Biomarkers Inform Immunotherapy Decisions in Advanced NSCLC

In Partnership With:

Balazs Halmos, MD, MA, discusses the advances that have been made in the treatment of patients with metastatic non–small cell lung cancer, ongoing biomarker research, and the importance of molecular testing.

Balazs Halmos, MD, MA, director of Thoracic Oncology and Clinical Cancer Genomics at the Montefiore Albert Einstein Cancer Center

Balazs Halmos, MD, MA, director of Thoracic Oncology and Clinical Cancer Genomics at the Montefiore Albert Einstein Cancer Center

Balazs Halmos, MD, MA

Molecular testing is a critical component in the treatment of patients with metastatic non—small cell lung cancer (NSCLC), according to Balazs Halmos, MD, MA, as the identification of biomarkers can help inform whether single-agent immunotherapy, chemoimmunotherapy, or combination immunotherapy should be used.

“In the past 5 years, we’ve seen a dramatic transformation in how these patients are being managed with the introduction of immunotherapy,” said Halmos. “Now, basically all patients will receive immunotherapy 1 way or another. [We now need to decide] whether we should use chemotherapy in addition [to immunotherapy] and when we should use combination immunotherapy approaches.”

PD-L1 is currently the strongest available biomarker to guide treatment decisions, according to Halmos. If a patient has a high PD-L1 tumor proportion score (TPS), for example, single-agent immunotherapy could be the optimal treatment approach, based on data from the phase III KEYNOTE-024 trial. The trial, which examined the use of pembrolizumab (Keytruda) versus chemotherapy in treatment-naïve patients with advanced NSCLC in patients with PD-L1 expression ≥50%, showed a significant improvement in overall survival (OS) with the single-agent immunotherapy at 30.0 months versus 14.2 months, respectively (HR, 0.63; 95% CI, 0.47-0.86).1

Although PD-L1 is less informative with regard to selecting the optimal patients for chemoimmunotherapy, a combination approach composed of pembrolizumab and chemotherapy demonstrated superiority over chemotherapy alone in several trials, including the phase III KEYNOTE-189 trial2 in patients with nonsquamous NSCLC and the KEYNOTE-407 trial3 in those with squamous disease. As such, for patients who did not have a high PD-L1 TPS, chemoimmunotherapy can be a beneficial treatment option, according to Halmos.

In an interview during the 2020 OncLive® State of the Science Summit™ on Lung Cancer, Halmos, director of Thoracic Oncology and Clinical Cancer Genomics at the Montefiore Albert Einstein Cancer Center, discussed the advances that have been made in the treatment of patients with metastatic NSCLC, ongoing biomarker research, and the importance of molecular testing.

OncLive: How has immunotherapy transformed the metastatic NSCLC paradigm?

Halmos: The treatment of patients with metastatic NSCLC has [evolved]. Molecular testing is very important and it helps us to identify patients with key alterations, such as EGFR, ALK, and ROS1. For patients [with those alterations], targeted therapies are certainly the best treatment approach.

Still, 80% to 85% of our patients with metastatic nonsquamous NSCLC will not have any of these driver alterations. Up until a few years ago, the conventional treatment [for those without these mutations] was systemic chemotherapy. Now, we have 3 key potential approaches: single-agent immunotherapy, chemoimmunotherapy, and combination immunotherapy. To put this into context, you need insight into the basic biology [of this disease]. Many lung cancers are inflamed tumors that are already recognized by the immune system; T cells are invading the tumor and the tumor cells hijack a normal pathway to prevent autoimmunity. By producing PD-L1, [tumor cells] can inactivate these incoming T cells. As such, offering checkpoint inhibition basically disengages this pathway and allows T cells to fully activate and kill cancer cells.

What pivotal trials are examining single-agent immunotherapy? In which patients is this approach most beneficial?

We’ve seen with the KEYNOTE-024 and KEYNOTE-042 studies that [treatment with] pembrolizumab induces dramatic responses and results in excellent outcomes compared with conventional combination chemotherapy in biomarker-selected patients with a high production of PD-L1.

In the KEYNOTE-024 study, patients with high PD-L1 expression, meaning a TPS score of more than 50%, had much better outcomes with single-agent pembrolizumab than they did with conventional chemotherapy. These patients also had better quality of life and fewer toxicities. These data basically transformed how we care for our patients with high PD-L1 [expression].

