Keywords

1 Introduction

Clearly, we are seeing that lung cancer improves its survival year by year; NSCLC 2-year relative survival increased from 34% for persons diagnosed during 2009 through 2010 to 42% during 2015 through 2016, including absolute increases of 5% to 6% for every stage of diagnosis; survival for small cell lung cancer remained at 14% to 15%. This is due to the improvement in treatments, where immunotherapy plays a fundamental role [1].

Immunotherapy treatment is now a reality in clinical practice, and knowledge mechanism of action is key in understanding the benefit of the improved survival of the lung [2]. The development of immune checkpoint inhibitor (ICI) agents targeting cytotoxic T-lymphocyte antigen-4 (CTLA-4), programmed cell death protein 1 (PD-1), or programmed cell death protein ligand 1 (PD-L1) has garnered tremendous interests in the field of immuno-oncology because of the recent successful applications in multiple advanced cancers. Although CTLA-4 is the first immune checkpoint molecule identified, the PD-1/PD-L1 axis has been widely investigated due to the role in the exhaustion of CD8+ T cells. Physiologically, PD-1/PD-L1 has the task of limiting the activity of T cells in peripheral tissues at the time of an inflammatory response to infection, thereby limiting autoimmunity. Similar to CTLA-4, PD-1 is expressed on activated T cells and inhibits T-cell responses by interfering with T-cell receptor signaling. PD-1 has two ligands, PD-L1 (B7-H1) that is expressed on antigen-presenting cells (APCs), macrophages, fibroblasts, and T cells and PD-L2 (B7-DC) that is predominantly expressed on antigen-presenting cells (APCs). PD-L1 is also overexpressed in several solid tumors, while PD-L2 is expressed relatively rarely. The role of CTLA-4 and PD-1/PD-L1 in immune suppression and their expression in solid tumors provided the rationale for their therapeutic exploitation. Moreover, CTLA-4 and PD-1 exert their effects through separate pathways, and therefore simultaneous targeting of both pathways has also been evaluated to restore antitumor immunity [3, 4].

Since the first demonstration of activity of PD(L)-1 agents in lung cancer in early clinical trials in 2012, immune checkpoint blockade (ICB) has emerged as a novel effective therapeutic strategy in different clinical settings and determined a dramatic shift in the therapeutic landscape of both NSCLC and SCLC (Fig. 1) [5]. Several biological prognostic and predictive factors in blood and tissue samples have been identified, but unfortunately no single biomarker can perfectly discriminate between responders and nonresponders, and PD-L1 immunohistochemical expression still remains the only applicable marker in clinical practice to date [6].

Fig. 1
figure 1

Role of immune checkpoint inhibitors in lung cancer and immune-modulatory activities of conventional treatment strategies [3, 7,8,9].

Abbreviations: MHC major histocompatibility complex, TAA tumor-associated antigens, IFN interferon, TH1 T-helper 1, CTL cytotoxic T-cell lymphocyte, NK natural killer, Tregs regulatory T cells, MDSCs myeloid-derived suppressor cells, TCR T-cell receptor, DC dendritic cell (Credit: Created with BioRender.com)

Full size image

To date, the primary biomarker used for lung cancer has been PD-L1 [10]. Different immunohistochemical assays have been developed, using different antibodies and scoring systems. However, multiple harmonization studies have consistently reported high concordance between most of these assays (22C3, SP263, 28-8, 73-10, and E1L3N) in terms of PD-L1 expression on tumor cells (TC) [11,12,13,14]. PD-L1 tumor proportion score (TPS) evaluated using the Dako 22C3 assay was developed and validated as the companion diagnostic for single-agent pembrolizumab in pretreated NSCLC in the randomized phase II/III KEYNOTE-010 trial [15]. Based on the positive results in pretreated patients, the use of ICIs was then moved to treatment-naïve patients, and PD-L1 expression represented the most extensively used biomarker for treatment selection of single-agent PD-1/PD-L1 inhibitors versus platinum-based chemotherapy [16]. PD-L1 TPS ≥50% identifies a subgroup of patients that accounts for approximately 30% of the patients with NSCLC that derives greater benefit from single-agent ICIs than platinum-based chemotherapy, and to date, three different agents (pembrolizumab, atezolizumab, and cemiplimab) [17,18,19] have been approved in this setting. Single-agent ICI in PD-L1 low expressors (TPS 1–49%) for first-line therapy is controversial, as the benefit seen in the KEYNOTE-042 trial with pembrolizumab [20] is likely driven by PD-L1 strong expressors [21].

Absence of PD-L1 expression does not conclusively identify patients who will not benefit from immunotherapy, leading to investigation of many other biomarkers [22,23,24].

In this review, we summarized the current status and progress of anti-PD-1/PD-L1 agents in lung cancer treatment.

2 Early-Stage NSCLC and Locally Advanced NSCLC

Approximately 40% of NSCLC patients are diagnosed with locoregional disease that is potentially resectable [25]. Adjuvant platinum-based chemotherapy has been shown to improve survival in patients with stages II–III disease and can be considered high-risk stage IB disease (4 cm, poorly differentiated carcinoma, post-wedge resection, lymphovascular invasion, visceral pleural involvement, unknown lymph node status) [26]. Meta-analyses of randomized phase III trials conducted in the 1990s and early 2000s reported an absolute survival benefit at 5 years of 5% from adjuvant/neoadjuvant approaches in stage IB–IIIA NSCLC compared with surgery alone [27, 28].

Since these trials, the therapeutic landscape of early-stage NSCLC has little improved over the last two decades, and only recently, a randomized phase III trial has reported a survival advantage in a selected patient population (activating EGFR mutations) using osimertinib after platinum-based chemotherapy [29]. ICIs might potentially revolutionize the adjuvant setting, given the well-known ability of immunotherapy of inducing long-term responses, and multiple clinical trials are ongoing (Table 1).

Table 1 Ongoing clinical trials evaluating immune checkpoint inhibitors as adjuvant therapy in radically resected NSCLC
Full size table

Recently, the preliminary results of a single-arm phase II study (NCT03053856) evaluated the postoperative role of pembrolizumab in stage IIIA-N2 NSCLC who has undergone neoadjuvant concurrent chemoradiotherapy (weekly carboplatin/paclitaxel and radiation therapy to 44 Gy in 22 fractions) with curative resection for up to 2 years or until disease recurrence. The primary endpoint is disease-free survival (DFS), with a statistical goal of more than 20 months. Thus far, of 37 patients treated in this trial, 14 patients have discontinued treatment owing to disease progression (n = 9), adverse events (n = 4), or consent withdrawal (n = 1). Adverse events have included grade 4 pneumonitis (n = 1) and grade 3 autoimmune hepatitis (n = 1), which have led to discontinuation, as well as grade 1 or 2 hypothyroidism (n = 6), pneumonitis (n = 5), and skin rash (n = 3) [30].

On March 2021, Roche announced that phase III IMpower010 trial met the primary endpoint, demonstrating a statistically significant improvement in terms of DFS with the use of the PD-L1 inhibitor atezolizumab as compared with best supportive care (BSC) in patients with PD-L1-positive, stages II–IIIA NSCLC who have undergone surgical resection and received up to four cycles of adjuvant cisplatin-based chemotherapy. The presentation of the full results of the study is eagerly awaited.

