The oncology field is the most concentrated area for global biopharmaceutical R&D investment. Cutting-edge technologies such as ADCs, bispecific antibodies, and CAR-T are advancing rapidly, and the market landscape long dominated by PD-1 monoclonal antibodies is being reshaped. The faster the technology iterates and the greater the R&D investment, the more critical the patent layout becomes—it determines whether R&D results can be transformed into sustainable commercial assets. There is a core premise in the patent layout in the oncology field: the “diagnosis and treatment methods for diseases” themselves cannot be granted patent rights in China. But this does not mean that innovation in the field of oncology cannot be protected by patents—the core strategy is: protect the “substance” rather than the “law,” protect the “intermediate result” rather than the “final diagnosis,” and protect the “technological implementation” rather than the “medical judgment.”

The following describes the types of innovations in the oncology field—drugs, diagnostics and prognosis, therapeutic devices and instruments—respectively.

---

1. Directions in tumor drug therapy

Oncology drug therapy is the direction with the most dense patent layout. Depending on the technological form of oncology drugs, the types of patents that can be laid out are as follows:

(1) Small molecule targeted drugs

The patents that can be laid out for small molecule targeted drugs in tumors are centered on the Markushton-formula compound patent. Taking EGFR-TKI as an example, the first-generation gefitinib and erlotinib patents protect the parent core structure of quinazoliline; The second-generation alfatinib introduces covalent binding groups (such as acrylamide) into its structure, achieving irreversible inhibition; The third-generation osimertinib introduces an indole ring and acrylamide warhead on the pyrimidine nucleus, achieving 32-fold selectivity for T790M mutations (mutant IC₅₀ 15 nM, wild-type 480 nM). Each generation of structural improvements corresponds to an independent compound patent.

Next are new forms of compounds—new salts, new crystal forms, solvent compounds. Lunan Pharmaceutical’s patent for osimertinib-malate eutectic is defined as “using Cu-Kα radiation to produce characteristic peaks at 3.68±0.2°, 6.87±0.2°, 10.55±0.2°, 14.86±0.2°, 21.37±0.2°, 25.57±0.2° in X-ray diffraction represented by Cu-Kα radiation. These patents, as peripheral protection for core compound patents, can extend the actual protection period.

The third is pharmaceutical compositions, which combine antitumor drugs with other active ingredients to form new compositions. The patent for composition is protected by drug combinations with specific ratios. Even if competitors use the same active ingredient, as long as the ratios or excipients differ, infringement is not guaranteed. Therefore, composition patents usually need to be used in conjunction with compound patents.

Fourth is pharmaceutical uses. When a marketed drug discovers a new tumor indication, it can use Swiss claims to arrange a use patent, written as “Use of XX compound in the preparation of drugs treating XX cancer.”

(2) Antibody drugs

The core of patents that tumor antibody drugs can be laid out is the antibody sequence itself. For newly discovered antigens, antibody claims can be limited by CDR sequences, variable region sequences, or by hybridomas that produce them. If antibodies have unique epitope-binding properties, epitope patents can serve as supplementary protection.

The structural design of bispecific antibodies—combining two antigen-binding domains and the design of linking peptides—is independent of the monoclonal antibody layout. Kangfang Bio’s Cardoniri (PD-1/CTLA-4 bispecific antibody) is a typical case in this direction. Its invention patent, “Bifunctional antibody against CTLA-4 and anti-PD-1, its pharmaceutical composition and its applications,” has consecutively won the China Patent Gold Award and the Guangdong Patent Gold Award, highlighting the strategic value of pharmaceutical composition and application patents beyond structural patents for bispecific antibody drugs.

(3) ADC drugs

ADCs are almost exclusively used in tumor treatment, and their patented layout naturally presents a multi-dimensional layered structure. The ADC consists of three parts: antibody, linker, and payload, with each layer independently eligible for patents. Antibody sequences, linker structures, and payload compounds can all be independently patented. When the linker is a completely new structure, the protective linker itself can extend to cover all ADCs containing it, allowing for freely selectable antibodies and payloads, with an extremely wide protection range.

The patent portfolio of detrastuzumab provides a typical example. The molecular patent for trastuzumab has expired, but the related patents for linkers and toxin drugs remain valid. The overall layout adopts a progressive application strategy of “structure-process-indications-combination drugs” to continuously extend patent life. This suggests the pace of ADC project layout: antibody patents may expire as early as the earliest, but linker-payload combination patents and complete molecular patents have longer protection periods, forming a cascading sequence.

(4) CAR-T cell therapy

Patents that can be laid out for tumor CAR-T therapy cover multiple structures: the CAR construct itself (antigen binding domain, hinge region, transmembrane domain, intracellular signaling domain), coding sequences, expression vectors, engineered cell products, therapeutic uses, and formulations.

