*4.3. Combination with Radiation Therapy*

In 2019, ICIs with high efficacy were reported for the treatment of unresectable stage III NSCLC. As noted above, in a PACIFIC study (NCT02125461) testing the efficacy of CRT followed by continued durvalumab treatment as consolidation, durvalumab treatment drastically improved the PFS and OS in comparison with the control group (PFS: 16.8 mo vs. 5.6 mo, HR 0.52, The 24-month overall survival rate: 66.3% vs. 55.6%, HR 0.68), leading to a major development of locally advanced-stage standard-of-care in the first 20 y [25]. In terms of the frequency of pneumonitis concerns associated with concomitant use of radiation therapy (RT) and ICIs, although pneumonitis was more common in the durvalumab group in all grades (13.1% vs. 7.7%), only a slight difference was observed between the two groups in ≥ grade-III NSCLC (4.4% vs. 3.8%, respectively), resulting in no apparent increase in the risk of serious pneumonitis [18]. The PACIFIC trial was designed to use durvalumab as a consolidation therapy in the first 42 d after the completion of CRT, whereas the ongoing PACIFIC2 trial is testing the efficacy of combination therapy with CRT and durvalumab, rather than sequential therapy (NCT03519971). The JCOG1508 studies have compared platinum-based chemotherapy + RT + durvalumab vs. platinum-based chemotherapy + RT → surgical resection + durvalumab in unresectable stage III NSCLC with N2 nodal involvement and have tested the effectiveness of durvalumab in the combined modality therapy including surgery.

#### **5. E**ff**ects on SCLC**

SCLC is a smoking-associated cancer type accounting for 10%–15% of all lung cancers. The median overall survival is 15–20 mo for limited-stage disease and 8–13 mo for extensive-stage disease, and the 5-y survival rate is 20–25% for limited-stage disease and 2% for extensive-stage disease among patients. In a phase-2 study on advanced SCLC conducted in 2011, the effectiveness of combination therapy with ipilimumab and chemotherapy was explored; however, the primary endpoint, i.e., OS prolongation, was not achieved [26]. Although ICIs for SCLC have yielded less encouraging results; the results of an IMpower133 trial published in 2019 led to the approval of CBDCA+etoposide+atezolizumab for untreated extensive-stage SCLC, as described previously [27]. The KEYNOTE-604 trial also compared pembrolizumab+EP vs. placebo+EP for untreated extensive-stage SCLC; this trial reported a significant prolongation in the PFS but not OS. In the ongoing CASPIAN trial, a three-arm comparative trial of durvalumab+tremelimumab+EP or durvalumab+EP vs. EP for untreated extensive-stage SCLC, among 268 patients receiving combination therapy with durvalumab and standard chemotherapy and 269 patients receiving standard chemotherapy alone, the median OS was significantly prolonged from 10.3 mo in the standard chemotherapy group to 13.0 mo in the combination therapy group [28]. Thus, future studies are required to develop more combination therapies with ICIs for SCLC.

Moreover, SCLC, unlike NSCLC and malignant melanoma, is generally characterized by a lower rate of PD-L1 expression [29]; however, the association between PD-L1 incidence and ICI efficacy has not been determined in SCLC. Nonetheless, TMBs are reportedly associated with the efficacy of ICIs in CheckMate026 trial [5]. The CheckMate032 trial compared the efficacy of nivolumab monotherapy with that of combination therapy with nivolumab+ipilimumab for previously treated SCLC, and the overall OS was 5.7 mo vs. 4.7 mo with no significant difference in efficacy between the combination therapy and monotherapy [30]. However, subgroup analysis revealed that combination therapy with nivolumab+ipilimumab displayed a higher efficacy than nivolumab monotherapy [30]. However, few studies have investigated the therapeutic utility of PD-L1 and the TMB in SCLC, warranting further validation in future studies.

#### **6. Biomarkers**

Although the efficacy of ICIs has been confirmed, the response rate to single agents is not as high as that of molecular-targeting agents, and the establishment of biomarkers to predict effective responses to ICIs remains challenging.

As a biomarker for therapeutic efficacy, PD-L1 has been recently used in actual clinical practice. The KEYNOTE-001 (NCT01295827) trial reported that pembrolizumab was more effective in decreasing the incidence of PD-L1 by ≥ 50%, by 1% to 49%, and by < 1% [31]. Particularly in the ≥ 50% group, patients with very high PD-L1 levels (≥ 90%) presented an even higher response rate than those with 50%–89% expression levels and presented prolonged PFS (objective response rate (ORR) 60.0% vs. 32.7%, PFS 14.5 mo vs. 4.1 mo, HR 0.50) [32]. Other studies have reported that PD-L1 upregulation, regardless of monotherapy or combination therapy, is associated with an increased efficacy of pembrolizumab [16,17,33]. Furthermore, the efficacy of ICIs other than pembrolizumab are also associated with PD-L1 upregulation [19,34], and generally, PD-L1 upregulation is associated with a higher ICI efficacy. Based on these results, we recommend using pembrolizumab monotherapy as the first-line therapy for PD-L1 positive (≥ 1%) advanced-stage NSCLC. It is also recommended to use PD-1/PD-L1 inhibitors for advanced-stage NSCLC in immune-checkpoint inhibitor-naïve patients as the second-line therapy.