In addition to the adjuvant setting, immunotherapy might have a role also in the neoadjuvant setting either as monotherapy or in combination with platinum-based chemotherapy. Preliminary data of these studies are encouraging, especially when considering chemo-immunotherapy combinations (Table 2).

Table 2 Selected clinical trials with neoadjuvant ICIs in resectable NSCLC
Full size table

Collectively, chemo-immunotherapy seems associated with higher ORR and increased probability of major pathological response (MPR)/pathologic complete response (pCR). This therapeutic strategy seems more promising than single-agent ICB and moved quickly to phase III. Several randomized trials evaluating the addition of a PD-1/PD-L1 inhibitor to a platinum-based doublet as neoadjuvant therapy in resectable NSCLC (stages I–IIIA) are underway, including KEYNOTE-671 (pembrolizumab), AEGEAN (durvalumab), NCT04316364 (adebrelimab/SHR-1316), NCT04379635 (tislelizumab), JS001 028 III (toripalimab), CheckMate-77 T (nivolumab), and CheckMate-816 (nivolumab/chemotherapy vs. nivolumab-ipilimumab).

In October 2020, Bristol Myers Squibb announced that the CheckMate-816 met its primary endpoint of improved pCR in patients who received nivolumab plus chemotherapy before surgery. The presentation of the full results of the study is expected in the next few months.

In patients with inoperable stage III disease, the use of chemoradiotherapy has been shown to increase survival as compared with radiotherapy alone [44], and concurrent chemoradiation (cCRT) increases 5-year overall survival by 4.5% as compared with a sequential approach [45].

Several studies have shown promise for immunotherapy following cCRT in patients with unresectable stage III LA-NSCLC. The PACIFIC trial (A Global Study to Assess the Effects of MEDI4736 Following Concurrent Chemoradiation in Patients with Stage III Unresectable Non-Small Cell Lung Cancer) reported encouraging phase III data on the use of the anti-PD-L1 antibody durvalumab in this context. PACIFIC was the first study to demonstrate improved outcomes in patients with LA-NSCLC who received an immune checkpoint inhibitor. In this phase III trial, patients with stage III unresectable NSCLC were randomly assigned in a 2:1 ratio to receive either durvalumab (a PD-L1 inhibitor) or placebo as consolidation therapy every 2 weeks for as long as 1 year [46]. The study population consisted of 713 patients who had received cisplatin-based chemotherapy with concurrent radiation to 66 Gy and had no disease progression following treatment. Progression-free survival (PFS), the primary endpoint, was significantly longer in the durvalumab group than in the placebo group (median PFS, 16.8 vs. 5.6 months; P < 0.001). In addition, the co-primary OS remained consistent with that previously reported (stratified HR = 0.69 [95% CI: 0.55–0.86]); the median OS was not reached with durvalumab but was 29.1 months with placebo. The 12-, 24-, and 36-month OS rates with durvalumab and placebo were 83.1% versus 74.6%, 66.3% versus 55.3%, and 57.0% versus 43.5%, respectively [47]. Improved OS with durvalumab was broadly observed irrespective of PD-L1 expression, which is consistent with findings from prespecified and post hoc analyses carried out at the time of the primary OS analysis [46]. Remember that PD-L1 data were based on pre-cCRT samples, which may not reflect changes in expression potentially incurred by cCRT, and should also be taken into consideration when drawing definitive conclusions. PACIFIC was not designed to evaluate the efficacy of durvalumab based on PD-L1 status. Overall, the findings of this analysis underscore the long-term survival benefit with durvalumab after cCRT and further establish the PACIFIC regimen as the standard of care in patients with unresectable stage III NSCLC who do not progress while undergoing cCRT. An exploratory analysis showed that patients who started treatment with durvalumab <14 days from completion of radiation therapy had improved efficacy outcomes compared with those who started treatment ≥14 days from completion of radiation therapy [48].

Besides consolidation after cCRT, other therapeutic strategies under active investigation include the concomitant use of ICIs during chemoradiotherapy (PACIFIC-2, CheckMate73L, EA5181, DETERRED-PART II, NICOLAS, KEYNOTE-799) and the use after sequential chemoradiotherapy (PACIFIC-6) or radiotherapy alone (DUART). The results of these trials will provide additional insights on the role of PD-1/PD-L1 inhibitors in inoperable stage III NSCLC.

3 Pretreated NSCLC

After few years since early clinical sights of activity of PD-1/PD-L1 inhibitors in lung cancer [49, 50], three PD-1/PD-L1 therapies have been approved by the US Food and Drug Administration (FDA) and the European Medicines Agency (EMA) in the second-line setting (nivolumab, pembrolizumab, and atezolizumab), on the basis of phase III studies demonstrating improved overall survival (OS) in comparison with the former standard-of-care therapy docetaxel.

In two phase III trials (CheckMate-017 and CheckMate-057), nivolumab showed an improvement in OS and favorable safety versus docetaxel in patients with previously treated, advanced squamous, and non-squamous NSCLC [51, 52]. After follow-up of 64.2 and 64.5 months for CheckMate-017 and CheckMate-057 [53], respectively, 50 nivolumab-treated patients and 9 docetaxel-treated patients were alive. Five-year pooled OS rates were 13.4% versus 2.6%, respectively; 5-year PFS rates were 8.0% versus 0%, respectively. Nivolumab-treated patients without disease progression at 2 and 3 years had an 82.0% and 93.0% chance of survival, respectively, and a 59.6% and 78.3% chance of remaining progression-free at 5 years, respectively. Treatment-related adverse events (TRAEs) were reported in 8 of 31 (25.8%) nivolumab-treated patients between 3 and 5 years of follow-up, 7 of whom experienced new events; one (3.2%) TRAE was grade 3, and there were no grade 4 TRAEs. Clearly, nivolumab compared to docetaxel exhibited a fivefold increase in OS rate, with no new safety signals. Interestingly, PD-L1 expression as a predictive biomarker produced contrasting results between the two trials, despite similar study designs and the same assessment methods. The different mutational burden of squamous and non-squamous histology, as well as the frequency of oncogene-addicted tumors, might have contributed to this discrepancy. Moreover, a landmark analysis of the CheckMate-057 demonstrated that, excluding patients who had died in the first 3 months, nivolumab was superior to docetaxel in both PD-L1-positive and PD-L1-negative patients [54]. Atezolizumab was compared with docetaxel in pretreated NSCLC in phase II (POPLAR) and phase III randomized studies (OAK), showing improved OS across all PD-L1 expression levels with incremental efficacy results at the increase of PD-L1 IHC expression in tumor cells (TC) or tumor-infiltrating immune cells (IC) using the SP142 assay [55, 56]. A longer OS was observed in patients receiving atezolizumab vs. docetaxel in POPLAR (median OS, 12.6 months vs. 9.7 months; HR, 0.76 [95% CI: 0.58–1.00]) and OAK (median OS, 13.3 vs. 9.8 months; HR, 0.78 [95% CI: 0.68–0.89]. Four-year OS rates in POPLAR were 14.8% (8.7–20.8) and 8.1% (3.2–13.0) for atezolizumab and docetaxel, respectively, and 15.5% (12.4–18.7) and 8.7% (6.2–11.3) in OAK. Most 4-year survivors in the docetaxel arms received subsequent immunotherapy (POPLAR, 50%; OAK, 65%). Of 4-year survivors, most had ECOG PS 0 and non-squamous histology; approximately half were responders (POPLAR, atezolizumab, 7/15; docetaxel, 3/4; OAK, atezolizumab, 24/43; docetaxel, 11/26). Treatment-related grade 3/4 adverse events occurred in 27% and 16% of atezolizumab 4-year survivors in POPLAR and OAK, respectively [57].

