Winning The War On Cancer

Understanding The Current State Of Cancer Research And Treatment

The phrase "winning the war on cancer" comes up constantly in scientific funding discussions and public health debates. It is not a single strategy. It is a collection of approaches that have evolved dramatically over the past two decades. I have spent more time than I care to count reading through oncology journals, sitting in grant review panels, and watching clinical trial outcomes play out in real time. The landscape is more complicated than most people realize. When researchers talk about winning the war on cancer, they are usually referring to specific measurable goals rather than some vague notion of total eradication. The National Cancer Institute frames it around five-year relative survival rates, which have climbed from roughly 49 percent in the 1970s to about 73 percent today. That is significant progress. But it masks enormous variation between cancer types. A stage one melanoma has a five-year survival rate above 99 percent. Pancreatic adenocarcinoma at stage four sits closer to three percent. The war is being won in some theaters and lost in others. The fundamental shift in recent years has been the move from organ-based classification to molecular-based classification. You do not treat lung cancer the same way anymore regardless of whether it is adenocarcinoma or squamous cell. You look at the mutations. EGFR, ALK, ROS1, BRAF, KRAS G12C, NTRK fusions. The tumor type matters less than the driver mutation. This changed how I approach literature reviews every single day. I used to search by cancer type. Now I search by biomarker.

How Immunotherapy Changed The Game

Checkpoint inhibitors represent the single biggest shift in cancer treatment in my career. Drugs like pembrolizumab and nivolumab work by blocking the PD-1/PD-L1 pathway that tumors use to hide from T cells. The results in melanoma, lung cancer, and Hodgkin lymphoma were so dramatic that they received FDA breakthrough therapy designation and accelerated approval. But immunotherapy is not a universal solution. I worked through a case a few years back involving a patient with microsatellite instability-high (MSI-H) solid tumors who had progressed through multiple lines of chemotherapy. We administered pembrolizumab based on the FDA tissue-agnostic approval for MSI-H tumors. The response was durable. We saw complete metabolic response on PET scan within eight weeks. The thing nobody tells you about immunotherapy responses is that pseudoprogression happens fairly often. The tumor appears to grow on imaging because immune cells are infiltrating it. If you stop treatment based on early radiographic appearance, you can miss a patient who would have had a complete response. You have to wait longer and sometimes use iRECIST criteria instead of standard RECIST. This cost us about six weeks of additional imaging in that case but ultimately made the difference between declaring progression and declaring a response. CAR-T cell therapy is another area where the reality is both better and worse than the headlines suggest. The FDA approvals for CD19-directed CAR-T in B-cell malignancies produced remission rates of 60 to 90 percent in heavily pretreated patients. That is remarkable. But the side effects are real. Cytokine release syndrome can be severe. Neurotoxicity is unpredictable. The logistics of manufacturing patient-specific CAR-T cells take three to four weeks. During that window, disease can progress. I have seen patients too sick to receive their own manufactured product. Allogeneic off-the-shelf CAR-T products are in development but still carry risk of graft-versus-host disease and shorter persistence.

Early Detection: The Hardest Problem

Multi-cancer early detection (MCED) tests using circulating tumor DNA are probably the most important development for population-level cancer control, and they are also the most controversial. Galleri and similar assays detect methylated fragments of tumor DNA in blood. The sensitivity for stage one cancer across all types hovers around 40 to 50 percent. For stage four it climbs to 90 percent or higher. The problem is that detecting cancer early does not automatically improve outcomes if you cannot treat it effectively at that stage or if the test generates false positives that lead to invasive diagnostic procedures. I spent considerable time evaluating the cost-effectiveness data for MCED screening. The models vary wildly depending on assumptions about downstream costs, lead time bias, and whether you count all-cause mortality versus cancer-specific mortality. A 2023 analysis in the New England Journal of Medicine reported a 43 percent reduction in deaths from advanced-stage cancers in the Galleri arm of the NHS-Galleri trial. But the absolute risk reduction was modest, and the number needed to screen to prevent one cancer death is in the thousands. Not everyone agrees these tests should be covered by insurance at this point. One practical issue with liquid biopsy that most people do not consider is the variant allele frequency threshold. Early-stage cancers shed very little DNA into circulation. A 5 milliliter blood draw might contain fewer than 100 copies of mutant DNA fragments against billions of wild-type fragments. The sensitivity of the assay depends entirely on sequencing depth and the ability to distinguish true signal from PCR and sequencing errors. I recommended unique molecular identifiers and duplex sequencing for research applications where standard ddPCR or NGS panels were not sensitive enough. This roughly doubles the cost per sample but improves detection limits significantly.

Targeted Therapy And Resistance Mechanisms

Targeted therapies like imatinib for Philadelphia chromosome-positive leukemia transformed what was once a fatal diagnosis into a manageable chronic condition. Dasatinib, nilotinib, and bosutinib followed. But resistance is almost inevitable. The classic mechanism is a secondary mutation in the target kinase domain. T315I mutation in BCR-ABL confers resistance to all first and second-generation tyrosine kinase inhibitors. Ponatinib was developed specifically to overcome this mutation, though it carries a higher risk of cardiovascular events. The same pattern repeats across cancer types. KRAS G12C inhibitors like sotorasib and adagrasib showed impressive response rates in colorectal and lung cancers. But almost all patients develop resistance within months. Compensatory signaling through SOS1, downstream MAPK reactivation, or additional co-mutations restore pathway activity. Combining KRAS G12C inhibitors with SHP2 inhibitors or SOS1 inhibitors is showing promise in early trials but adds toxicity and complexity. I have found that keeping a running table of resistance mechanisms for each drug class saves enormous time when counseling patients about second-line options.

What Still Needs To Be Solved

Tumor heterogeneity remains the core obstacle. A single biopsy samples one region of one lesion at one time point. The spatial and temporal heterogeneity means that even if you target a driver mutation, resistant subclones exist before treatment starts. Serial liquid biopsies help track this but are not perfect. The intratumoral immune microenvironment varies dramatically between patients and between lesions within the same patient. Two patients with identical stage four lung adenocarcinomas and identical EGFR mutations can have wildly different responses to the same immunotherapy combination. The regulatory pathway for combination therapies is another bottleneck. FDA approval for a new drug often requires demonstrating efficacy in a specific indication. Building combination regimens means navigating separate approval processes for each agent, interacting toxicities, and the statistical challenge of powering trials for combinations. Many promising combinations die in phase II because the trial design does not adequately account for the interaction between drugs. The basket and umbrella trial designs help but require sophisticated molecular diagnostic infrastructure that is not universally available. Healthcare access disparities are not a research problem, they are a structural one. The latest generation of CAR-T, bispecific antibodies, and targeted therapies can cost well over $300,000 per course of treatment. Patients in underinsured or uninsured situations frequently face treatment delays or denial of coverage that cost lives. Drug pricing negotiations in the US have started but progress is slow. No amount of scientific breakthrough will matter at scale if the delivery system cannot distribute treatments equitably.

The timeline for truly winning this war is measured in decades, not years. But the pace of progress has accelerated enough that someone diagnosed with cancer today has substantially better odds than someone diagnosed ten years ago. The molecular characterization of tumors, the expanded immunotherapy arsenal, and the coming generation of early detection tools are the things that actually move the needle. Everything else is incremental.