Many patients with cancer experience infections requiring treatment. However, David Pinato, MD, MRes, PhD, and colleagues recently found that in patients with non-small cell lung cancer (NSCLC), melanoma, and other tumor types, broad-spectrum antibiotic treatment administered within 30 days prior to the commencement of immune checkpoint inhibitor therapy dramatically worsened overall survival (2 vs 26 months) and increased the risk of treatment-refractory disease. In this interview with i3 Health, Dr. Pinato discusses the possible causes behind his study's startling results. He also answers an important practical question: what should be done about patients with cancer who need both antibiotics and checkpoint inhibitor therapy?
Did any of your findings surprise you?
David J. Pinato, MD, MRes, PhD: Yes, very much so. We were surprised in particular by the very strong difference in overall survival—over 20 months between the two groups—and the fact that the difference in survival was preserved across many different tumor sites. This is not something that we would have expected here.
We saw that one of the potential explanations for the antibiotics decelerating prognosis was the association with comorbidities, especially in a group of patients like this one, where we had around 120 patients with NSCLC and therefore a strong background of chronic pulmonary disease. Obviously, the provision of antibiotics could just be a signal of a subset of patients who are a bit more unwell than others and require antibiotics for that reason. However, there was a lot of surprise, especially when we looked at the response rates, which were based on computed tomography (CT) scans 6 to 8 weeks after having started the immunotherapy. We saw that there was a much higher chance of progression compared to no antibiotic treatment. That was quite striking.
In the study, you mention gut dysbiosis as a possible cause. Can you explain more about that?
Dr. Pinato: Gut dysbiosis is basically a shift from a normal state in the flora that constitutes the bacteria that are typically present within the gut. If you have healthy gut microbiota, that associates with a lot of different things; it's not only cancer that is influenced by the gut microbiome. There are many other chronic diseases for which the gut microbiota is a potential link in terms of how the disease accelerates or decelerates, potential onsets, and predisposition.
In the specific setting of immunotherapy, there have been a couple of studies in the past that examined the diversity of the gut bacteria. What has been found is that the more diverse the flora is, the more likely it is that patients will be able to respond well to immunotherapy.
Based on this finding, we think that the disruption of a good ecosystem within the gut might potentially change the education, so to say, of the immune system towards the cancer. It's almost like the maturation of a sustainable and active anticancer response is being disrupted by exogenous treatments, such as antibiotics.
It's not really just in cancer; as I said, we know that a lot of pro-inflammatory conditions, a lot of chronic diseases, have certainly got a component that stems from the microbiota. The issue is really how this relationship works. I think our study was quite important because it's one of the first to validate the concept of this deterioration, which was exhibited so strongly here in terms of both survival and response rates. I think that the reason why this paper is receiving a lot of attention is because we are genuinely seeing that antibiotics might play a role and that this message could be real.
What do you think is behind the discrepancy between the extremely strong association of poor checkpoint inhibitor outcomes with prior antibiotic use and the lack of association with concurrent antibiotic use?
Dr. Pinato: Everything we know about cancer immunology is linked with this idea of the cancer immunity cycle, whereby there are basically a series of predefined events that have to happen in a sequence so that anticancer immunity can work at the maximum. If all those steps do not happen in a correct sequence, immunity can be impaired as a result. Part of this sequence is the fact that the antigens that are specifically associated with cancer have to be processed and presented to the immune cells, and therefore the immune system has to be primed to recognize cancer. A failure in any of these steps leads to the presence of a deficient immunity against cancer.
It's quite interesting that the possible reason why we have the effects for prior dosing but not concurrent dosing is because antibiotics might potentially act on the priming phase of the cancer immunity interaction. This is consistent with what we know in terms of the influence of the microbiota and what the microbiota does: it seems to be able to educate the effector T-cell response, which is technically what is being upregulated as a result of immune checkpoint inhibitor therapy.
So if a patient is lacking that preexisting indication, that prompting that might be stemming from exposure to a healthy microbiota or may be potentially facilitated by it, then the patient is not going to be able to mount an effective response because whatever is amplified when you give the immune checkpoint inhibitors is actually not there in the first place.
Obviously, we can't really say whether this is true or not, so I think this is perhaps the next question that needs to be asked: how is it possible that such a relationship is shaped only with the preexisting antibiotic treatment? What we are doing at the moment in the lab is to try and understand the basis of this association, but even just looking at the epidemiological evidence for prior versus concurrent antibiotic use, there is a bit of a signal there that justifies our interest and our appetite for more studies.
What should physicians do for patients who require checkpoint inhibitor therapy but are also in need of antibiotics?
Dr. Pinato: This is a really crucial question. I personally think that we have to use our data on checkpoint inhibitors with a lot of caution. Obviously, this is further evidence—ours is not the only study that has looked at this particular phenomenon—but I think that the use of antibiotics should always be dictated on a case-by-case basis, with the treating physician being able to fully make a judgment as to whether the use of antibiotics is required in the individual patient.
However, what we found in looking at the charts of our own patients exposed to antibiotics was that some of these patients were technically treated for a suspected infection. Perhaps our study might be of help in further strengthening the concept of antibiotic stewardship in patients with cancer: now that there is more evidence that there could be worrisome adverse effects from antibiotic treatment, maybe clinicians should think twice about giving antibiotics if there are no clear signs of infection or if it's just a suspicion.
It's a very difficult message to give out because obviously, patients need to receive all the adequate care, including antibiotics if required. However, the other end of the question is concurrent antibiotic treatment and outcome. Because concurrent treatment seems to be safe in the context of checkpoint inhibitor therapy, I think we can at least rest assured that once the patient is on treatment, there doesn't really seem to be a lot of evidence that we should withhold antibiotic treatment in the case of potential infection. That's a really good point to reinforce. Patients that come with toxicity often have a very grave symptomatology, where infection is frequently part of the differential diagnosis; withholding antibiotics in such cases would really almost be criminal. I think we have to take things with a pinch of salt and wait for more data to come through, especially data that will solidly tell us what is happening behind these survival issues. In the meantime, it may be good to proceed with a bit more caution and a bit more education.
About Dr. Pinato
David J. Pinato, MD, MRes, PhD, is a Clinical Senior Lecturer, Clinician Scientist, and Consultant Medical Oncologist in the Department of Surgery and Cancer of the Faculty of Medicine at Imperial College London. He leads a translational research program that concentrates on the early clinical implementation of experimental anticancer therapies, particularly in the area of immunotherapy. He has led a number of first-in-class studies of immune checkpoint inhibitors in liver cancer.
For More Information
Pinato DJ, Howlett S, Ottaviani D, et al (2019). Association of prior antibiotic treatment with survival and response to immune checkpoint inhibitor therapy in patients with cancer. JAMA Oncol. [Epub ahead of print] DOI:10.1001/jamaoncol.2019.2785