The development of pembrolizumab in NSCLC started with the phase I multi-cohort study KEYNOTE-001, which evaluated the safety and activity of this compound and also validated the companion diagnostic 22C3 IHC assay for PD-L1 expression. Patients with squamous and non-squamous tumors were enrolled; however, PD-L1 expression had to be 1% or greater. All patients had progressed on first-line platinum-doublet therapy, and those with driver mutations had also progressed on appropriate TKI therapy [58]. The updated analysis with a 42.6 months follow-up [59] showed that the risk of death was reduced with pembrolizumab versus docetaxel in both the PD-L1 TPS ≥50% group (HR 0.53; P < 0.00001) and the TPS ≥1% group (HR 0.69; P < 0.00001). Median OS was 16.9 months (95% CI, 12.3 to 21.4 months) versus 8.2 months (95%CI, 6.4 to 9.8 months) in the TPS ≥50% group and 11.8 months (95% CI, 10.4 to 13.1 months) versus 8.4 months (95% CI, 7.6 to 9.5 months) in the TPS ≥1% group. Kaplan-Meier estimates of OS at 36 months were higher with pembrolizumab versus docetaxel in both TPS groups, with OS rates of 34.5% versus 12.7% in the TPS ≥50% group and 22.9% versus 11.0% in the TPS ≥1% group. The risk of disease progression or death (per RECIST v1.1 by BICR rather than per investigator) was reduced with pembrolizumab versus docetaxel in the PD-L1 TPS ≥50% (HR 0.57; P = 0.00001) and TPS ≥1% groups (HR, 0.83; P = 0.005). Kaplan-Meier estimates of PFS at 36 months were higher with pembrolizumab versus docetaxel in both TPS groups, with PFS rates of 21.9% versus 1.2% in the TPS ≥50% group and 12.7% versus 1.0% in the TPS ≥1% group.

Not all trials using PD-1 and PD-L1 checkpoint inhibitors for the second-line treatment of advanced NSCLC have yielded positive results. Avelumab, an anti-PD-L1 monoclonal antibody, was compared with docetaxel in the JAVELIN Lung 200 trial [60]. As a result, the OS was not significantly different between the avelumab and docetaxel groups, even in the subgroup with positive tumor PD-L1 expression. High post-study use of ICIs and the non-blinded design of the trial might have affected the results.

Currently, ICIs are now well established as the standard of care for second-line treatment of advanced NSCLC, but there is no data to suggest that one agent is superior to another in that setting. No head-to-head comparison has been conducted. Indeed, the meta-analyses of published studies with ICIs in pretreated NSCLC did not demonstrate significant evidence of survival differences between these agents [61, 62]. Therefore, in clinical practice, factors that could influence ICI selection might include drug access, dosing schedule, costs, and PD-L1 expression.

4 First-Line Metastatic NSCLC

The introduction of the anti-PD-1 nivolumab for metastatic lung cancer in second-line setting was just the beginning in the development of checkpoint inhibitors in different clinical scenarios [16], including first-line treatment either as monotherapy in selected patient populations or in combination with chemotherapy +/− antiangiogenic drugs or in combination with CTLA-4 inhibitors with or without chemotherapy.

Within the revolutionary first-line setting for metastatic lung cancer patients, two pivotal randomized phase III clinical trials have compared pembrolizumab vs. platinum-doublet regardless histology and without driver mutations (EGFR/ALK wild type) in patients with PD-L1 TPS ≥50% (KEYNOTE-024) and in those with PD-L1 TPS ≥1% (KEYNOTE-042).

KEYNOTE-024 met its primary endpoint, reaching its goal of demonstrating the superiority of pembrolizumab vs. chemotherapy for patients with strong PD-L1 expression (TPS ≥50%) regardless of tumor histology [17]. By reaching a median OS of 30 months [63], pembrolizumab is positioned as a less toxic and more effective treatment than platinum-doublet-based chemotherapy in this selected patient population, demonstrating for the first time a survival advantage over platinum-based chemotherapy as first-line treatment in non-oncogene addicted NSCLCs. Based on these results, PD-L1 ≥ 50% in absence of concomitant driver mutations identified a novel subgroup of patients that accounts for approximately 30% of all NSCLCs that benefits from a chemotherapy-free regimen in first line.

Similar results were more recently reported with two ICIs, atezolizumab and cemiplimab, in patients with strong PD-L1 expression without EGFR mutations and/or ALK rearrangements and have been recently approved by the US FDA as first-line options.

Atezolizumab was compared with platinum-based chemotherapy in the randomized phase III trial IMpower110 in PD-L1 selected NSCLCs (PD-L1 expression on at least 1% of tumor cells or at least 1% of tumor-infiltrating immune cells as assessed by the SP142 immunohistochemical assay) [18]. The study demonstrated a statistically significant improvement in OS in the intention-to-treat (ITT) population (patients whose tumors were wild-type with respect to EGFR mutations or ALK translocations) within the subgroup of patients with strong PD-L1 expression (20.2 months vs. 13.1 months; hazard ratio for death, 0.59; P = 0.01). Furthermore, OS and PFS favored atezolizumab in the subgroups with a high blood-based tumor mutational burden (bTMB), assessed through the plasma 394-gene NGS panel FoundationOne CDx Liquid, suggesting a potential utility of this biomarker for patient selection [18].

Cemiplimab was compared with platinum-based chemotherapy in the randomized phase III trial EMPOWER-Lung 1 as first-line treatment in advanced NSCLC with PD-L1 tumor expression ≥50% and no EGFR mutations, ALK translocations, or ROS1 fusions. Patients were ineligible if they had never smoked (defined as ≤100 cigarettes in a lifetime). This is the largest study in this setting (563 patients in the PD-L1 ≥ 50% population) and showed that cemiplimab was superior to chemotherapy in improving PFS (8.2 months vs. 5.7 months, HR 0.54; p < 0.0001) and OS (not reached vs. 14.2 months, HR 0.57; p = 0.0002) in PD-L1-strong positive NSCLC patients [19].

Recently, a multicenter retrospective study analyzed the impact of different PD-L1 expression levels on pembrolizumab outcome in the subgroup of patients with NSCLC PD-L1 TPS ≥50% without EGFR/ALK aberrations. Compared with patients with PD-L1 expression of 50%–89% (N = 107), patients with an expression level of 90%–100% (N = 80) had a significantly higher ORR (60.0% versus 32.7%), a significantly longer PFS (14.5 versus 4.1 months), and a significantly longer OS (not reached versus 15.9 months). These results suggest that in patients with NSCLC and PD-L1 expression ≥50% treated with first-line pembrolizumab, clinical outcomes are significantly improved in NSCLC with a very high PD-L1 expression of ≥90% [64].

