Prognostic Markers in Peripheral Blood Shine Light on Future Directions in NSCLC

Using data from a post-hoc analysis of the phase 3 EMPOWER-Lung1 (NCT03088540) and the EMPOWER-Lung3 trials (NCT03409614), investigators sought to validate the prognostic implications of neutrophil/lymphocyte ratio (NLR) and other peripheral myeloid cells in patients with non–small cell lung cancer (NSCLC) treated with cemiplimab-rwlc (Libtayo).

Improvements in overall survival (OS) and progression-free survival (PFS) were significantly associated with NLR, neutrophil/monocyte/lymphocyte ratio (NMLR), monocytes, and log eosinophils. Specifically, the median OS was 10 months (95% CI, 9-13) for patients with a mean NLR greater than 3.98 and was 20 months (95% CI, 17-24) for those with a mean NLR of 3.98 or lower (HR, 1.11; 95% CI, 1.08-1.13; P < .001). The estimated 1-year OS rates were 45% and 69%, respectively.

For patients with a mean NMLR of greater than 4.25, the median OS was 10 months (95% CI, 9-13) with a 1-year OS rate of 46% compared with 22 months (95% CI, 17-24) and a 1-year OS rate of 70% for patients with a mean NMLR of 4.25 or lower (HR, 1.10; 95% CI, 1.08-1.13; P < .001). The median OS for those with a maximum monocyte cut of 1 or higher had a median OS of 10 months (95% CI, 9-14) and a 1-year OS rate of 45% compared with 19 months (95% CI, 16-22) and a 1-year OS rate of 65% for those with a maximum monocyte cut of 1 or lower (HR, 1.63; 95% CI, 1.26-2.10; P < .001).

Among those with a maximum cut of log esosinophlis of greater than –2.13, the median OS was higher at 17 months (95% CI, 16-22) and a 1-year OS rate of 66% compared with a median OS of 12 months (95% CI, 10-15) and an estimated 1-year OS rate of 50% for those with log esosinophlis of –2.13 or lower (HR, 0.91; 95% CI, 0.86-0.96; P < .001). Finally, in findings from a decision tree analysis, investigators noted that NMLR was the most prognostic of OS, and that, regardless of treatment, those with a mean NMLR of 4.25 or higher were associated with a greater risk of death.

Although the findings are hypothesis-generating only, investigators added in the presentation that the addition of prospective data can turn these preliminary findings into clinically actionable markers used to predict outcomes with anti–PD-1 therapy for patients with advanced NSCLC.

“As a thoracic oncologist, I’m keenly interested in all the work in lung cancer and also esophageal cancer that I treat, and I am looking for ways that we can use prognostic information throughout a patient’s diagnostic and therapeutic journey,” Joseph Christopher Murray, MD, said in an interview with OncLive®. “This information can be as detailed as white blood cell differential counts, like we did in this study, or as complex as cell-free DNA analysis from tumor specimens from patients. But what I want to do is be able to inform patients better [using data from] this work. That’s the reason for doing this. And no matter how many complex experiments we do, bringing it back to the patient is the key part.”

Murray, an instructor of oncology and codirector of the Lung Cancer Precision Medicine Center of Excellence at Johns Hopkins Sidney Kimmel Cancer Center in Baltimore, Maryland, spoke with OncLive about the role of prognostic markers in NSCLC and how data from the EMPOWER trials will guide future directions in the space.

OncLive: What was the rational for evaluating these markers for prognostic influences in the EMPOWER studies?

Murray: We use chemotherapy and immunotherapy in the first line for patients with metastatic or advanced NSCLC. The EMPOWER-Lung1 and EMPOWER-Lung3 studies used similar treatments of chemotherapy and immunotherapy across a population [of patients with NSCLC] that was broad and inclusive, independent of known biomarkers, including PD-L1 status. What we wanted to do is look to see how peripheral [myeloid] cells affected prognostication for patients in these 2 sets of trials.

These patients, as many patients who are enrolled on clinical trials, were quite fit with a good performance status. We looked at specific biological markers in the blood that are sometimes indicative of systemic inflammation that can affect response to chemotherapy and immunotherapy. These patients [had characteristics that] were well balanced across the clinical trial arms. And we did have baseline data [from] their peripheral blood to start. Those were the patients who were included [in the analysis]—those with baseline peripheral blood data, as well as on treatment data.

What were the methods of this study?

We use those baseline blood draws to look at a complete blood count, which differentiates the white blood cells into different classes, which have different functional effects. We used that baseline metric to stratify the response for those patients as far as PFS and OS.

Please summarize the topline data presented at the 2023 ASCO Annual Meeting.

We looked at the relevant relationship between these different white blood cell subsets obtained at baseline. Specifically, we looked at NLR, but also introduced a new cell population that’s been less studied in this area called monocytes, as well as eosinophils. We found that the NMLR was the strongest predictor for prognostication for OS in our patient population. There have been nascent data studied in other retrospective work. But this is a post hoc analysis from a prospective clinical trial where we were able to identify this.

This has important implications for patients, but also clinicians and how to think about a patient who presents with different changes in their white blood count, and how that could affect their response to therapy. That’s useful to be able to talk to patients about.

What are some of the next steps for this research?

One key thing is to use information that we get a baseline and to follow it on treatment. For example, do these changes, [such as] an increased NLR or an increased neutrophil monocyte to lymphocyte ratio hold up on treatment? If patients have persistent findings on their white blood cell counts that reflect this, does that mean they do better or worse or the same as other patients? [In addition to] baseline metrics, an important follow-up for the study will be on treatment metrics.

With every post hoc, secondary analysis, even from a prospectively designed clinical trial, there are caveats. This was not one of the initial primary end points for the study and was an exploratory analysis. We need more prospective data to elucidate what's going on with these different cell populations so we can use it in a prognostic way. For example, [we should aim to] identifying nomograms that enable clinicians to talk to patients about how these findings at baseline may affect their on-treatment response.

What is the main takeaway message for colleagues regarding this research?

The takeaway is for my clinical colleagues, we’ve all seen in patients who present to us with elevated white blood counts. And the differential can give us a little bit of clue [of what to expect on-treatment]. If we think there may be benefit for certain therapeutic paradigms, it doesn’t mean we should change how we treat the patient, but it does give us insight into what those patients may experience through their therapeutic journey. This is a key part about prognostication: We want to enable patients to be informed.

Reference

Murray JC, McIntyre DAG, Anagnostou V, et al. Peripheral myeloid cells as prognostic markers in patients (pts) with non-small cell lung cancer (NSCLC) treated with cemiplimab: pooled analysis of EMPOWER-Lung 1 and EMPOWER-Lung 3 phase 3 trials. J Clin Oncol. 2023;41(suppl 16):9028. doi:10.1200/JCO.2023.41.16_suppl.9028