors TPX-0131 and NVL-655, and the RET inhibitors LOXO-260 and HM06. Reprotrectinib received three FDA breakthrough therapy designations for the treatment of patients with ROS1 fusion–positive NSCLC or NTRK1/2/3 fusion–positive cancers. This drug is a small macrocycle that targets on-target, solvent front (e.g., ROS1 G2032R, NTRK3 G623R) and gatekeeper mutations that emerge with early-generation ROS1 or TRK TKIs. Intracranial and extracranial activity has been observed in patients with TKI-naïve or TKI-pretreated ROS1/NTRK fusion–positive cancers.

The next-generation ALK inhibitors TPX-0131 and NVL-655 address putative double ALK-resistance mutations in ALK fusion–positive NSCLC that have been sequentially treated with the early-generation ALK inhibitors crizotinib (incidentally also approved for the treatment of ALK fusion inflammatory myofibroblastic tumors this year), alectinib, brigatinib, and lorlatinib. Lastly, the selective next-generation RET inhibitors LOXO-260 and HM06, designed with resistance mutation coverage against solvent front mutations such as RET G810S, are currently being explored in trials for patients with RET-dependent cancers.

TKIs with the potential for improved therapeutic windows have also emerged this year.In January 2022, the EGFR TKI CLN-081 received breakthrough therapy designation for EGFR exon 20–mutant NSCLC. Compared with other agents like mobocertinib, CLN-081 has higher selectivity for mutant EGFR versus HER2. Additionally, the highly selective ROS1 inhibitor NVL-520 is designed to improve resistance mutation coverage and CNS activity as well as avoid TRK inhibition to ameliorate unfavorable TRK inhibition–mediated neurologic toxicities.

Although small-molecule inhibitors have been transformative for advanced/metastatic cancers, these drugs have not always hit the mark for every oncogene-driven cancer. In these instances, pivoting to antibody–drug conjugates (ADC) has represented a potentially more efficacious treatment strategy. This was resoundingly illustrated in the case of HER2-mutant or -amplified tumors. Trastuzumab deruxtecan, an ADC composed of trastuzumab and a topoisomerase I inhibitor payload, received accelerated approval in August this year for HER2-mutant NSCLC. This represented the first regulatory approval of any targeted therapy for this genomic subset of NSCLC (10). Approval was based on an ORR of 55%, a median progression-free survival (PFS) of 8.2 months and a median overall survival (OS) of 17.8 months, an improvement from the relatively poor activity of many TKIs in the same molecularly enriched population.

Beyond ADCs, naked antibodies have demonstrated utility in fusion–positive cancers in which the oncoprotein localizes to the tumor cell surface. Fusions involving NRG1, a ligand that binds HER3 to promote HER3/HER2 dimerization and consequent downstream AKT activation, are found uncommonly across virtually all tumor types (∼0.6% in the AACR Project GENIE patient cohort). Zenocutuzumab, an anti-HER3/HER2 bispecific antibody, demonstrated activity across nine NRG1 fusion–positive tumor histologies, including the typically treatment-refractory PDAC (ORR 39%), mucinous lung adenocarcinomas (ORR 35%), breast cancer, and cholangiocarcinoma in an update at the ASCO 2022 Annual Meeting (11). At the same meeting, seribantumab, an anti-HER3 IgG2 monoclonal antibody, was reported to have an ORR of 30% and a disease control rate of 90% in 11 patients with NRG1 fusion–positive NSCLC and one with an NRG1 fusion–positive PDAC (12).

The number of targeted therapy programs that utilize a basket trial strategy continues to increase. The growing popularity of these designs has been accompanied by the development of novel statistical methodology enabling an efficient and robust analysis of efficacy data even with small patient numbers and rare alterations or histologies (13). Basket trials may demonstrate that response to targeted therapy is conditioned by lineage (14). Conversely, activity may be tumor agnostic, as is the case for NTRK fusions and TRK inhibitor therapy. A total of five genomic biomarkers [microsatellite instability–high (MSI-H), 2017; NTRK fusions, 2018; tumor mutational burden–high (TMB-H), 2020; BRAF V600E, 2022; and RET fusions, 2022] are now associated with an FDA-approved tumor-agnostic therapy, and all of them are OncoKB level 1 biomarkers.

