A Review of Immunotherapy in Non-Small-Cell Lung Cancer

Cancer immunotherapy in the form of immune checkpoint inhibitors has led to a dramatic increase in the survival of patients with lung cancer across all stages. Over the past decade, the field has experienced rapid maturation; however, several challenges continue to complicate patient management. This review aims to highlight the data that led to this dramatic shift in practice as well as to focus on key challenges. These include determining the optimal therapy duration, managing frail patients or those with brain metastases, addressing the challenges posed by immune-related adverse events, and defining the various patterns of clinical and radiological responses to immunotherapy.


Introduction
Lung cancer is the leading cause of death from cancer for both men and women worldwide.In 2022, 253,537 cases were diagnosed in Northern America, 30,000 of which were in Canada [1][2][3].About 85% of all lung cancers are non-small-cell lung cancer (NSCLC), divided into adenocarcinoma (40-50%) and squamous cell carcinoma (20-30%) subtypes [2].Of these, 60% are diagnosed at a locally advanced stage and 40% with de novo metastatic disease [2].In the absence of a driver mutation, the multimodality approach to the treatment of NSCLC relies on chemotherapy, immunotherapy, surgery, and radiation therapy.Notably, immunotherapy has been the cornerstone of the most significant advancements in lung cancer survival over recent decades.This review will provide an overview of the data supporting its use across all NSCLC stages, highlighting key challenges and unresolved questions.As we review the relevant clinical trials, their summary can be found in Table 1.OS significantly longer for pembro 2 mg/kg vs. doce (p = 0.0008) and for pembro 10 mg/kg vs. doce (p < 0.0001).
Stage IIIB or IV SCC and non-SCC NSCLC that had received 1-2 previous treatment.
The first randomized phase 3 study that reported results of a PD-L1-targeted therapy, with atezo resulting in a clinically relevant improvement of OS vs. docetaxel in previously treated NSCLSC regardless of PD-L1 expression or histology.
OS significantly longer with atezo in the ITT and PD-L1-expression populations.
In the ITT OS was improved with atezo compared with doce (median OS 13.8 months vs. 9.6 months; p = 0.0003).Patients in the PD-L1 low also had improved survival with atezo (median overall survival 12.6 months vs. 8.9 months.OS improvement was similar in patients with SCC in the atezo group or non-SCC.Pembro exhibited improved OS compared to chemo. OS was significantly longer in the pembro group than in the chemo group.The median survival were 20 months for pembro vs. 12.2 months for chemo.IMpower110 (NCT02409342) [8] Atezo vs. Chemo.
Metastatic non-SCC or SCC NSCLC that had not previously received chemo and that had PD-L1 expression on at least 1%.
Atezo resulted in longer OS than platinum-based chemo in patients with NSCLC with high PD-L1 expression, regardless of histologic type.
OS was longer by 7.1 months in the atezo group than in the chemo group (20.2 months vs. 13.1 months; HR, 0.59; p = 0.01).
Untreated stage IV or recurrent NSCLC and a PD-L1 expression of 1% or more.
Nivo was not associated with significantly longer PFS than chemo.
First-line treatment of Metastatic NSCLC.
The study did not meet primary end points of improved OS with durva vs. chemo or improved OS or PFS with durva plus tremelimumab vs. chemo in patients with ≥25% of tumor cells expressing PD-L1.

Exploratory analyses identified
a bTMB threshold of ≥20 mutations per megabase for optimal OS benefit with durvalumab plus tremelimumab.
Patients with platinum-treated advanced NSCLC.
Stage IV or recurrent NSCLC and a PD-L1 expression level of 1%.
First-line treatment with nivo plus ipi resulted in a longer duration of OS than did chemo in patients with NSCLC, independent of the PD-L1 expression level.
Pembro had an acceptable side-effect profile and showed antitumor activity.PD-L1 expression in at least 50% correlated with improved efficacy of pembro.
ORR was 19.4%, and the median duration of response was 12.5 months.The median duration of PFS was 3.7 months, and the median duration of OS was 12.0 months.
No head-to-head phase III trials compare single-agent immunotherapy to immunotherapy-chemotherapy in PDL-1-high patients.A recent cohort study analyzed 3086 advanced NSCLC patients receiving first-line therapy.Among them, 32% (978) received chemoimmunotherapy and 68% (2108) received immunotherapy alone.The study found no significant difference in OS between the groups, but chemoimmunotherapy of-fered an early survival advantage in both the overall cohort and the high PD-L1 subgroup (≥90%) [34].
Further evidence comes from a network meta-analysis comparing chemoimmunotherapy and different immunotherapy combinations across studies.This analysis included 12 eligible trials with a total of 7845 patients.In patients with high PD-L1 expression, chemoimmunotherapy improved objective response rate (ORR) and PFS compared to single-agent immunotherapy (not dual-agent immunotherapy).However, there was no difference in overall survival across treatment groups regardless of PD-L1 status [35].