Other studies have sought to expand the potential number of patients that might benefit from upfront PD-1/PD-L1 blockage as monotherapy, evaluating these agents in patients with PD-L1 expression ≥1%. The CheckMate-026 failed to demonstrate a survival benefit with nivolumab versus platinum-based chemotherapy. Nivolumab was not associated with significantly longer PFS than chemotherapy among patients with previously untreated stage IV or recurrent NSCLC with a PD-L1 expression level of 5% or more, the primary endpoint of the trial. Furthermore, no differences were observed in OS between groups, and no advantage was seen in the PD-L1 ≥ 50% subgroup (HR for PFS 1.07 and 0.90 for OS) [64]. However, an exploratory analysis evaluating the tumor mutation burden (TMB) with whole exome sequencing (WES), performed in a subgroup of patients (58% of the randomized patients), showed that nivolumab was associated with higher ORR (47% vs. 28%) and longer PFS (9.7 vs. 5.8 months; HR 0.62) in patients with high TMB (≥243 mutations). No correlation between TMB and PD-L1 expression level was observed. Interestingly, the subgroup of patients with both high TMB and strong PD-L1 expression identified the subgroup of patients with higher response rate (75%) than those with only one of these factors (32% among patients with a high TMB only and 34% among those with a PD-L1 ≥ 50% only) or neither factor (16%) [65].

In contrast, the KEYNOTE-042 met its primary endpoints, demonstrating a statistically significant OS benefit in patients with a TPS ≥50% (HR 0.69, p = 0.0003), ≥20% (HR 0.77; p = 0.0020), and ≥ 1% (HR 0.81; p = 0.0018) leading to the FDA approval of pembrolizumab in treatment-naïve EGFR/ALK wild-type NSCLC patients with a TPS ≥ 1% [20]. However, this decision raised some concerns as pembrolizumab monotherapy may not represent the best treatment strategy for patients with tumor PD-L1 expression of 1–49%, as survival curves cross approximately 7 months after treatment initiation, with chemotherapy performing better than pembrolizumab during the first 6 months from randomization. These results suggest that a substantial number of patients progress rapidly and die within the first 6 months of treatment without obtaining any meaningful benefit from immunotherapy, and therefore other therapeutic strategies might be preferable in this subgroup of patients [21], especially in light of the positive results of chemo-immunotherapy trials in the first line.

Multiple randomized phase III trials have investigated the efficacy and safety (Table 3) of different chemo-immunotherapy trials.

Table 3 Safety of chemo-immunotherapy regimens in first-line NSCLC
Full size table

Phase III KEYNOTE-189 trial evaluated the use of pembrolizumab in association with platinum-pemetrexed chemotherapy in patients of non-squamous EGFR/ALK wild-type NSCLC, regardless of PD-L1 expression [66]. The trial met the two primary endpoints, demonstrating a statistically significant improvement in terms of both OS and PFS with the combination, as assessed by blinded, independent central radiologic review. First-line pembrolizumab plus chemotherapy demonstrate substantially improved OS and PFS in metastatic non-squamous NSCLC, regardless of PD-L1 expression or liver/brain metastases, with acceptable safety profile [66, 67]. Pembrolizumab plus platinum-pemetrexed was associated with a median OS of 22.0 months vs. 10.6 months with chemotherapy alone (HR 0.60) with a 3-year OS almost doubled (31.3% vs. 17.4%). Median PFS was longer for the experimental arm (9.0 vs. 4.9 months; HR 0.50), with a 3-year PFS rate of 11.8% vs.1.3% [68]. The PFS/OS benefit was seen across all the PD-L1 subgroups, with the strong PD-L1-positive subgroup benefitting more from the addition of pembrolizumab. Patients who completed the planned 35 cycles of treatment (2 years) were associated with durable responses and were most still alive at the 4-year follow-up (79.6% OS rate after 2 years from treatment completion) [68].

In a similar study design, pembrolizumab in combination with carboplatin plus paclitaxel/nab-paclitaxel demonstrated a PFS/OS as compared with chemotherapy in squamous NSCLC (KEYNOTE-407). The study met its two primary endpoints, demonstrating a statistically significant advantage for chemo-immunotherapy in terms of both OS (15.9 vs. 11.3 months, HR 0.64; P < 0.001) and PFS (6.4 vs. 4.8 months, HR 0.56; P < 0.001) [69]. Similar to the KEYNOTE-189 trial, the addition of pembrolizumab to chemotherapy was associated with survival benefit across all the PD-L1 subgroups, including among PD-L1-negative (TPS <1%) tumors [70]. At a 3-year follow-up, pembrolizumab plus carboplatin and paclitaxel/nab-paclitaxel continued to provide OS and PFS benefit vs. placebo plus chemotherapy (median OS 17.2 vs. 11.6 months with a 3-year OS rate of 29.7% vs. 18.2%; median PFS 8.0 vs. 5.1 months with a 3-year PFS rate of 16.1% vs. 6.5%). Among patients who completed the planned 35 cycles of treatment, durable responses were seen with a 1-year OS rate from completion of pembrolizumab of 96% [71].

In the IMpower150 was tested the addition of atezolizumab to bevacizumab plus chemotherapy as first-line treatment for metastatic non-squamous NSCLC, regardless of PD-L1 expression. In contrast with other chemo-immunotherapy trials, patients with known EGFR or ALK aberrations were included in the study but were excluded from the ITT population. Patients were randomly assigned, in a 1:1:1 ratio, to receive atezolizumab plus carboplatin plus paclitaxel (ACP group), atezolizumab plus bevacizumab plus carboplatin plus paclitaxel (ABCP group), or bevacizumab plus carboplatin plus paclitaxel (BCP group) [72].

The two primary endpoints were PFS both among patients in the ITT WT population and among patients in the WT population who had high expression of an effector T-cell (Teff) gene signature in the tumor (Teff-high WT population) and overall survival in the WT population. ABCP was associated with longer PFS (8.3 versus 6.8 months, HR 0.62; P < 0.001) and longer OS (19.2 versus 14.7 months, HR 0.78; P = 0.02) as compared with BCP in the ITT population [72]. Interestingly, an exploratory analysis of the study showed that ABCP was associated with improved OS compared with BCP in patients with sensitizing EGFR mutations (HR 0.31) and in those with baseline liver metastases (HR 0.52). In contrast, no OS benefit was seen with ACP versus BCP in patients with sensitizing EGFR mutations (HR 0.90), in the ITT population (HR 0.85), or in patients with baseline liver metastases (HR 0.87) [73]. These data should be interpreted with cautions, given the low number of patients included in this analysis, but suggest a potential synergistic effect between bevacizumab and atezolizumab.

