Metastatic Gastric Cancer: Identifying Disease Progression

David H. Ilson, MD, PhD: How do we determine that a patient is progressing on first-line treatment? Typically, we do CT [computed tomography] scan imaging every 2 to 3 months; we follow blood work, liver function studies, and blood counts; and we see the patients and assess symptoms. Certainly, trigger signs would be the development of increasing fatigue or weight loss or abdominal pain or pain at the sites of known metastatic disease. Clinical deterioration would certainly prompt us to order earlier imaging. It’s always challenging when we have a patient who’s on frontline treatment and they’re doing well symptomatically but their imaging shows progression of disease. Then we have to determine if the progression of disease is significant enough such that we would change treatment. Certainly, if we see substantial disease progression on imaging, even in the absence of symptoms, that would be a trigger to change to second-line treatment.

Daniel V. Catenacci, MD: In addition to clinical symptoms that signal progression on first-line therapy, we have nonclinical signs that we can see. These include increasing tumor markers, for example. We often follow tumor markers CEA [carcinoembryonic antigen] and CA [carbohydrate antigen] 19-9. Even if they’re normal at onset in the newly diagnosed setting, occasionally these will start to rise over time on therapy, whether it be on full doublet chemotherapy or even on maintenance therapy.

In addition to that, we’re looking at scans and we routinely use imaging: preferably CT scans of the chest, abdomen, and pelvis every 8 weeks or so. We will obviously be looking at those. Sometimes patients are progressing on tumor markers and/or scans without having clinical progression. These would be grounds for considering changing therapy. Finally, there is the notion that we can identify progression earlier than with clinical symptoms, tumor markers, and CT scans. One day, we might be looking at ctDNA [circulating tumor DNA] analysis in the blood by next-generation sequencing. We know that we can identify this in the blood at baseline in 85% or more of patients with newly diagnosed disease.

There is a subgroup of patients who aren’t shedding DNA. Those would be patients who have low tumor burden and patients with diffuse peritoneal disease only. They don’t have high levels in the blood. The vast majority, on the other hand, will have ctDNA in the blood at baseline. Over time, we know that with therapy, most patients will have a drop in the ctDNA in the blood. It acts like a tumor marker. Over time, when they’re generating resistance, you will see that to start to rise. We see that happen several months before we would see tumor markers rise, CT scans change, or symptoms worsen.

This would be another nonclinical way of identifying progression. But at this time, we don’t really use that to change therapy because there’s not enough evidence to show that would change much. But that’s an area of investigation, and changing to a more effective therapy in the second line earlier may allow for continued control for longer. That remains to be proven as opposed to waiting for our normal clinical methods of identifying progression, which are the tumor markers, CT scans, and clinical symptoms.

Salah-Eddin Al-Batran, MD: The most important trigger to proceed from first-line to second-line treatment is disease progression. In gastric cancer, time to progression is unfortunately relatively short. The majority of patients experience disease progression within five months in the first-line setting. What we try to do is to be very cautious and perform MRI [magnetic resonance imaging] or CT scans to discover disease progression early and switch the treatment or change the treatment to a second-line treatment. In terms of planning the treatment or in terms of sequencing the treatment, I think that we should make a treatment plan that enable patients to receive as many lines of treatment as appropriate. We know that survival is longer when patients receive multiple lines of treatment. The scheduling is dependent on the previous therapy. Depending on the previous therapy, we plan the second-line therapy and we try to use drugs that are not completely cross-resistant to the compounds the patient has received in first-line therapy, or even in more recent settings.

Transcript edited for clarity.