Dual Immunotherapy without Chemotherapy
Another FDA-approved first-line option for metastatic NSCLC with PDL-1 ≥ 1% without EGFR/ALK alterations is nivolumab plus ipilimumab.OS benefit was positive in the phase III trial Checkmate 227, which comprised 1739 patients across all PD-L1 expression subgroups and compared ipilimumab/nivolumab vs. chemotherapy.Patients on ipilimumab/nivolumab had better long-term survival with a 4-year OS rate of 29%, compared to 18% for patients on chemotherapy [12].
Tumor mutational burden's (TMB) validity as a biomarker to select patients for dual ICI remains unclear.In an exploratory analysis of the MYSTIC trial, comparing durvalumab, durvalumab/tremelimumab, and chemotherapy in NSCLC, a TMB ≥ 20 mutations/Mb was significantly correlated with improved OS for durvalumab/tremelimumab vs. chemotherapy (21.9 months vs. 10.0 months, HR 0.49, 95% CI [0.32-0.74])and was not correlated with PDL-1 expression, suggesting its potential as a predictive biomarker for responsiveness to dual ICI [10].In Checkmate-227, progression-free survival (PFS) benefit was seen for TMB ≥ 10 mut/Mb vs. chemotherapy (7.2 months vs. 5.5 months, HR 0.58, 97.5% CI [0.41-0.81]),but no significant OS benefit between high-TMB patients and low-TMB patients [12].The phase III trial NEPTUNE explored TMB by comparing durvalumab/tremelimumab vs. platinum-based chemotherapy in metastatic NSCLC with TMB ≥ 20 mutations/Mb.However, NEPTUNE failed to meet its primary endpoint for OS and the full results have not yet been published [13].
An exploratory analysis from IMpower150, with a median follow-up of 39.3 months, hinted at an OS advantage when atezolizumab was added to the bevacizumab-carboplatinpaclitaxel regimen.This potential benefit (HR 0.60, 95% CI [0.31-1.14]),suggests that bevacizumab's counteraction of VEGF-mediated immunosuppression may enhance T-cell responsiveness [18].
Building on these insights, the phase 3 ATTLAS trial further evaluated the efficacy of atezolizumab in combination with bevacizumab and chemotherapy (ABCP) against the standard pemetrexed plus carboplatin or cisplatin in EGFR or ALK-mutated NSCLC who experienced progression following prior TKI therapy.The study's findings revealed an ORR improvement in the ABCP arm (69.5% vs. 41.9%) and a median PFS benefit (HR 0.62, 95% CI [0.45-0.86]),with the overall benefit increasing as PD-L1 expression rose.However, the OS was similar (HR 1.01, 95% CI [0.69-1.46])[19].
For advanced squamous NSCLC, the phase III trial KEYNOTE-407 examined firstline pembrolizumab combined with platinum-doublet chemotherapy versus platinumdoublet chemotherapy.The trial met its primary endpoints for PFS and OS.Statistically significant OS and PFS benefits were present in subgroup analyses for PDL-1 1-49%.There was also a positive trend in the PDL-1 < 1 subgroup [20].As a result of KEYNOTE 407, pembrolizumab with chemotherapy was approved in 2018 by the FDA for firstline treatment of metastatic squamous NSCLC [33].Meanwhile, atezolizumab is not approved for the squamous population; the phase III IMpower131 trial (atezolizumab plus chemotherapy vs. chemotherapy in metastatic squamous NSCLC) demonstrated statistically significant benefit for PFS, but not for OS [16].