Another therapeutic strategy explored in treatment-naïve advanced NSCLC is the dual immune checkpoint blockage with PD-1 plus CTLA-4 inhibitors. Checkmate-227 (Part 1) trial was a randomized phase III study evaluating the role of nivolumab plus ipilimumab in either PD-L1-positive (≥1%) versus chemotherapy or nivolumab (Part 1a) or PD-L1-negative (<1%) NSCLC patients versus chemotherapy +/− nivolumab (Part 1b). In Part 1a, nivolumab-ipilimumab was significantly associated with a longer median duration of OS as compared with chemotherapy alone (17.1 vs. 14.9 months; P = 0.007). The OS benefit was also observed in the Part 1b of the study (PD-L1 < 1%) with a median duration of 17.2 months with nivolumab plus ipilimumab and 12.2 months with chemotherapy. This combination was associated with similar serious adverse event (G3–4 AEs) rates compared with chemotherapy (32.8% with nivolumab plus ipilimumab and 36.0% with chemotherapy) [77]. At a 3-year follow-up, nivolumab-ipilimumab continues to provide a survival benefit as compared with chemotherapy in both PD-L1 ≥ 1% and PD-L1 < 1% with a similar 3-year OS rate (33% and 34%, respectively) [78]. This chemotherapy-free regimen was recently FDA-approved in PD-L1 ≥ 1%.

In contrast with nivolumab-ipilimumab, the dual blockage with durvalumab plus tremelimumab was not associated with significant survival benefit in the randomized phase III MYSTIC trial. The primary endpoints, assessed in patients with PD-L1 ≥ 25%, were OS for durvalumab vs. chemotherapy and OS and PFS for durvalumab plus tremelimumab vs. chemotherapy. The study did not meet its primary endpoints with no statistically significant improvement in terms of OS with durvalumab vs. chemotherapy (HR 0.76, p = 0.04) or OS/PFS with durvalumab plus tremelimumab vs. chemotherapy in patients PD-L1-positive tumors (HR 0.85 and 1.05, respectively) [79]. However, this combination was associated with OS improvement in patients with high blood TMB (≥20 mutations per megabase), assessed with the 500-gene plasma NGS platform GuardantOMNI [80].

Whether dual PD-1/CLTA-4 blockage is superior to PD-1 inhibition alone in patients with PD-L1 TPS ≥50% is still debated. The randomized, double-blind, phase III trial, KEYNOTE-598, addressed this issue and compared pembrolizumab plus ipilimumab vs. pembrolizumab alone. The primary endpoints were OS and PFS. The trial failed to demonstrate a survival benefit in terms of both OS (21.4 months for pembrolizumab-ipilimumab vs. 21.9 months for pembrolizumab-placebo; HR 1.08, p = 0.74) and PFS (8.2 months for pembrolizumab-ipilimumab vs. 8.4 months for pembrolizumab-placebo; HR1.06, p = 0.72). Differences in grades 3–5 AEs occurred in 62.4% vs. 50.2% and resulted in death in 13.1% versus 7.5%. Despite the study being early stopped due to futility by the external data and safety monitoring committee, it provides evidence that the addition of an anti-CTLA-4 inhibitor to pembrolizumab in PD-L1-strong positive NSCLC patients does not improve the efficacy but also worsens the toxicity profile [81].

To increase the disease control during the first few weeks of immunotherapy, another therapeutic strategy recently investigated is the addition of a limited course (two cycles) of a platinum-based chemotherapy to the dual checkpoint blockage. In the CheckMate-9LA, patients were randomly assigned (1:1) to nivolumab (360 mg intravenously every 3 weeks) plus ipilimumab (1 mg/kg intravenously every 6 weeks) combined with histology-based, platinum-doublet chemotherapy (intravenously every 3 weeks for two cycles, experimental group) or chemotherapy alone (every 3 weeks for four cycles, control group). Randomization was stratified by tumor histology, sex, and PD-L1 expression. The primary endpoint was OS in all randomly assigned patients [82]. The experimental group was associated with a significantly longer OS than control group (14.1 vs. 10.7 months, HR 0.69; p = 0.00065) at the preplanned interim analysis. In contrast with the CheckMate-227 study, the two OS curves early separated, suggesting that the addition of a short course of chemotherapy might overcome the limits of chemotherapy-free regimens that might be associated with a lower disease control in the first 3 months of treatment. No differences were observed across all PD-L1 TPS subgroup. This regimen was associated with an increased incidence of serious AEs as compared with chemotherapy alone (30% vs. 18%), although treatment-related deaths were similar in both groups (2%) [82]. Recently, the results of an exploratory analysis of the study, analyzing the role of tissue and blood TMB (tTMB and bTMB), were presented. Collectively, 64% and 73% of all randomized patients had tTMB (FoundationOne CDx assay) and bTMB (GuardantOMNI) evaluable samples, respectively. Similar to the CheckMate-227, the OS benefit with nivolumab-ipilimumab plus chemotherapy was observed regardless of TMB status with higher tTMB and bTMB associated with greater ORR and PFS benefit but similar OS outcomes. Collectively, these results support the use of nivolumab-ipilimumab plus two cycles of chemotherapy as first-line treatment option for patients with advanced NSCLC regardless of PD-L1 expression, TMB status, or their combination [83].

In summary, we have an arsenal of options when choosing first-line treatment for patients with metastatic lung cancer. The choice of the scheme will depend on a series of factors to take into account, considering that patients in everyday practice often do not always resemble the group selected and suitable for clinical trials.

5 ICIs and SCLC

Small cell lung cancer (SCLC) accounts for ~15% of all lung cancers and ~ 30,000 deaths in the USA annually, owing to the elusive pathophysiology of the disease, the poor prognosis of patients, and the minimal improvement in the effectiveness of therapies over the past decades. By the time that small cell lung cancer (SCLC) is diagnosed, nearly two-thirds of patients already have extensive stage disease (ES-SCLC) [84, 85].

ES-SCLC has a poor prognosis and a 5-year survival rate of <7% [84, 86]. For more than 20 years, the standard of care for ES-SCLC was platinum chemotherapy, which is associated with high initial response rates but a median survival of only 10 months. These findings highlight an unmet need for first-line (1 L) treatment of ES-SCLC [84, 86, 87]. In recent years, PD-L1/PD-1 inhibitors have demonstrated improved outcomes in patients with ES-SCLC.

Recent studies have shown that the efficacy of immunotherapy is related to a high tumor mutation burden (TMB), high genomic instability, and high immunogenicity in tumor cells. Some studies have shown that SCLC may have some advantages in immunotherapy.

PD-L1 expression in >1% of tumor cells is present in only a minority (~20%) of SCLC specimens [87, 88]. High counts of tumor-infiltrating lymphocytes (TILs) have been associated with better prognosis in SCLC in the pre-immunotherapy era [89]. Indeed, the presence of suppressive FOXP3+ regulatory T cells has been associated with a better prognosis in patients with LS-SCLC (HR 0.37; P = 0.013), and the presence of CD45RO+ memory T cells in brain metastases from ED-SCLC has been correlated with prolonged OS (11 vs. 5 months; P = 0.007) [90, 91]. However, data from studies designed to investigate the presence or absence of alternative, potentially clinically important immune checkpoints in SCLC, such as LAG3, TIM3, TIGIT, OX40, and ICOS, are currently unavailable. A better understanding of the immune microenvironment is an important area of unmet need in the immunobiology of SCLC. To better understand the ES-SCLC treatment, it can be divided into first line, maintenance, and second or more lines.

