Malignant pleural mesothelioma is a rare type of cancer. A growing number of mesotheliomas are being treated with novel immunotherapies. But clinicians still lack a meaningful way to measure mesothelioma treatment response.
An article in Seminars in Nuclear Medicine lists five ways clinicians can measure how well a treatment plan is working. And a new artificial intelligence tool can measure mesothelioma tumors without any human input. Artificial intelligence may provide the key to treatment response in the future.
Measuring Mesothelioma Treatment Response
In malignant pleural mesothelioma, complex tumors result in inconsistent clinical assessment. There are five common ways to measure if a treatment plan is working. Generally, a clinician will look to see if the size or weight of a tumor has changed after treatment.
The first is: Response Evaluation Criteria In Solid Tumors (RECIST). It has been the main choice for objective treatment response assessment in clinical trials. But, this response assessment was designed for tumors with a spherical shape, which is not mesothelioma. Malignant pleural mesothelioma has a different shape and growth pattern. This makes it difficult to use RECIST measurement for mesothelioma.
The second is: mRECIST (modified Response Evaluation Criteria In Solid Tumors). This option works better for malignant pleural mesothelioma. This measurement focuses more on the thickness of a tumor, rather than its shape or diameter.
The third introduces immunotherapy to the first and second choices: iRECIST/imRECIST/iRC. Immunotherapy treatments are common. But they lead to problems with the conventional treatment assessments. Immunotherapy causes the shape and size of a tumor to change during treatment. This third option adjusts the treatment assessment based on immunotherapy factors.
The fourth choice is: EORTC/PERCIST. That’s the European Organization for Research and Treatment of Cancer (EORTC). And the PET Response Criteria in Solid Tumors (PERCIST). This method uses imaging and scans to track if a treatment is working. PET/CT scans are most common. This method is a quicker assessment than the other options. Clinicians can get results sooner after starting therapy. This method has not yet been validated in a clinical trial.
The fifth choice is Volumetric Response Assessment. This method compares the weight and measurement of a tumor to assess treatment success. But, again, the growth pattern of malignant pleural mesothelioma can make this difficult.
Machine Learning to Support Measurement of Treatment Response
Some have looked at using machine learning to help with this problem. Machine learning has been used to look at mesothelioma type and genomic data. But no one has used machine learning with medical imaging to assess treatment response.
The German team is looking to apply new machine learning techniques to assess mesothelioma treatment response.
The team thinks that artificial intelligence could enable better measurement of malignant pleural mesothelioma tumor volume. This will help measure how well a treatment plan is working. And it will also help clinicians come to a quicker diagnosis. This tool’s performance may be further improved with more training and calibration.
Sandach, Patrick, Robert Seifert, Wolfgang P. Fendler, Hubertus Hautzel, Ken Herrmann, Sandra Maier, Till Plönes, Martin Metzenmacher, and Justin Ferdinandus. “A Role for PET/CT in response assessment of malignant pleural mesothelioma.” In Seminars in Nuclear Medicine. WB Saunders, 2022. https://doi.org/10.1053/j.semnuclmed.2022.04.008