Optimal Duration of ICI Therapy in Advanced Disease
Trials with ipilimumab and nivolumab (e.g., CheckMate-9LA), the KEYNOTE trials with pembrolizumab (e.g., 024, 042, 189, and 406) and the EMPOWER trials with cemiplimab (e.g, Lung 1 or 3) set a limit of two years, 35 cycles, and 108 weeks, respectively.There was no limit in the IMpower trials (e.g., 130 or 150) with atezolizumab or in the POSEIDON trial with durvalumab (although a maximum of 5 doses of tremelimumab was administered in combination with durvalumab).Checkmate 153, a phase III/IV real-world population study that evaluated the safety and efficacy of nivolumab in previously treated patients with stage IIIB or stage IV NSCLC, sought to answer the question of the optimal duration of ICI therapy [23].The protocol was amended to randomly assign patients to either 1 year of nivolumab therapy or until progression.A treatment with duration beyond one year conferred a significant survival advantage compared to an early stoppage (PFS 24.7 months vs. 9.4 months; hazard ratio [HR], 0.56 [95% CI, 0.37-0.84]).Despite limitations such as small sample size, the unplanned nature of the analysis, and insufficient power to demonstrate a statistically significant survival advantage, this research offers the only prospective data evaluating the relationship between therapy duration and survival outcomes.Further analysis, presented in a retrospective cohort study utilizing the Flatiron Health EHR database and involving 113 patients on fixed duration and 593 on indefinite ICI therapy, revealed no significant difference in survival between the two groups (univariable HR 1.26, 95% CI [0.77-2.08];multivariable HR 1.33, 95% CI [0.78-2.25]).These findings suggest that extending ICI treatment beyond two years may not provide a significant OS benefit, offering reassurance for those considering discontinuing therapy at the two-year threshold.
Although lacking randomized clinical trials, these analyses offers an unprecedented insight into an unanswered yet clinically significant issue.Some promising biomarkers such as minimal residual disease and measuring circulating tumor DNA are under study to guide the optimal duration of therapy [39][40][41].
The AEGEAN study presented the durvalumab combination as significantly extending event-free survival, with a HR of 0.68 (95% CI, 0.53 to 0.88, p = 0.004), translating to a 32% reduction in risk.The pathological complete response was achieved more frequently with durvalumab than with placebo (17.2% vs. 4.3%; 95% CI, 8.7 to 17.6; p < 0.001).The OS data are not yet published.Checkmate 77T's interim analysis highlighted neoadjuvant nivolumab plus chemotherapy's benefit, improving event-free survival with a HR of 0.58 (97.36%CI, 0.42-0.81),suggesting a 42% reduction in the risk of progression or death.The pCR rates were significantly higher in the nivolumab group (25.3% vs. 4.7%; odds ratio, 6.64; 95% CI, 3.40-12.97),demonstrating enhanced tumor response pre-surgery.The OS data are also awaited.
In the Neotorch trial, perioperative toripalimab markedly improved EFS with a HR of 0.40 (95% CI, 0.277-0.565;p < 0.001), indicating a 60% reduction in the risk of disease progression or death.The MPR and pCR rates were significantly elevated in the toripalimab arm (48.5% vs. 8.4% and 24.8% vs. 1.0%, respectively), underscoring the potent anti-tumor activity of the regimen.The OS data are not yet released.

The Effect of Radiation Therapy on Immunotherapy
Radiation therapy may have an impact on priming the immune system and enhancing immune response, as does chemotherapy.In the KEYNOTE-001 phase I trial, compared with the patients who received pembrolizumab only, the patients who received radiotherapy as a precondition prior to pembrolizumab showed a significantly longer PFS (4.4 vs. 2.1 months) and OS (10.7 vs. 5.3 months) [24].Similarly, in the randomized phase II PEMBRO-RT trial, patients who received stereotactic body radiation therapy (SBRT) to a single metastatic site within 7 days before initiation of immunotherapy showed an increase in the overall response rate (18% vs. 36%) and disease control rate (40% vs. 64%) at 12 weeks, as well as improved median PFS (1.9 vs. 6.6 months) and median OS (7.6 vs. 15.9 months), compared to those who did not [48].
The effect witnessed in PEMBRO-RT might be in part due to the abscopal effect.This phenomenon is characterized by tumor regression of untreated metastatic disease outside the target of local therapy [49,50].This effect is defined as a size reduction of 30% in a non-irradiated metastasis, irrespective of other lesions [51,52].This effect remains rare, and its incidence is not well understood.As one example, a study in Germany screened 168 patients with metastatic disease, including NSCLC patients, to evaluate the abscopal effect during simultaneous irradiation and anti-PD-1 therapy.The effect was observed in 29% of all patients.Three of the seven patients with NSCLC had an abscopal effect [53].
Other studies have shown improvement in patient outcomes with the addition of radiation therapy either before the initiation of immunotherapy or after its failure.These studies are small, non-randomized, and serve the role of hypothesis generation.Currently, numerous phase 1-3 trials of ICI with radiation therapy are underway, as reviewed elsewhere [54].

