According to global cancer statistics 2020 reported by the International Agency for Research on Cancer, the incidence of lung cancer in China ranks first in males and second in females, while the mortality ranks first in both sexes. Recently, due to the emergence of targeted therapy and immune checkpoint inhibitors (ICIs), the survival of patients with lung cancer, specifically non-small cell lung cancer (NSCLC), has been strikingly prolonged. Lung cancer has the highest risk of brain metastasis (BM) among all solid carcinomas. Moreover, 10%–30% of patients with NSCLC have brain involvement during the course of their disease. The median time for patients with lung cancer to develop BM was 11 months. The emergence of BM has a significant impact on the clinical evaluation and oncologic treatment selection of patients. Chemotherapy is often ranked behind local therapy for BM, such as radiotherapy, due to its limited penetration through the blood-brain barrier/blood-tumor barrier (BBB/BTB) and the existence of specialized efflux pumps. With the application of lung cancer molecular testing, targeted therapy has been proven effective against NSCLC BM. The third-generation tyrosine kinase inhibitor osimertinib targeted epidermal growth factor receptor mutation, which is noted in 35%–50% of Asian patients with lung cancer, and the third-generation anaplastic lymphoma kinase (ALK) inhibitor lorlatinib targeted ALK rearrangement, which is reported in 4%–7% of patients with lung cancer, have better BBB penetration and greater intracranial objective response rate (ORR). ICI remains the most promising treatment for a substantial proportion of patients without druggable mutations. An increasing amount of clinical evidence has shown that ICI has a certain effect on lung cancer BM with good safety and tolerance. Patients with NSCLC BM are more often studied as a subgroup in clinical trials. In clinical trials of all phases, the intracranial ORR of ICI in patients with NSCLC BM ranges from 9% to 70% with great heterogeneity, mostly around 30%. Overall, the intracranial efficacy of ICI is shown to be durable and comparable to extracranial efficacy, facilitating research on the tumor immune microenvironment of BMs, which drives advances in understanding the ICI activity in BMs.
Therefore, this review aimed to summarize the evidence for ICI treatment efficacy for NSCLC BMs and compare the tumor immune microenvironments of BMs and primary lesions. Accordingly, we proposed the possible mechanisms of ICI treatment for NSCLC BMs based on existing evidence.
Survival Benefits of ICI for Patients with NSCLC BM in Phase III Studies
To date, it remained speculative whether ICI could provide a survival benefit to small cell lung cancer (SCLC) patients with BM. According to updated results from CASPIAN (Durvalumab ± Tremelimumab in Combination With Platinum Based Chemotherapy in Untreated Extensive-Stage Small Cell Lung Cancer [CASPIAN]), the overall survival (OS) benefit with the addition of durvalumab was shown in SCLC patients with BM albeit with high deviation (hazard ratio 0.79 [95% confidence interval, 0.44–1.41]). Nonetheless, no survival benefits were found in other randomized trials for BM subgroup with SCLC like Impower 133, Keynote 604, and Reck 2016. Overall, the magnitude of benefit with ICI in SCLC with BM did not mirror that shown in NSCLC with BM which may be explained by a more immunosuppressive environment and more rapid progression in SCLC. Meanwhile, substantial clinical research has demonstrated that ICI can provide survival benefits to patients with NSCLC BM. Patients with BMs are more often studied as a subgroup in large multicenter phase III trials of ICIs. Despite the diversity of inclusion criteria and regimens of these trials, these studies obtained the same conclusion that ICI can significantly improve the survival of patients with NSCLC BM.
ICI monotherapy or dual immunotherapy
The survival outcomes of large multicenter phase III clinical trials involving patients with NSCLC BM are summarized in Table 1. CheckMate 227 found that patients with NSCLC with brain involvement may benefit even more from ICI dual immunotherapy than patients without BM and whether they benefit from ICI regimen may not depend on the expression of programed death ligand-1 (PD-L1) on tumor cells. The same conclusion was reached by CheckMate 017/057 and OAK study which applied ICI monotherapy. As shown by CheckMate 017/057, nivolumab improved 5-year survival rate of patients with CNS metastases from 0% to 8%. All these studies have observed that after administration of ICI, the OS of patients with BM showed no significant difference from that of patients without BM. Another sizeable real-world study OAK confirmed that patients with BM receiving nivolumab have similar OS, progression-free survival (PFS), and ORR compared with the cohort without BM, which is consistent with the results of pembrolizumab as first-line treatment.
Table 1 -
Summary of the survival outcomes of NSCLC patients with BM in phase III trials of ICI.
||ICI arm vs. comparator arm
||OS (ICI arm vs. comparator) (months)
||HR for OS (95% CI)
||PFS (ICI arm vs. comparator) (months)
||HR for PFS (95% CI)
||Nivolumab + ipilimumab vs. chemotherapy
||16.8 vs. 13.4
||Nivolumab vs. docetaxel
||7.6 vs. 6.2
||Atezolizumab vs. docetaxel
||16.1 vs. 8.6
||Nivolumab + ipilimumab + chemotherapy vs. chemotherapy
||NR (12.3-NR) vs. 7.9 (5.0–10.7)
||9.8 (5.7–11.4) vs. 4.1 (2.8–5.0)
||Pembrolizumab + pemetrexed-platinum vs. pemetrexed-platinum
∗The number of patients participating in the trial. BM: Brain metastases; CI: Confidence interval; HR: Hazard ratio; ICI: Immune checkpoint inhibitor; NR: Not reached; NSCLC: Non-small cell lung cancer; OS: Overall survival; PFS: Progression-free survival.