For BRAF V600E, the June 2022 accelerated approval of combination dabrafenib and trametinib in all solid tumors stands on the accumulated FDA approvals of combination dabrafenib with trametinib in BRAF V600E–mutant melanoma, NSCLC, Erdheim-Chester disease, and anaplastic thyroid cancer. These historic approvals added to the successes of three separate trials, the phase II ROAR trial in BRAF V600E–mutant biliary tract cancers, the multitumor lineage umbrella NCI-MATCH trial, and the phase I/II CTMT212 × 2101 trial in BRAF V600E–mutant Langerhans cell histiocytosis. Across the ROAR and NCI-MATCH trials, the ORR was 46% in biliary tract cancers, 33% in high-grade gliomas, and 50% in low-grade gliomas. Although both TRK fusion– and BRAF V600E–driven solid tumors feature single-driver oncogene addiction, the underlying biology of these molecularly distinct neoplasms is different. TRK-driven cancers have simple genomes lacking concurrent drivers, which may explain the dramatic and tumor-agnostic responses observed with TRK inhibitors. Conversely, BRAF V600E–driven malignancies harbor lineage-specific characteristics, including differing comutation patterns that may modulate the efficacy of dual RAF/MEK inhibition. Nonetheless, 7 years after the first basket trial of vemurafenib for any BRAF V600E–mutant tumor (15), BRAF V600E is the second molecular tumor-agnostic biomarker (as opposed to genomic features such as MSI or TMB status) to be associated with an FDA-approved therapeutic.

Similarly, the RET inhibitor selpercatinib was initially developed with a focus on cancers in which RET is most commonly altered, showing an ORR of 64% in RET fusion–positive NSCLC (16) and 69% in previously treated RET-mutant medullary thyroid cancers (MTC; ref. 17). This year, selpercatinib showed an ORR of ∼44% across multiple RET fusion–positive non-lung or thyroid solid tumors, including pancreas, colon, breast, and sarcoma. Although the ORR in these solid tumors was lower than that observed in NSCLC or MTC, possibly reflecting tumor heterogeneity and comutation patterns, over half of the patients enrolled on the study had GI malignancies, a group of cancers in which single-agent targeted therapies have rarely been successful. In September 2022, selpercatinib received tumor-agnostic approval for RET fusion–positive cancers (18).

The foundation of precision oncology stands on biomarker discovery efforts guided in part by large genomic studies such as those conducted by The Cancer Genome Atlas or the International Cancer Genome Consortium. Although these consortia-based studies and others have led to the emergence of novel treatment paradigms in molecularly selected patient cohorts, they have relied on patient sample datasets in which Black, Latinx, Native American, and Asian populations are vastly underrepresented despite these populations often showing an increased incidence or mortality in many cancer subtypes. Moreover, studies in pancreatic cancer this year reported that Black patients, who have higher incidence of and mortality rates due to pancreatic cancer compared to white patients, and those on Medicaid were underrepresented or disproportionately found to be ineligible for enrollment onto clinical trials (19, 20). As we evolve novel clinical trial designs and drug development efforts, considering the diversity of our patient population and underscoring inclusivity are paramount.

Although the vast majority of approved or guideline-listed therapies have been associated with a genomic biomarker, next-generation sequencing has its limitations as exemplified by the discovery of the role of embryonic mosaicism in seemingly sporadic cancers (21) and by the identification of silent mutations conferring oncogenicity in so-called single-driver cancers (22).

The newest developments in clinical oncology portend a transition to the increasing adoption of predictive protein-based biomarkers in addition to genomic biomarkers. The activity of trastuzumab deruxtecan in HER2-low breast cancer as reported in the DESTINY-04 trial exemplifies this trend. In this trial, HER2-low breast cancer was defined as a score of 1+ on immunohistochemistry (IHC) analysis or as an IHC score of 2+ but with negative fluorescence in situ hybridization (FISH) results. Trastuzumab deruxtecan showed a significantly longer median PFS (9.9 vs. 5.1 months) and median OS (23.4 vs. 16.8 months) versus physician's choice chemotherapy (18). The HER2-low group likely comprises more than half of all breast cancers. As such, a significant unmet need is addressed by this therapy (23).

These data underscore the critical task of accurately measuring HER2 protein expression. A survey showed that pathologists do not always agree in differentiating HER2-low from HER2-negative breast cancers using IHC. Determining the cutoff that distinguishes HER2-low versus HER2-negative becomes paramount and separates those who may be eligible or benefit from trastuzumab deruxtecan.

Further, the mechanism of action for trastuzumab deruxtecan is not entirely understood. Recent studies suggest that increased HER2 expression does not necessitate increased drug activity. Indeed, emerging modeling data have demonstrated that HER2 internalization and endocytosis rather than the degree of HER2 overexpression in tumors with normal or increased HER2 copy number as in HER2-low breast cancer or HER2-mutant NSCLC is a key mechanism of anti-HER2 ADC efficacy (24). Preclinical studies to assess the rates of receptor internalization or endocytosis in tumor tissue samples may become necessary as the use of ADCs is further refined.

The use of protein expression as a biomarker of therapeutic activity is not limited to HER2. Telisotuzumab, an anti-MET ADC with a microtubule inhibitor payload, received FDA breakthrough designation in January 2022 for EGFR wild-type nonsquamous NSCLC with high MET overexpression (25). The drug demonstrated differential activity based on the level of MET expression (ORR 52% in the MET-high group; ORR 24% in the MET-intermediate group), underscoring the importance of determining optimal agent- and lineage-specific protein expression cutoff levels.