Immunotherapy for Frail Patients with NSCLC
Aging is thought to be associated with decreased responsiveness to new antigens, low-grade inflammation, defective T-cell memory responses, decreased naïve T cells, and increased susceptibility to autoimmunity [55].
A pooled analysis of KEYNOTE010/024/042 (pembrolizumab monotherapy) demonstrated that patients over 75 years old still had an OS benefit, particularly if PDL-1 ≥ 50% [7].This was also noted in another pooled analysis of phase III monotherapy trials spanning nivolumab, pembrolizumab, and atezolizumab [56].However, in subgroup analyses for trials combining pembrolizumab with chemotherapy, patients older than age 65 had less benefit, particularly with squamous NSCLC [57].In another pooled analysis of phase III data, patients 65 and older with advanced NSCLC, including those ≥75 years, seem to derive similar survival benefits from treatment with PD (L)-1 therapies as patients <65 years of age.Patients 75 and older enrolled in these trials appear to tolerate ICI and have a similar incidence of grade 3 or 4 irAEs compared to the subgroup of patients <65 years of age [56].Although the incidence of irAEs was similar, the duration of steroid treatment is often longer in this patient population, which can lead to worse geriatric outcomes including delirium and fractures [58].A limitation of these data is that older patients enrolled in clinical trials are often fitter than what is seen in real practice.

Patients with Brain Metastases
It is estimated that about one-third of newly NSCLC diagnosed patients may develop central nervous system (CNS) metastases throughout the disease, and 20% present with brain metastasis at the time of diagnosis [59].Patients with CNS disease have mostly been excluded from large registration trials.However, data from smaller phase 2 studies as well as real-world evidence suggest that ICI monotherapy can induce objective intracranial responses on the order of approximately 30%, with higher responses seen in PDL positive tumors [60][61][62][63].As an example, in a phase II trial of patients with NSCLC with untreated brain metastases 11 of 37 (30%) patients with PD-L1-positive tumors experienced objective CNS responses with pembrolizumab whereas that number was 0 out of 5 (0%) in those with PD-L1-negative tumors [62].Responses were mostly concordant with extracranial objective responses in nearly 80% of cases that were evaluable for both CNS and systemic responses.The concordance of responses was also seen in a retrospective study of 255 patients treated in an ICU, in which only 13% discordant cranial-extracranial responses were seen.Analysis of the nivolumab expanded-access program in Italy identified 372 squamous NSCLC patients, of whom 38 had asymptomatic brain metastases [63].The disease control rate in this population was 39%.Median PFS and OS in brain metastasis patients were 5.5 and 6.5 months, respectively.A preliminary presentation of pooled data of patients enrolled in one of five treatment studies with atezolizumab suggests that this drug also has some CNS activity [64].Furthermore, in one of the trials included in this analysis, among patients with baseline brain metastases, there was a nonsignificant trend toward reduction in the risk of developing new CNS lesions with atezolizumab compared with docetaxel (hazard ratio [HR] 0.42, CI [0.15-1.18]).Although initial data are promising, we await further studies to determine the optimal sequencing and combination therapies in patients with NSCLC and CNS metastasis.

Disease Response Assessment/Hyperprogressive Disease
Patients on ICI can experience different tumor responses than what was seen on chemotherapy.The guidelines of the Response Evaluation Criteria in Solid Tumours (RECIST) working group were updated for response assessment on immunotherapy (iRE-CIST) [65].Pseudoprogression, delayed responses and hyperprogressive disease are new types of responses incorporated into the iRECIST guidelines [65].
Pseudoprogression is extremely rare.About 5% of patients on ICI or less will experience a transient increase in tumor volume, resulting from an influx of inflammatory cells into tumor sites, which is then followed by a real response and decrease in tumor burden [66][67][68][69].Clinical outcomes in patients with pseugoprogression are better, with longer OS compared to patients with progressive disease (median OS not reached vs. 6.4 months; p = 0.001) [70,71].
Meanwhile, the definition of hyperprogressive disease remains controversial [72].Several definitions have been proposed, such as a tumor growth rate being at least twofold greater during ICI than immediately before the start of therapy, or a disease progression of >50% at the time of the first evaluation as compared to the onset of the treatment [73,74].Essentially, it is characterized by a fast and unexpected accelerated tumor progression, suspected to be induced by ICI [75].Hyperprogressive disease has been reported in 6-29% of advanced cancer patients [76].In NSCLC, the exact incidence is unknown but is thought to be low.Studies showed that patients who were treated with immunotherapy and developed hyperprogressive disease had a worse prognosis, with significantly inferior PFS and OS compared to patients on chemotherapy.As an example, a retrospective study analyzed 335 patients with advanced NSCLC treated with PD(L)-1 monotherapy [77].A total of 135 had PD by RECIST 1.1 criteria, 44 of which were found to have hyperprogression.This latter group had an inferior OS compared to patients without hyperprogression q (4.7 vs. 7.9 months).

