Early lung mets often are not particularly large contributors to CEA levels vs interferences.
Also your CEA may not be your best marker. Clearly your CEA range is lower than the CRC average.
There are several directions for making blood work more meaningful, or to at least clarify the story.
1. Monitor more of the traditional blood data, some more frequently.
2. Smooth the data in terms of active control of inflammation, chemo, sugar and other interferences to lower the scatter.
3. add more expensive, difficult or less specific markers
More frequent CEA data with inflammation markers would at least measure the scatter and part of the cause, a relative cold comfort. Post 2nd surgery, we achieved mirror smooth low CEA with anti-inflammatories and supplements, like IV vitamin C, other nutraceuticals, and baby aspirin (80-100 mg, more is likely not better), giving us more insight or certainty when things started to move ~13 months later, following supply loss of a chemo component.
A number of forum commenters have had to depend on CA199 as their best marker (using one brand of test if not the same lab). One even got their oncologist to do the tissue stain for CA199 to get both the oncologist and insurance on board for monitoring CA199 since the CEA was useless.
For some fraction of CRC patients, CA199 is a more prominent CRC marker but CA199 gives relatively noisy data from various interferences. US medicine studiously ignores CA199 for CRC and US drs know few specifics about CRC and CA199 (some EU countries do use CA199). This stmt probably includes your pancreatic cancer oncologist where CA199 is the primary marker.
LDH, a liver panel, is likewise often sensitive to cancer mets although it is a non specific marker.
DNA blood tests are expensive and more specific, but often not as sensitive as CEA for most people, but may provide re-assurance where CEA fails. I would clarify about potential chemo interference with any particular DNA blood test version and timing.
All of these work best with series in combination, more data sooner.
Traditional cyclical chemo itself is an interference with blood data that has to be considered.