6 First-Line Therapy

The first immune checkpoint inhibitor evaluated in SCLC was the CTLA-4 inhibitor ipilimumab, following the promising results of a randomized phase II study [92]. In a phase III, placebo-controlled randomized trial, ipilimumab was evaluated in combination with platinum-etoposide with a phased schedule (two cycles of chemotherapy followed by two cycles of ipilimumab plus chemotherapy and then two additional cycles of ipilimumab) vs. chemotherapy alone in patients with ES-SCLC. Addition of ipilimumab to chemotherapy did not prolong OS versus chemotherapy alone in patients with newly diagnosed ES-SCLC (13.4 vs. 12.4 months, HR 0.91; p = 0.25) and was associated with higher serious AEs and discontinuation rates due to treatment-related AEs [93].

In the IMpower 133 trial, the efficacy of atezolizumab in combination with carboplatin-etoposide was assessed in patients with ES-SCLC. Patients were randomized to receive four 21-day cycles of carboplatin-etoposide plus atezolizumab or placebo and then maintenance atezolizumab or placebo until unacceptable toxicity, disease progression, or loss of clinical benefit. The study met its two primary endpoints (investigator-assessed PFS and OS) [94]. The addition of atezolizumab was associated with a significantly longer median OS (12.3 vs. 10.3 months, HR 0.76; p = 0.154) compared with chemotherapy alone with an 18-month OS of 34% vs. 21%. The survival benefit was seen regardless of PD-L1 expression or bTMB status [95]. Atezolizumab was the first ICI approved in first-line ES-SCLC, and this trial was the first randomized phase III study reporting a survival benefit in this setting as compared with platinum/etoposide after three decades of inconsistent results.

A second PD-L1 inhibitor that demonstrated a survival benefit in first-line ES-SCLC was durvalumab in combination with cisplatin/carboplatin plus etoposide. The randomized phase III trial CASPIAN randomized 805 ES-SCLC patients to receive durvalumab/tremelimumab plus platinum/etoposide or durvalumab plus platinum/etoposide or platinum/etoposide alone. Primary endpoint was OS [96]. Durvalumab/tremelimumab plus platinum/etoposide failed to demonstrate a significant improvement in OS versus platinum/etoposide (10.4 vs. 10.5 months, HR 0.82; p = 0.045). In contrast, durvalumab plus platinum/etoposide showed a sustained improvement in OS versus platinum-etoposide (12.9 vs. 10.5 months, HR 0.75, p = 0.0032). The survival benefit observed with durvalumab plus platinum/etoposide versus platinum/etoposide consistently favored the combination across all prespecified patient subgroups, as well as post hoc subgroups defined by liver metastases at baseline [97]. The overall survival benefit observed with durvalumab plus platinum-etoposide in CASPIAN aligns with findings from the IMpower133 trial, adding a novel therapeutic option in the therapeutic armamentarium of ES-SCLC.

In contrast with the positive results of IMpower133 and CASPIAN, the randomized phase III trial KEYNOTE-604 failed to demonstrate a statistically significant OS benefit with the addition of pembrolizumab to platinum/etoposide. The study randomized 453 ES-SCLC patients to receive pembrolizumab plus platinum/etoposide for 4 cycles followed by pembrolizumab for up to 35 cycles vs. platinum/etoposide for 4 cycles. Primary endpoints were PFS (by blinded central review) and OS with prespecified efficacy boundaries where one-sided P = 0.0048 for PFS and P = 0.0128 for OS. The addition of pembrolizumab significantly improved PFS (HR 0.75; P = 0.0023) and was associated with durable responses (12-month PFS rates: 13.6% vs. 3.1%). Albeit median OS was longer in the experimental arm, the significance threshold was not met (HR, 0.80; P = 0.0164). Twenty-four-month OS estimates were 22.5% and 11.2%, respectively. The PFS and OS HRs were similar between PD-L1-positive and PD-L1-negative tumors and regardless of the choice of platinum [98]. Albeit formally negative, the results of this trial along with those of IMpower133 and CASPIAN consolidate the use of platinum/etoposide plus an ICI as the novel standard of care for first-line ES-SCLC, which is associated with long-term clinical benefit in a small subgroup of patients. Inherited differences in the three trials and the enrollment of a poorer prognosis population in the KEYNOTE-604 trial might account for the survival differences seen in these studies. The identification of predictive biomarkers and the correlation with SCLC molecular subtypes might provide novel insights on patients benefitting most from this strategy.

7 Second-Line or Later Monotherapy

ICI monotherapy with nivolumab or pembrolizumab is FDA-approved for patients with ES-SCLC, independent of PD-L1 status, as a third or subsequent line of therapy.

The approval of nivolumab was based on the preliminary results of the phase II study CheckMate-032. Nivolumab monotherapy provided durable responses (median duration of response 17.9 months with 12-month and 18-month OS rates of 28.3% and 20.0%, respectively) and was well tolerated as a third- or later-line treatment for recurrent SCLC [99]. The randomized study compares nivolumab plus ipilimumab with nivolumab alone in pretreated ES-SCLC with ORR by blinded independent central review as primary endpoint. Although nivolumab plus ipilimumab was associated with higher ORR compared with nivolumab monotherapy (21.9% vs. 11.6%; p = 0.03), addition of ipilimumab did not prolong OS (median OS 4.7 vs. 5.7 months; 24-month OS rates 16.9% vs. 17.9%, respectively), at cost of higher-grade 3/4 treatment-related AEs (37.5% vs. 12.9%) [100].

The randomized phase III trial CheckMate-331 evaluated the efficacy of nivolumab in second line versus an active comparator (topotecan or amrubicin). The primary endpoint was OS. The trial failed to demonstrate a significant OS benefit with nivolumab compared with chemotherapy (7.5 vs. 8.4 months, HR 0.86; P = 0.11). No differences were noted between PD-L1-positive and PD-L1-negative tumors. Patients with baseline lactate dehydrogenase (LDH) under the upper limit of normal and those without baseline liver metastases seemed to benefit from nivolumab. A delayed separation in the survival curves at 12 months was observed, suggesting long-term benefit with nivolumab [101].

The FDA approval of pembrolizumab as third or subsequent line of therapy for ES-SCLC was based on the results of KEYNOTE-028 and KEYNOTE-158 trials. In the phase Ib KEYNOTE-028 trial, pembrolizumab was evaluated in PD-L1 selected patients with a tumor cell, immune infiltrate, and stromal summative PD-L1 combined positive score (CPS) ≥1%. The study included 24 patients (31.7% of all samples evaluated for PD-L1) with relapsed SCLC (12.5% receiving pembrolizumab as second line and 50% as third line). Pembrolizumab showed encouraging signals of activity in this setting with an ORR of 33%, a median PFS of 1.9 months (1-year PFS 23.8%), and a median OS of 9.7 months (1-year OS 37.7%) [102]. The KEYNOTE-158 was a phase II basket trial that enrolled 107 patients with relapsed SCLC (79% received pembrolizumab in the second-line or third-line setting), regardless of PD-L1 status (47% of patients had PD-L1-negative tumors). This study confirmed that promising antitumor activity (ORR 18.7%, median PFS 2.0 months, and OS 9.1 months) and durable responses (77% of the patients had a duration of response ≥9 months) were seen with pembrolizumab in pretreated SCLC, especially in patients with PD-L1-positive tumors (ORR 35.7% vs. 6.0% for PD-L1-positive and PD-L1-negative subgroups, respectively) [103]. A pooled analysis of these two trials, including 83 patients with recurrent SCLC, confirmed these findings. Pembrolizumab was associated with an ORR of 19.3% (2 complete responses and 14 partial responses) and a median duration of response not reached (61% of responders had responses lasting ≥18 months) [104].