Lastly, in addition to protein-based biomarkers, complementary detection approaches such as mass spectrometry and blood-based proteomics are now emerging. Multiplexing proteomic markers can generate a so-called proteomic signature that has the potential to dramatically increase the diagnostic and therapeutic utility of solid tumor protein assays.

Efforts to apply precision medicine matching strategies to immunomodulatory therapies continue to grow. Novel bispecific agents are likely to work best when patients are matched to these drugs based on both tumor- and host-specific factors. Tebentafusp-tebn, an affinity-enhanced T-cell receptor fused to an anti-CD3 effector (also called an ImmTAC, or immune-mobilizing monoclonal T-cell receptor against cancer), is engineered to redirect T cells toward uveal melanoma cells that express the tumor-associated antigen gp100 even at very low cell-surface epitope densities. Beyond harboring the cancer-specific antigen, patients must also carry a specific HLA genotype, as gp100 presentation is restricted to the HLA-A*02:01 serotype.

The IMCgp100-202 trial randomized patients to tebentafusp or investigator's choice of pembrolizumab, ipilimumab, or dacarbazine. As proof of concept of the activity of this agent, the primary outcome of median OS was significantly higher in patients who received tebentafusp compared with those who received investigator's choice of standard-of-care therapy (21.7 months vs. 16 months; HR = 0.51, P < 0.0001; ref. 26). In January 2022, the FDA approved the use of tebentafusp for HLA-A*02:01–positive patients with metastatic or unresectable uveal melanomas. Similar agents that require combined HLA genotyping and the presence of the targeted cancer antigen are being tested in ongoing trials. For example, data from the IMC-F106C trial of a PRAME cancer antigen–targeting bispecific for HLA-A*02:01–positive patients with advanced cancers were presented at the ESMO Annual Meeting this year (27).

Although tebentafusp was approved in a manner requiring only demonstration of the correct HLA genotype in patients with a single cancer type, it is not unreasonable to presume that future indications may require demonstration of a “double hit.” Specifically, a label could potentially require both patient HLA and tumor-associated antigen positivity to match patients to therapy, and the diagnostic implications of such a label would be profound.

Thus far, 2022 has been a landmark year for advances in precision medicine. Several targeted therapies were approved or received breakthrough designation, including two new tumor-agnostic approvals for BRAF V600E and RET fusions. Rational drug design continues to evolve. Many small-molecule classes have undergone iterative changes that have produced next-generation agents with increasing mutation selectivity and the potential for a widened therapeutic index. Protein-based biomarkers arguably represent a space with the highest potential for growth, particularly in our approach to diagnostics. Many protein-targeting ADCs, bispecific naked antibodies, and HLA-matched bispecific immunotherapies have received regulatory approval or demonstrated activity across various cancers this year. This growing number of active large molecules represents an encouraging movement beyond TKIs in the small-molecule saturated field of cancer-directed targeted therapy.

E. Rosen reports an NIH P30 CA008748 grant during the conduct of the study. A. Drilon reports personal fees from Ignyta, Genentech, Roche, MORE Health, AXIS, Loxo, Eli Lilly and Company, Bayer, AbbVie, EPG Health, Takeda, Ariad, Millenium, 14ner, Elevation Oncology, Harborside Nexus, TP Therapeutics, ArcherDX, Liberum, AstraZeneca, Monopteros, RV More, Pfizer, Novartis, Ology, Blueprint Medicines, EMD Serono, Amgen, Helsinn, Medendi, TouchIME, BeiGene, Repare RX, Janssen, BerGenBio, Nuvalent, Entos, Hengrui Therapeutics, Merus, Treeline Bio, Exelixis, Chugai Pharmaceutical, Prelude, Tyra Biosciences, Remedica Ltd, Applied Pharmaceutical Science, Inc., Verastem, mBrace, Treeline, MonteRosa, and AXIS during the conduct of the study; other support from Pfizer, Exelixis, GSK, Teva, Taiho, and PharmaMar outside the submitted work; a patent for osimertinib-selpercatinib pending; an NIH P30 CA008748 grant; and royalties from Wolters Kluwer, other (food/beverage) from Merck, Puma, Merus, and Boehringer Ingelheim, and CME honoraria from Medscape, OncLive, PeerVoice, Physicians’ Education Resources, Targeted Oncology, Research to Practice, Axis, Peerview Institute, Paradigm Medical Communications, WebMD, MJH Life Sciences, AXIS, EPG Health, JNCC/Harborside, and I3 Health. D. Chakravarty is the Lead Scientist of OncoKB, a precision oncology knowledge base that provides licenses to commercial companies that use the OncoKB database for research or hospitals that provide OncoKB data in clinical sequencing reports.

We thank Dr. Sarah P. Suehnholz for her critical review of Table 1 of the manuscript.