Immune-Related Adverse Events in NSCLC
Patients with NSCLC being treated with ICI often experience the same common irAEs seen across all tumor types which often include dermatitis, colitis, and thyroid dysfunction.Though rarer, of particular interest in NSCLC is pneumonitis, as many patients have lower lung reserves due to pre-existing diseases like COPD or exposure to radiation therapy.In a retrospective study of 205 NSCLC patients on ICI, a higher rate of ICI pneumonitis of 19% was found, compared to the 3-5% previously reported in trials, and 53% of the pneumonitis case were grade 3-5 [78].Prior chest radiation has been disproven to be a risk factor for ICI pneumonitis [79].Time to ICI pneumonitis onset ranges from 1.5 to 127 weeks (median of 34 weeks) [80].Unlike diffuse bilateral pneumonitis induced by targeted therapy, ICI pneumonitis may present more focally, sometimes with unilateral radiographic changes [81].Patterns include chronic obstructive pneumonia-like, ground glass opacities, hypersensitivity type, and interstitial type [82].
Treatment of symptomatic pneumonitis is with corticosteroids (≥1 mg/kg), with a taper over 4-6 weeks.Pneumonitis is steroid responsive in >80% of cases [82].However, chronic pneumonitis has been described, defined as pneumonitis for 12 weeks or more despite corticosteroids and ICI suspension, and warrants further immunosuppression [83].In retrospective studies or case reports, success has been observed with the use of infliximab, intravenous immune globulin (IVIG), infliximab with IVIG, mycophenolate mofetil, and cyclophosphamide [80].
Overall, irAEs in NSCLC are associated with improved survival outcomes, particularly PFS [84,85].However, a multivariate analysis of two retrospective studies has shown ICI pneumonitis in NSCLC to be associated with worse outcomes, though they did not examine whether the death was due to pneumonitis, disease progression, or other causes [86,87].A retrospective study of NSCLC confirmed that pneumonitis, compared to other irAEs, was the most common toxicity associated with ICI suspension and steroid use (95% and 93% of cases, respectively) [88].ICI rechallenge after pneumonitis resolution is challenging, given the risk of recurrence.Data on safety are limited.Two retrospective studies on ICI pneumonitis had small subgroups of patients rechallenged with ICI (n = 12, n = 16) and demonstrated a 32% recurrence rate [82,89].
Another emerging area of interest is the association of the microbiome with irAEs.A higher abundance of Streptococcus, Paecalibacterium, and Stenotrophomonas was associated with developing grade 3-5 irAEs [90].In another study higher levels of Dorea, Butyricicococcus, and Eubacterium ventriosum were associated with grade 2-5 irAEs.In metagenomics profiling, alpha diversity was decreased in patients with grade 2-5 irAEs, with an overrepresentation of Dorea, Anaerosporobacter mobilis, Butyricicococcus, and Enterococcus faecium [91].
In future studies, it will be important to track microbiota changes during ICI therapy and at the time of irAEs.Additionally, it will be interesting to explore whether the respiratory microbiome may be correlated with pneumonitis.

Future Perspectives/Conclusions
ICI therapy has radically revolutionized the treatment of NSCLC.In the span of a decade, the field also matured quite rapidly.It is less likely that a novel ICI will bring a dramatic leap forward in the survival of cancer patients, as most targets have been extensively studied.Akin to what happened with the maturation of the field of chemotherapeutics, incremental gains will likely be achieved through optimization of the treatment schedule, finding the appropriate dose intensity, exploring biomarkers for efficacy and toxicity, and drawing solutions for the challenging patient populations (such as those with brain metastases or frail patients).A few fields, nevertheless, hold a lot of promise.The modulation of the microbiome or the blockade of inflammatory cytokines in conjunction with ICI still has the potential to yield impressive results.The results of those trials are eagerly awaited.Although it is not explored in this review, the basic science of immunotherapy is also a rapidly evolving field.Several key breakthroughs in the field of genomics, metagenomics, and special profiling have given us a glimpse into how the immune checkpoint blockade works in NSCLC.The clinical application of these findings will undoubtedly lead to further advancement in the field [92,93].

Table 1 .
Summary of trials.