In a phase II randomized clinical trial, the efficacy of atezolizumab monotherapy was compared with that of chemotherapy (with either topotecan or platinum rechallenge) in second-line SCLC, independent of PD-L1 expression. The study included 73 patients (49 in the atezolizumab arm and 24 in the chemotherapy arm), and 64% had platinum-sensitive disease (defined as disease progression ≥90 days after completion of induction chemotherapy). No significant differences were observed in median OS (9.5 vs. 8.7 months, HR 0.84; P = 0.60), and the median PFS was statistically inferior in patients who received atezolizumab (1.4 vs. 4.3 months; P = 0.004). ORRs were low in both groups (2.3% in the atezolizumab arm and 10% the chemotherapy arm) [105].

In summary, both nivolumab and atezolizumab have failed to improve OS compared with standard chemotherapy in RCTs involving patients with relapsed SCLC requiring second-line therapy. FDA approval of ICI monotherapy, with either nivolumab or pembrolizumab, has been granted only in the third-line or later setting based on ORRs of 10–30% in single-arm studies.

8 Activity of ICIs in Special Populations

Poor Performance Status (PS)

Patients with disease burden-determined poor performance status (PS) have generally poor prognosis. Evidence on first-line ICIs in PS ≥2 NSCLC with PD-L1 ≥ 50% expression is relatively scant, as this population is usually excluded from clinical trials. A recent retrospective multicenter study in a real-world setting addressed this issue. Among 153 patients included, the median PFS and OS were 2.4 (95% CI, 1.6–2.5) and 3.0 months (95% CI: 2.4–3.5), respectively. The 6-month PFS rate was 27% (95% CI, 21–35%). Patients with a PS 2 determined by comorbidities had significantly better results compared with PS 2 induced by disease burden (6-month PFS rate, 49% vs. 19%; median OS 11.8 vs. 2.8 months, respectively) [106]. Additional data are required to determine the best therapeutic approach for this poor prognosis subgroup of patients.

HIV/AIDS

Anti-PD-1/PD-L1 checkpoint inhibitors have been approved for a variety of cancers that occur with higher incidence in people with HIV, including lung cancer. However, HIV-infected patients were excluded from all registrative trials with ICIs in solid tumors, and therefore the evidence on safety and activity of these agents in this population are relatively poor and mostly derived from small case series or case reports [107].

A recent prospective study explored the safety of pembrolizumab immunotherapy in solid tumor patients with HIV infection (CD4 count greater than or equal to 100 cells/μL, antiretroviral therapy for 4 or more weeks, and an HIV viral load of less than 200 copies/mL were eligible) and showed that pembrolizumab has a similar irAE profile for people with HIV and advanced cancer who have suppressed antiretroviral treatment for HIV as seen in HIV-negative participants in published studies [108]. The proportion of serious events was similar to that previously described in patients receiving anti-PD-1 therapy for FDA-approved indications. Hypothyroidism was the most frequent immune-mediated event in 20% of the participants and was adequately controlled with standard treatment [108].

Evidence available to date in HIV-infected NSCLC suggests that single-agent PD-1/PD-L1 inhibitors can be used safely in this subgroup of patients with similar efficacy results observed in the overall NSCLC population. The results of ongoing clinical trials evaluating ICIs in HIV-infected patients with NSCLC (CHIVA-2/NCT03304093) and/or different solid tumors (NCT03094286, NCT02408861) will provide definitive conclusions in this setting [109].

Preexisting Autoimmune Disorders

The vast majority of clinical trials have excluded patients with significant preexisting autoimmune disorders (AID). However, AIDs are relatively common in clinical practice. Safety and efficacy of ICIs in patients with preexisting AIDs are largely unknown, and evidence available to date are mostly based on retrospective analyses.

In a large retrospective study including 751 patients, of whom 65.5% had an advanced NSCLC, 11.3% had preexisting AID, including both clinically active (17.6%) and inactive (82.4%) diseases. Patients with preexisting AID experienced higher incidence of immune-related adverse events (irAEs) of any grade compared with patients without AIDs (65.9% vs. 39.9%). However, no significant differences were observed regarding grade 3/4 irAEs. Interestingly, preexisting AIDs were not significantly associated with ICI efficacy [110]. Similarly, another retrospective multicenter study evaluating the safety of PD-1/PD-L1 inhibitors in NSCLC patients with preexisting AID showed that exacerbation of AID occurred in a minority of patients (23%). Thirty-eight percent of the patients experienced an irAE (74% G1/2, 26% G3/4), and 14% discontinued treatment because of irAEs [111].

Given the paucity of data, treatment with ICIs in this patient population should be evaluated with caution and after an accurate evaluation of the risk-benefit ratio within a multidisciplinary team [112].

Solid Organ Transplant

A scenario of the daily clinic surrounds the aspect that carries safety problems, solid organ transplant recipients (SOTR) who are routinely excluded from immunotherapy trials; therefore, there is limited data for these agents in this population. A first approximation to the information in relation to cancer patients and solid organ transplants was published in 2018 evaluating 26 solid organ recipients treated with ICIs. 3/7 had graft rejection with ipilimumab, 6/15 with PD-1 inhibitor, and 2/4 patients treated with sequential ipilimumab and PD-1 inhibitors. Graft rejection was observed in 7/10 patients who received prednisolone with or without cyclosporine and 4/16 patients treated with different immunosuppressive regimens (tacrolimus, everolimus, sirolimus, or mycophenolate mofetil), suggesting that solid organ transplant recipients considered for ICIs might need more intensive immunosuppressive therapy than prednisolone monotherapy. Tumor response was reported in 27% of patients treated with ipilimumab and in 32% in those treated with PD-1 inhibitors. Of the nine patients who obtained a CR/PR to ICIs, four patients were immunosuppressed with tacrolimus or sirolimus, while the other five were treated with prednisolone. This could suggest that immunosuppressive regimens containing tacrolimus or sirolimus can be continued when ICIs are administered to organ transplant recipients [113].

A recent meta-analysis of published data reported that 37% of the patients experienced organ rejection and 14% died as a result of graft rejection. Nivolumab was associated with the highest rejection rate (52.2%), followed by pembrolizumab (26.7%) and ipilimumab (25%). When analyzing transplant rejection by organ, the highest rejection rate was observed in patients with kidney transplants (40.1%), followed by liver (35%) and heart (20%) transplants, and 64% presented disease progression. In terms of efficacy, the response rate was highest for pembrolizumab (40%), followed by nivolumab (30%) and ipilimumab (25%) [114].

