For most cancers, there is visible evidence of a cancer on scans such as CT scans that are done periodically during the course of a patient’s treatment. A baseline scan is done, ideally just shortly before the start of treatment, and new scans done after some fixed duration of treatment are then compared with the new scans. The general concept is to see whether the repeat scans demonstrate tumor shrinkage, an increase in the size of measurable disease or new lesions (indicating progression), or stable findings. For clinical trials, there is a formal definition of complete response, partial response, stable disease, or progression that are incorporated into the RECIST criteria (Response Evaluation Criteria in Solid Tumors), but clinical practice doesn’t tend to be as precise. Obviously, we are happy to see tumor shrinkage even if it falls shy of the formal definition of a partial response, and stable disease is often very welcome compared to an alternative of disease progression.

The role for PET scans in advanced disease to assess response to therapy remains a controversial area. A CT can provide plenty of helpful information for assessing response to therapy once stage has been established, and CT scans are the well studied and validated metric for assessing interval change for cancers with measurable disease. Some people favor getting PET/CTs to clarify response in extreme detail, but there is a real risk of identifying clinically insignificant changes, such as by a minimal increase in the PET uptake of a tumor that remains stable in size, that might lead to a change in management that isn’t clearly necessary.

A related issue is the use of serum tumor markers, which are proteins produced by the cancer, to guide treatment decisions. For some cancers that often don’t have visible evidence of cancer, tumor markers are a favored approach to assessing response (an example is prostate serum antigen (PSA) to measure the ongoing course of prostate cancer in a man). In other cancers, such as pancreatic cancer, a marker like CA 19-9 is generally accepted as a useful index of disease activity, as CA-125 is for colon cancer. However, not all cancers make these markers, leading to their being used with less of a clear role in many cancers. Breast cancer, lung cancer, and some others may have patients with increased serum tumor markers, but not that reliably. Such serum tumor markers are not universally accepted as a standard measure of monitoring disease, and oncologists tend to vary in their level of enthusiasm for using these in decision-making. A leading concern of those who do not favor using them to guide treatment decisions is that, like subtle changes on a PET scan, changes in a serum tumor marker when scans show stable disease (assuming there is visible evidence of disease on imaging) might lead to a decision to change treatments in the setting of clinically insignificant changes.

Individual physicians have different perspectives about their reliance on PET scans and serum tumor markers in monitoring the course of a cancer, but for most solid tumors (cancers of solid organs where there is visible evidence of the cancer), changes in the size of known cancer on serial CT scans at regular intervals of follow-up remain the best studied and most validated way to assess response to treatment or monitor for progression off of treatment.

See additional links for more details:

Stable disease: Is the glass half-empty or half full?
Practical Principles on the approach to mild progression
Incorporating tumor cavitation into response assessment
Podcast discussion of difficulty assessing response after chemo/radiation

Cancer 101 FAQ: Primer on PET Scans