The unique layout scenario of tumor CAR-T is target iteration—from CD19 to BCMA to solid tumor targets, each new target CAR design corresponds to an independent patent. The choice of co-stimulatory domain is also a key focus in the layout, as different choices directly affect the durability and safety of CAR-T. If the antigen-binding domain has a unique design, applying separately can cover a wider range of application scenarios.

(5) Nucleic acid drugs

Patents that can be laid out for tumor nucleic acid therapy include siRNA, mRNA, ASO sequences themselves, chemical modifications (such as thiophosphate skeletons, 2′-O-methyl modification), and delivery systems (such as formulations and preparation methods for lipid nanoparticles LNP). The unique layout scenario for tumor nucleic acid drugs is the tumor-targeted delivery system—how to specifically deliver nucleic acids to tumor tissues rather than normal tissues, which is the core competitive point of patent layout:

Sequence patents protect specific base arrangements

Chemical modifications that modify patent protection to improve stability and reduce immunogenicity

Delivery patents protect vector systems that deliver nucleic acids into tumor cells

(6) Drug Uses and Combination Therapy

The indication expansion chain for oncology drugs is especially long: from backline to first-line, from advanced to adjuvant, from monotherapy to combination therapy. Every expansion is a new use patent opportunity. In the field of oncology, therapeutic patents include delivery methods, indications, specific patient subgroups, and combination therapies. In the context of precision medicine, inventions for pharmaceutical use using Swiss-type disease classification and staging claims are an important form of protection.

---

2. Tumor Diagnosis and Prognosis Directions

The core of the patent layout in tumor diagnosis is to shift the protection from “diagnostic methods” to “testing products” and “data processing systems.”

(1) Biomarker and Test Kits

In tumor diagnosis, biomarkers can be used in patent directions including:

Test kits: Kits containing antibodies, probes, or primers with specific binding markers are the most direct layout form. The claims of the kit need to specify its composition and detection principle, rather than the method of “detecting XX markers to diagnose XX cancer.”

Use of markers in the preparation of detection reagents: The format should be “Use of XX markers in the preparation of reagents for detecting XX cancer.” The core of this approach is to demonstrate the degree of association between the marker and the disease—if the test result requires further analysis by a doctor to judge, it is considered an “intermediate result,” with more room for deployment.

Marker combination: a joint detection solution for multiple markers. The value of a combination of markers is higher than that of a single symbol—the more markers there are in the group, the harder it is for individual markers to be replaced. In early tumor screening, a combination strategy of “core markers + auxiliary markers” is commonly adopted, with core markers providing basic protection and auxiliary markers extending the protection period.

Technical implementation of detection methods: sample processing flow, signal amplification methods, quantitative analysis algorithms. These patents at the implementation level protect “how to obtain reliable detection signals from samples,” rather than “derive diagnostic conclusions from the signals.”

(2) Image processing system

AI patents for tumor imaging diagnosis can be deployed as computer processing systems for medical imaging, including: image preprocessing module, lesion segmentation module, feature extraction module, and efficacy evaluation module.

Image preprocessing module. Specific preprocessing methods for tumor images—such as noise suppression in CT images, bias correction in MRI images, and attenuation correction in PET images. The claims of the preprocessing module should specify the objects and objectives it handles, rather than the general concept of “image enhancement.”

lesion segmentation module. Automated segmentation methods for tumor lesions—such as liver tumor segmentation based on U-Net and glioma segmentation based on multimodal fusion. The claims of the split module should specify the specific steps and output forms of the splitting.

Feature extraction module. Extract quantitative features from tumor images—such as texture, shape, and function. These features are used for subsequent efficacy evaluation or prognosis prediction. The claims of the feature extraction module should specify the specific feature types and calculation methods to be extracted.

Efficacy evaluation module. The oncology-specific RECIST and iRECIST efficacy evaluation systems determine how the claims should automatically generate evaluation results according to these standards—such as automatic measurement of the longest diameter of the target lesion, automatic detection of new lesions, and automatic determination of disease progression. The protection of the efficacy evaluation module lies in “automatically generating evaluation results according to RECIST standards,” rather than “determining whether a tumor is present.”

(3) Liquid biopsy

The patented layout of liquid biopsy centers on the technical implementation of the detection method and the capture/enrichment device. Directly stating “diagnosing cancer by detecting ctDNA” will be rejected, but writing “a method for capturing circulating tumor cells from peripheral blood” or “a microfluidic chip for ctDNA detection” can be applied. Technical implementation of detection methods, structural design of capture devices, and data analysis algorithms are all directions that can be laid out.

---

3. Directions of tumor treatment equipment and instruments

This is the safest and most direct type of patent in the field of oncology. The structure, materials, and control systems of the device are considered “products” and are not subject to exclusion by diagnostic/treatment methods. The core of the patent layout for tumor treatment equipment and devices is to protect the device structure itself, control methods, and data processing logic.