However, the incidence and effects do not necessarily coincide, suggesting that tumor cell heterogeneity is one of the causes along with the potential involvement of host immune evasion mechanisms not mediated by PD-1/PD-L1 [35]. PD-L1 is often debated to be an incomplete biomarker, and several studies have as attempted to develop new biomarkers and to combine PD-L1 with other biomarkers.

While ICIs exert antitumor effects by activating immune cells, tumor infiltrating lymphocytes (TILs) are important in mediating these effects, along with PD-L1 [36]. T cells included in TILs may be enriched with clones specific for tumor antigens, but are suppressed by an immunosuppressive tumor microenvironment, and so on, and thus, they are believed to be incapable of exerting effective anti-tumor responses [37]. Results from a KEYNOTE-061 trial (NCT02370498) examining the usefulness of pembrolizumab as a second-line treatment for advanced gastric cancer reported that the effect of pembrolizumab may be predicted by the combined positive score, the number of PD-L1 positive cells among tumor cells, lymphocytes, and macrophages divided by the total number of tumor cells multiplied by 100 [38]. When the tumor microenvironment is subtyped into four types according to the presence or absence of PD-L1 expression and the presence or absence of TILs, TIL-positive/PD-L1 positive Type I and TIL-positive/PD-L1 negative Type IV are considered as "Hot tumors" in which anti-PD-1 antibodies are effective alone or in combination [36]. The usefulness of PD-L1 as a TIL biomarker has been reported in breast cancer [39], and similar studies are expected in future.

Other potential biomarkers include the total number of genetic mutations in cells, called tumor mutation burden (TMB). Mutated genes invariably yield mutated proteins, which are recognized as non-self by immune cells; hence, cells containing numerous mutated proteins are more susceptible to be attacked by immune cells, and cells with a high TMB are considered to display a more effective response to ICIs. In general, the TMB tends to be higher among smokers [40], and the relatively higher efficacy of ICI among smokers is speculated to result from the TMB [12,41]. In the 2017 CheckMate026 trial (NCT02041533), nivolumab was compared with platinum-based chemotherapy for first-line treatment of advanced-stage NSCLC with PD-L1 ≥ 5%, and nivolumab did not demonstrate superiority for the primary endpoint, PFS [5]. However, this study on exploratory TMB stratification analysis suggested that ICIs may be more effective in the high TMB group (TMB ≥ 243 nonsynonymous mutations) than in the low TMB group (TMB < 243 nonsynonymous mutations) (9.7 mo vs. 5.8 mo HR 0.62). A CheckMate227 trial (NCT02477826) (2019) comparing the efficacy of nivolumab+ipilimumab combination therapy vs. nivolumab monotherapy vs. platinum-based chemotherapy with TMB as a biomarker demonstrated the superiority of nivolumab+ipilimumab combination therapy to that of platinum-based chemotherapy; however, this tendency was stronger in the group with a high TMB (≥10 Mut/Mb) (global high: 23.0 mo vs. 16.4 mo, HR 0.68; low: 16.2 mo vs. 12.6 mo, HR 0.7, PD-L1 < 1% high: 20.4 mo vs. 11.2 mo, HR 0.51, low: 15.5 mo vs. 13.0 mo, HR 0.69) [23]. A CheckMate568 trial (NCT02659059) evaluated the safety and efficacy of concomitant nivolumab+ipilimumab combination therapy in the same year, reporting that the high-TMB group had a significantly prolonged PFS (7.1 mo

vs. 2.6 mo) [42]. Again, the results showed that TMBs and ORRs were predominantly associated in the group with PD-L1 < 1% (AUC 0.90) [42]. Consistent with the results of CheckMate568, the ORR with nivolumab plus ipilimumab in CheckMate227 was higher for tumors with PD-L1 > 1% compared with that for tumors with PD-L1 < 1% [21,42]. However, the relationship between the PD-L1 biomarker and efficacy of the combination of nivolumab and low-dose ipilimumab is complex, as in CheckMate227, there was a similar survival advantage for nivolumab and low-dose ipilimumab compared with that for standard chemotherapy in PD-L1-positive and PD-L1-negative tumors. Thus, the TMB (particularly in cases with PD-L1 < 1%) is reportedly associated with the efficacy of ICIs. At present, clinical trials considering TMBs as biomarkers are underway for atezolizumab (BFAST study: NCT03178552), and future studies are expected to yield clinically significant results.

The other reported potential benefits of these biomarkers include the prognostic nutritional index (PNI) and its association with the frequency of irAEs, neutrophil-to-lymphocyte ratio (NLR) [43], enterobacterial status [44,45], and early reduction in tumor markers [46].

Although the PNI was proposed as a predictor of surgical risk in the 1980s, it has been subsequently considered a useful marker to predict the efficacy of drugs to treat malignant diseases. Studies have reported that ICIs are more effective among patients with a high PNI, i.e., high nutritional status [47]. Further, numerous studies have reported that ICIs are more effective among patients with irAE [22,48], and that the management of adverse events is potentially important for the continuation of effective treatment.
