Frequently asked questions

Yes. Many current studies evaluate CAR-T cell therapy and other advanced immunotherapies for myeloma.

No. Some trials focus on newly diagnosed or smoldering myeloma, while others are designed for relapsed or refractory cases.

Eligibility depends on disease stage, prior treatments, measurable disease status, lab values, organ function, and genetic markers. Each study defines specific inclusion and exclusion criteria.

Yes. Many current studies evaluate CAR-T cell therapy and other advanced immunotherapies for follicular lymphoma.

No. Some trials focus on early-stage or indolent lymphoma, while others are designed for relapsed or refractory disease.

Eligibility depends on stage, prior treatments, measurable disease, lab values, and overall health. Each study defines its own inclusion and exclusion criteria.

Trial drugs and study procedures are usually free. Standard care is billed normally.

Trials undergo strict scientific and ethical review. All treatments are monitored for safety, and patients receive close clinical follow-up.

No. Many studies target early-stage, postoperative, neoadjuvant, or prevention settings. Others focus on metastatic disease or specific subtypes like TNBC or HER2-positive cancer.

Eligibility depends on factors like cancer subtype, stage, prior treatments, biomarkers (HER2, ER/PR), age, organ function, and location. CancerBot helps match these factors automatically.

Yes, we will supply all the source code to such organizations. They can modify it as they choose and truly make it their own service. Based on our discussions we believe that most such organizations will eventually choose to have CancerBot host the service for them (with their branding). CancerBot will also enhance the software for the foundation’s needs. CancerBot will then charge either a hosting fee or broker patients.

Patients can sign up for free to CancerBot directly. CancerBot makes money by brokering patients to trial researchers from for-profit pharma.

There are some open source projects, notably TrialGPT and TrialMatchAI, that present “vignettes” for synthetic patients (short paragraphs describing the patient and a handful of symptoms) and then recommend a trial from a curated dataset of trials. It then evaluates whether the top ranked trial is accurately selected and evaluates the precision of the top n trials. This is not true precision matching based on patient labs, diagnostics and biomarkers.

The accuracy results are quite marginal (in the 60% range), and reflect what the accuracy of presenting some patient info and then asking the LLM whether the trial is relevant can achieve. This is not accurate enough to be a useful system for patients. CancerBot’s EXACT system is more involved than this to get around the limitations of this naive approach.

We use AI LLMs in a complex system called EXtracting Attributes from Clinical Trials (EXACT) described here. The result is extracting specific attribute criteria from trials with very high accuracy as described here.

It is a notoriously difficult problem. Other patient centered matching services ask a few basic questions (disease, stage, grade, previous therapy) and then get a lay consultant involved in the matching process. In my interactions with these consultants they knew less about my disease than I did. This is not precision matching. 

CancerBot was created because there are no patient-centered clinical trial matching services that offer precision matching based on patient labs, diagnostics and biomarkers. Other patient-centric matching services don’t even ask the questions about those attributes that would enable such matching.