(1) Tumor electric field therapy equipment

Patents for tumor electric field therapy equipment can be laid out, covering multiple aspects:

Structural design of electrode sheets: spatial arrangement of electrode arrays, adhesion between electrodes and skin, and selection of flexible electrode sheet materials are all directions of structural patents. The shape, size, and array density of the electrode sheet directly affect the uniformity of electric field distribution in the tumor region, and the specific design of these parameters can all be patented.

Electrode sheet detection and control methods: fault detection, temperature detection, impedance monitoring. These patented methods protect the real-time monitoring and feedback adjustment logic of the equipment during treatment.

Control methods for electric field parameters: dynamic adjustment strategies for frequency, intensity, and direction of alternating electric fields. For example, adjust the direction of the electric field according to the tumor’s location and size, adjust the frequency according to the stage of treatment, and adjust the intensity according to the patient’s response. These control methods are patent-protected technical implementations of “how to adjust treatment parameters based on real-time data.”

Overall architecture of the treatment system: electric field generator, electrode connection, and control system integration. System-level patents protect the coordination and control processes between components.

In addition, the application of alternating electric fields in the preparation of anti-tumor drugs can also be applied for use patents, extending the scope of protection from “devices” to “pharmaceutical uses.” The significance of this strategic layout is that even if competitors develop electric field therapy devices with different structures, as long as their use falls within the scope of patent protection for that use, infringement may still occur.

(2) Surgical robots

The patents that tumor surgical robots can be laid out are divided into two layers:

Product claims protect mechanical structures and controller design. Specifically, it includes: the joint structure of the robotic arm, the end effector design, and the hardware architecture of the controller. Product claims are not limited by treatment exclusions and represent the safest part of the surgical robot patent portfolio.

Method claims protect control logic, trajectory planning, and sensor data processing. Specifically, these include: path planning algorithms, force feedback control methods, and intraoperative imaging registration methods. The method claims need to focus on device control and data processing, rather than the operational steps of “inserting instruments into the human body.” If the claims involve therapeutic steps such as “inserting the puncture needle into tumor tissue,” they may be deemed non-therapeutic surgical methods lacking practicality.

The patent layout for surgical robots also needs to focus on human-machine interaction directions: control mapping methods for master-slave operations, methods for generating tactile feedback, and logic for switching operational permissions. These directions are especially important in tumor interventional surgery—doctors need to precisely control instruments under imaging guidance, and the quality of human-machine interaction directly affects surgical safety and efficiency.

(3) Ablation equipment

Patents that can be laid out for tumor ablation therapy include:

Structural improvements for ablation needles/catheters. The radiation antenna design of the microwave ablation needle, the refrigerant channel structure of the cryoablation catheter, and the array arrangement of the RF ablation electrodes. Structural patents protect the physical form of the device itself and are not limited by treatment exclusions.

Power control methods. Output power is adjusted in real time according to changes in tissue impedance, ablation range is controlled based on temperature feedback, and output is automatically stopped based on treatment time. These patented methods protect the control logic of “how to adjust energy output based on real-time monitoring data,” rather than the treatment step of “inserting the ablation needle into the tumor.”

Ablation range assessment method. Real-time evaluation of ablation areas based on imaging or impedance data, automatic delineation of ablation boundaries, and prediction models for ablation effects. These patents are especially important in tumor ablation—doctors need to confirm whether the ablation coverage fully covers the tumor, and the evaluation method patent protects the technical implementation of “how to determine whether ablation is sufficient from data.”

(4) Radiotherapy equipment

Patents for tumor radiotherapy equipment can be laid out, covering device structure, control systems, and dosing algorithms.

Equipment structure. Superconducting magnet design for proton therapy devices, beam flow transmission system for heavy ion accelerators, blade structure and drive mechanism for multi-leaf gratings. The core components of high-end radiotherapy equipment are the focus of the patent layout.

Control system. Real-time adjustment of beam energy, motion control of multi-leaf gratings, and positioning accuracy control of treatment beds. Control system patents protect the coordination logic and real-time response mechanisms among components.

Dosage algorithm. Accelerated implementation of Monte Carlo dose calculation methods, energy deposition models of proton beams in tissues, and optimization algorithms for dose distribution. The core of the radiotherapy planning system is how to generate plans—target plotting algorithms (automatically segmenting tumors and organs at risk), dose optimization algorithms (minimizing normal tissue dose while meeting target dose requirements), and adaptive planning adjustment strategies (adjusting plans based on anatomical changes at each treatment).

---

4. Conclusion

Innovation in the oncology field is advancing at an unprecedented pace, from molecular targeting to cell therapy, from liquid biopsy to intelligent radiotherapy, with technological iterations continuously reshaping the diagnostic and treatment paradigm. However, the value of innovation lies not only in “making it,” but also in “protecting it.” Under China’s current patent system, although “methods for diagnosis and treatment of diseases” cannot be granted, by focusing on “objects” rather than “laws,” and protecting “technological implementation” rather than “medical judgment,” a solid and flexible patent barrier can still be built.

---