KEYNOTE-042 looked at a larger patient population; any patients with a PD-L1 TPS of 1% or more were evaluated. Investigators sought to confirm the benefit of single-agent immunotherapy. They also demonstrated that the biggest benefit [with this approach] is in the patients with the highest PD-L1 expression scores. Furthermore, in patients with intermediate PD-L1 scores [ranging from] 1% to 49%, immunotherapy appears to perform just about the same as chemotherapy. However, I would believe that we have better treatment [options available] for that particular patient population. Nonetheless, it was also nice to see confirmation of this approach.

Another checkpoint inhibitor, atezolizumab (Tecentriq), also demonstrated similar benefits in a biomarker-selected patient population. [This trial utilized a different scoring, which is] based on PD-L1 expression in both the tumor cells (TC) and the immune cells (IC). Patients with the highest [PD-L1] expression, [TC3 or IC3] scores, benefited greatly from atezolizumab compared with chemotherapy.

What are some of the data with chemoimmunotherapy in this space?

With combination chemoimmunotherapy, we have seen a similarly transformational change [in care]. In the KEYNOTE-189 trial, which focused on patients with nonsquamous disease, and the KEYNOTE-407 trial, which focused on patients with squamous disease, chemoimmunotherapy was compared with chemotherapy alone and showed very substantial benefits across the board—not just for the patients with high PD-L1 expression. The benefit was observed in patients with intermediate or low PD-L1 scores as well, including those with a TPS of 0. This is certainly very important for treating clinicians to remember.

Similar benefits were seen with atezolizumab as well, in the IMpower150 trial in the context of quadruplet therapy including bevacizumab (Avastin), as well as in the IMpower132 trial, albeit with somewhat lesser benefits. The FDA approved the combination explored in the IMpower150 study, which was composed of atezolizumab plus bevacizumab (Avastin), paclitaxel, and carboplatin, [for use in the frontline treatment of patients with metastatic nonsquamous NSCLC without EGFR or ALK aberrations].

Are chemotherapy-free approaches under exploration?

More recently, we're seeing emerging data from the pivotal CheckMate-227 study, which is evaluating the chemotherapy-sparing combination immunotherapy approach composed of the anti-PD-1 and anti-CTLA-4 agents nivolumab (Opdivo) and ipilimumab (Yervoy). This combination showed improvement in progression-free survival in the tumor mutational burden (TMB)—selected patient population, and more importantly, an OS benefit in the PD-L1–selected patient population.

A significant P value and a hazard ratio of around 0.7 [was observed in] [PD-L1—positive] patients who received the immunotherapy combination. This [combination is being reviewed by] the FDA and is something to watch out for in the coming weeks.

PD-L1 expression is used often to inform treatment decisions. What are some of the challenges still faced with its use? Is TMB being used to inform treatment decisions?

A number of biomarkers help us make treatment choices. There's PD-L1 and TMB, as well as other factors that are quickly emerging. Clearly, PD-L1 is currently the best biomarker that we have for treatment selection, although it suffers from a number of issues related to different antibodies that have been used, different scoring systems, as well as heterogeneity of PD-L1 expression over time and within a particular tumor geography as well.

Still, using PD-L1, we can select patients with a high PD-L1 score for single-agent immunotherapy, which is clearly superior to chemotherapy. Lately, the higher the PD-L1 score [that we saw], the more confidence we can have in the outcome with single-agent immunotherapy; that's something important to remember and watch out for.

In the context of chemoimmunotherapy, PD-L1 is not helpful across the board. [However, we do know that the] combination of immunotherapy and chemotherapy is superior to chemotherapy alone. As such, in a way, we can select patients [with high PD-L1 scores for] single-agent immunotherapy and [use] chemoimmunotherapy for the rest of the patient population. Of course, we need to keep other factors in mind as well.

One that has been emerging but hasn't been fully validated is TMB. TMB, which also suffers from a bit of a lack of harmonization, suggests some benefits in terms of being a predictive biomarker in the context of single-agent immunotherapy, as we've seen recently from the KEYNOTE-010 and KEYNOTE-042 studies. However, again, TMB is not predictive in the context of chemoimmunotherapy.