Integrating Special Populations

An example of an inclusive clinical trial seeking to shed light on a daily problem in healthcare practice was given by a prospective cohort investigation with ICIs in special populations with stage IV or recurrent NSCLC, and no known sensitizing EGFR or ALK alterations, regardless of PD-L1 expression. CheckMate-817 was a phase IIIb/IV trial initiated due to limited available data on safety and efficacy of immunotherapy in patients with advanced NSCLC with poor performance status (ECOG PS 2) or other comorbidities, such as kidney and renal disease, and HIV-infected. First-line flat-dose nivolumab plus weight-based ipilimumab showed a consistent safety profile in special populations with advanced NSCLC, including those with ECOG performance score 2. Patients with either high TMB or higher PD-L1 1 expression exhibited improved outcomes. The safety profile was similar between the special population and a reference cohort. The mean time to the appearance of adverse events was similar between the cohorts [115]. Similarly, the TAIL study evaluated the safety and activity of atezolizumab in pretreated NSCLC with ECOG PS 2, renal failure, or preexisting autoimmune disease [116]. 615 patients received atezolizumab. Serious AEs occurred in 7.8% of patients and irAEs in 8.3%. The median OS was 11.1 months (95% CI: 8.9, 12.9), the ORR was 11.1% (95% CI: 8.7, 13.8), and the median of DOR was 14.6 months (95% CI: 8.4, 15.4) [116]. Medium- and long-term safety profile data are awaited for this population.

Clinical trials in lung cancer with anti-PD-1/PD-L1 treatment have generally excluded patients with ECOG PS ≥ 2, organ transplantation, AIDS, chronic viral infection, or organ dysfunction. This group of patients does not have scientific support that the use of immunotherapy in these special populations is scarce and is derived mainly from case series or real-world experience. Therefore, cautions should be used in clinical practice when considering these agents in special patient populations, as the evidence available to date are low. The results of ongoing clinical trials in these peculiar clinical scenarios will provide definitive conclusions on the safety and efficacy of ICIs in these subgroups of patients.

9 Impact of Molecular Characterization in the era of Immunotherapy and Future Directions

In the era of personalized medicine, the increasing use of next-generation sequencing (NGS) in both tissue and plasma is rapidly expanding our knowledge on the molecular characteristics of lung tumors. The multitude of information that can be obtained with these techniques has considerably improved the therapeutic landscape of advanced NSCLC trough the identification of oncogene drivers exploitable with targeted therapies [117]. In the context of non-oncogene addicted tumors, the molecular characterization of the tumor might provide useful prognostic and predictive information, overcoming the limits of PD-L1 tumor expression.

There are reports regarding the increase in the acquisition of somatic mutations during tumorigenesis which is associated with the formation of neoantigens and the subsequent development of immunogenicity; therefore, it has been postulated that tumors with a higher number of somatic mutations could be more sensitive to blocking immune checkpoints. According to international consensus, tumor mutational burden (TMB) is defined as the total number of non-synonymous mutations per coding area of ​​a tumor genome and is calculated as mutations per DNA megabase (Mb). This emerging biomarker has been variably associated with ICI efficacy, although its clinical utility in clinical practice is unclear. On the one hand, TMB on either tissue (tTMB) or plasma (bTMB) has been clearly associated with improved efficacy with single-agent PD-1/PD-L1 inhibitors in exploratory analyses of large randomized studies in advanced NSCLC [65; 118, 119] and is a tumor-agnostic FDA-approved biomarker for pembrolizumab [120]. On the other hand, the predictive role of tTMB and bTMB is questioned when using chemo-immunotherapy combinations [121, 122].

Recently, the presence of concomitant mutations has been associated with inferior outcomes in NSCLC patients treated with single-agent ICIs.

Retrospective studies have shown that the presence of TP53 mutations without co-occurring STK11 or EGFR alterations (TP53-mut/STK11-EGFR-WT), independent of KRAS mutations, identifies a group of tumors with the highest CD8 T-cell density and PD-L1 expression that is associated with a prolonged PFS with single-agent ICIs. In contrast, STK11/LKB1 alterations are the most prevalent genomic driver of primary resistance to PD-1 axis inhibitors in KRAS-mutant lung adenocarcinoma [123, 124]. One of the possible explanations for these findings is that STK11, EGFR, or SMARC4 mutations are usually enriched among PD-L1-negative tumors, whereas TP53 mutations are more often seen in PD-L1-strong positive patients, as recently reported [125]. Collectively, these data suggest that concomitant mutations might influence ICI activity due to the presence of a tumor microenvironment less immunogenic (“cold tumors”), despite a higher TMB than wild-type tumors [126]. However, the debate on the prognostic/predictive role of these mutations is still open, and recent studies have shown that these mutations represent a poor prognostic factor rather a predictive factor that is independent of treatment received. Interestingly, the blood biomarker analysis of the MYSTIC trial showed that OS was shorter for patients with KEAP1-mutated or STK11-mutated NSCLC compared to wild-type patients irrespective of treatment received (durvalumab, durvalumab-tremelimumab, or chemotherapy) [127], suggesting that these mutations are likely a poor prognostic factor rather than a predictive factor to ICIs. Similar conclusions were recently reported in a large pan-cancer analysis evaluating the prognostic and predictive role of STK11 mutations. Across multiple solid tumors, STK11 alterations correlated with a poor prognosis regardless of therapy and were not associated with inferior immunotherapy outcome in the pan-cancer setting or in NSCLC. Furthermore, pan-cancer patients with co-altered STK11/KRAS did worse, regardless of treatment type [126]. In addition, the impact of these concomitant mutations among patients treated with chemo-immunotherapy combinations is still unclear, as initial retrospective studies reported that STK11 and KEAP1 genomic alterations are associated with shorter PFS with both platinum-pemetrexed-pembrolizumab chemo-immunotherapy and platinum-pemetrexed chemotherapy in non-squamous NSCLC and therefore represent adverse prognostic biomarkers, but the addition of pembrolizumab to platinum-pemetrexed does not result in prolonged PFS in PD-L1-positive STK11 and/or KEAP1-mutant non-squamous NSCLC [127], suggesting a potential negative predictive role [128, 129]. However, an exploratory analysis of the KEYNOTE-189 study did not confirm these findings, as pembrolizumab plus platinum/pemetrexed is associated with better outcomes than chemotherapy regardless of STK11 or KEAP1 mutational status [130].

The scenario of permanent evolution in the search for the best therapeutic strategy for a first line entails the need for robust predictive biomarkers to advanced NSCLC [131], which can potentially allow counseling of these patients who do not benefit from the use of ICI alone or in combination with chemotherapy or a combination of different checkpoint inhibitors.

The cornerstone of treatment for advanced NSCLC is focused on the search for biomarkers capable of predicting the response with an adequate safety profile. The biomarkers reported so far showed limitations in the capacity to effectively predict therapeutic efficacy of ICIs either alone or in different combinations. The use of ICIs is rapidly revolutionizing the therapeutic landscape of lung tumors, providing a significant improvement in the overall survival of these patients in multiple clinical settings. Long-term follow-up of registrative trials of these agents is constantly demonstrating long-term survivals in an unprecedented percentage of the patients, transforming an incurable disease into a chronic disease. The next step will be to extend the survival benefit to a higher percentage of patients, through the identification of novel predicting biomarkers and the introduction of more effective therapeutic strategies in tumors with less immunogenic microenvironment.