Patient and Treatment Monitoring
Why do we monitor patients during treatment?
In cancer care, monitoring those receiving treatment is done to ensure they're doing well, the treatment is effective, and there are no worsening health issues. This can include regular check-ups with the treating physician, diagnostic exams like blood work or imaging to monitor specific signs, standardized guidelines to ensure that monitoring is consistent and comparable across doctors and centers, etc. Patient monitoring is the only way we can know if a treatment is working or not.
How do we know if the treatment is not working?
When treatment isn't effective, we call this “progressive disease” – commonly shorten by PD. Detecting PD isn't based on just one test. Doctors might do a biopsy: taking a tiny piece of the tumor to see if cancer cells have increased, indicating “pathological progression”. Or they could use radiological scans to show if the tumor got bigger, known as “radiological progression”. Or they could see worsening of symptoms, termed “clinical progression”. Additionally, progression can be identified through specific response criteria, for example RECIST criteria (used often for patients in clinical trials) defines it as a 20% increase in tumor size. There is no unique definition of cancer progression. Each method used has its own pros and cons. For example, observing cancer cells under the microscope cannot tell us if and how these cells have spread across the body. Through radiological scans, on the other hand we can see the whole body, but modern machines have resolution of 1/10 millimeter at most, and do now allow to see left over cancer cells, which could cause a the cancer to return later. Subjective interpretation of medical exams based on the different experinece and perception of doctors can also result in uncertain outcomes, with the same patient being responding and progressing at the same time, depending on the doctor performing the exam. Patient monitoring remains challanging.
How can we improve patient monitoring?
Relying on just one method in clinical trials might not always give the full story, especially if that method is subject to variability, possibly needing more doctors to consult on a patient-by-patient level, or more clinical trials to run on a treatment-by-treatment level. Improving the accuracy and reproducibility of each existing method, and understanding how to use them together is the solution. Our research group aims to do achieve that goal, by (1) studying the limitations of current methods, (2) improving the accuracy and reproducibility of diagnostic tests, and (3) understanding how these can be used together, in a synergistic manner.
What have we discovered so far?
By examining current criteria for evaluating treatment, we've gained insights into techniques that enhance the reliability of measurements. Our use of computer simulation models revealed that even minor variations in how radiologists interpret treatment responses can significantly affect patient outcomes. Although radiologists generally assess responses consistently, differences in their experience levels and subjective interpretations can lead to notable discrepancies in evaluations. Better measurment of the tumor burden, and better understanding the changes happening during treatment in the body of the patient are evident priorities. We've used computer algorithms and artificial intelligence to address these inconsistencies, achieving more reliable, accurate, and reproducible results. These algorithms help us better understand tumor growth, how the body reacts to treatment, and the effects on patient survival. However, these tools are not yet perfect, and there is still much research needed to improve them.