I also wanted to highlight that the 2 tests complement each other; they are not overlapping. For example, a patient with a high PD-L1 score and high TMB will have much superior benefit with immunotherapy than with just either factor alone. These are helpful complementary factors, with PD-L1 being fully validated and TMB being an emerging marker.

TMB also suffers from practical issues; usually it takes a couple of weeks to obtain, which makes it very difficult to integrate into day-to-day treatment approaches. However, we're now seeing ctDNA-based blood TMB emerging as a potential option as well; this is also something to watch out for as this could greatly improve the practicality of integrating TMB into decision making.

Many biomarkers are emerging, some of which are negative. For example, lately we have seen that tumors with STK11 abnormalities are very unlikely to respond to immunotherapy. In the same vein, EGFR and ALK alterations also identify low—mutation-burden tumors, which benefit much more from targeted treatment interventions than they would from immunotherapy.

As clinicians, we need to look at the whole spectrum of available biomarkers. PD-L1 is the key treatment selection marker, but it’s not the only one we need to [pay attention to]. We need to [look at all] biomarker data to make the best treatment selection for our patients. The main treatment decision [that we face is choosing] between single-agent immunotherapy versus chemoimmunotherapy, with the possibility of combination immunotherapy dependent on the FDA's ultimate decision, based on data from CheckMate-227.

What is your advice on biomarker testing?

We have to keep in mind that biomarker testing is an absolutely key integral part of us providing the best treatment to our patients, not just in the frontline setting, but throughout the patient's treatment continuum. As such, it's very important to communicate with our colleagues about obtaining tissue and then working through to optimize the usage and to provide the right stewardship through the diagnostic path that we use. We want to test for PD-L1 as well as the key molecular biomarkers. What are the key molecular biomarkers? Of course, that's a dynamic and ever-changing area. Currently, that includes 8 or so molecular biomarkers that we should continue to test for, and we usually call for multigene panels.

Fortunately, ctDNA testing has allowed us to expand our ability to complete these tests for our patients, even with limited tissue availability or major time constraints, such as patients needing treatment decisions quickly. As such, I call on my colleagues to consider ctDNA as an additional helpful tool as they complete their biomarker testing.

What is your take-home message to your colleagues?

At the moment, our most important guide is PD-L1. As such, with regard to anyone with a PD-L1 TPS score of more than 50%, a consideration can be made for single-agent immunotherapy versus chemoimmunotherapy. In helping make this distinction, single-agent immunotherapy has lesser toxicities. Usually I offer the [option that provides my patients with] better quality of life, so for the right patient, that might be the better choice. However, we do want to consider other markers into this calculation. For example, a patient with a low mutation burden with STK11-mutated disease, or someone with either an EGFR or ALK alteration, should be viewed very differently.

I do look at patient-specific factors as well, such as age, comorbidities, frailty, and the patient's willingness to undergo more aggressive treatment approaches. For a patient with a PD-L1 TPS score of less than 50%, the best treatment at the present time is chemoimmunotherapy, unless for whatever reason the patient is not a good candidate for that approach. For example, elderly, frail patients might not be good candidates for chemotherapy. Additionally, certain types of organ dysfunction might exclude patients or make them poor candidates for chemotherapy. We have several factors to consider, but for the large majority for the patients, chemoimmunotherapy will be the right decision.


  1. Reck M, Rodríguez-Abreu D, Robinson AG, et al. Updated analysis of KEYNOTE-024: pembrolizumab versus platinum-based chemotherapy for advanced non-small-cell lung cancer with PD-L1 tumor proportion score of 50% or greater. J Clin Oncol. 2019;37(7):537-546. doi: 10.1200/JCO.18.00149
  2. Gadgeel SM, Garassino MC, Esteban E, et al. KEYNOTE-189: updated OS and progression after the next line of therapy (PFS2) with pembrolizumab (pembro) plus chemo with pemetrexed and platinum vs placebo plus chemo for metastatic nonsquamous NSCLC. J Clin Oncol. 2019;37(suppl 15; abstr 9013). doi: 10.1200/JCO.2019.37.15_suppl.9013
  3. Paz-Ares L, Luft A, Vicente D, et al. Pembrolizumab plus chemotherapy for squamous non-small-cell lung cancer. N Engl J Med. 2018;379(21):2040-2051. doi: 10.1056/NEJMoa1810865
Related Videos
View All
Related Content