ICI combined with chemotherapy
The improvement in OS and PFS was more profound in patients receiving ICI combined with chemotherapy for patients with BM, such as CheckMate 9LA and KEYNOTE-189 [Table 1], which may imply that the additive use of ICI can amplify the antitumor effect of chemotherapeutic drugs. Consistent with clinical trials of ICI monotherapy, the PD-L1 expression in the tumor tissue was not related to survival outcome or efficacy in ICI trials combined with chemotherapy.
ICI combined with radiotherapy
Presently, most studies on immunotherapy in combination with radiotherapy for NSCLC BM are retrospective. It is recognized that the addition of ICI to radiotherapy is safe and tolerable and provides better local control rate, PFS,[18,19] and OS[17-19]vs. radiotherapy alone, although whether the order and time interval between these two therapies affect the efficacy remains unclear. More research evidence has shown that synchronization of the two therapies or a time interval of <1 week can improve the remission rate and survival. Nevertheless, it seems that the delivery of programed death-1 (PD-1)/PD-L1 prior to radiation could not achieve equally good intracranial control or prolonged survival. A study claimed that concurrent use of ICI and stereotactic radiosurgery (SRS) does not lead to better local response or longer survival than concurrent use of chemotherapy and SRS. However, another study found that ICI plus baseline radiotherapy was superior to chemotherapy plus radiation when restricted to the PD-L1 ≥ 50% subgroup. Whether ICI performs better than chemotherapy in combination with radiation requires further study.
Based on the results of the abovementioned major multicenter phase III clinical trials, we concluded that ICI can effectively improve PFS and OS of patients with NSCLC BM, and the improvement is more significant when combined with chemotherapy. To determine whether to deliver immunotherapy, it is not necessary to consider PD-L1 expression levels in the tumor tissue, as patients with PD-L1 expression <1% can also benefit from immunotherapy. Moreover, for patients with symptomatic BMs, a meta-analysis showed that glucocorticoid administration has no effect on intracranial PFS, while it predicts shorter OS and systemic PFS. However, considering that most studies did not exclude other confounding factors, the effect of glucocorticoids on patients with NSCLC BMs treated with ICIs requires more in-depth research. The combination of ICI and anti-angiogenic agents also showed a potential synergistic effect with acceptable tolerability according to several phase I trials.
Comparison of the Efficacy of ICI Between Primary and Brain Metastatic Lesions
Accordingly, we confirmed that ICI can confer a survival benefit for patients with NSCLC BMs; however, the specific effects of ICI on BMs and primary lesions are still unclear. Several studies have evaluated the systemic response and intracranial response individually after ICI treatment, which more directly reflects the efficacy of ICI treatment on BMs, providing insights into the mechanism of ICI action on BMs.
The efficacy outcomes of these clinical studies are summarized in Table 2, and the responding patterns of intracerebral and paired extracerebral lesions are presented in Table 3. Because of the limitations associated with retrospective data collection, these studies included BMs under various circumstances while bringing more confounding factors. All these studies excluded patients who received ICI concurrent with local treatment, such as radiotherapy for BMs. The intracerebral ORRs of these studies in Table 2 are almost the same, which is approximately 30% and close to that of primary lesions. Two[27,28] of them decreased to approximately 10%, which may be attributed to the inclusion of active BMs. As shown in Table 2, intracerebral disease control rates (DCRs) generally reached 50%, similar to paired systemic DCRs. Therefore, it can be implied that the curative effects of ICI on BMs and extracerebral lesions are analogous.
Table 2 -
Efficacy outcomes of ICI in treatment for NSCLC patients with BM.
||Phase II trial
||PD-1/PD-L1 inhibitor monotherapy
||J Thorac Oncol
||Clin Lung Cancer
||Onco Targets Ther
||PD-1/PD-L1 inhibitor monotherapy
||Cancer Immunol Immunother
||Clin Transl Oncol
∗73 patients measured CNS response and 255 patients measured systemic response.
†11 patients measured CNS response and 126 patients measured systemic response.BM: Brain metastases; CNS: Central nervous system; DCR: Disease control rate; ICI: Immune checkpoint inhibitor; No.: Number; NSCLC: Non-small cell lung cancer; ORR: Objective response rate; PD-1: Programed death-1; PD-L1: Programed death ligand-1.
Table 3 -
Concordance of ICI response in BMs and paired extracranial lesions.
||No extracranial response
||No intracranial response
||No intracranial response
||Clin Lung Cancer
||PD-1/PD-L1 inhibitor monotherapy
||Clin Transl Oncol
No., the ordinal number of the trial in Table 2
. Data are presented as n
∗Response means CR or PR according to RECIST Version 1.1. BM: Brain metastases; CR: Complete response; ICI: Immune checkpoint inhibitors; PD-1: Programed death-1; PD-L1: Programed death ligand-1; PR: Partial response; RECIST: Response Evaluation Criteria in Solid Tumors.
Table 3 illustrates more clearly how people responded to ICI and whether the efficacy of ICI is consistent between BMs and its matched primary lesions. By searching Pubmed, we found that only six studies separately evaluated responses in the central nervous system (CNS) and lung including 87 patients. Discordant responses were shown in four studies and the rates of discordance ranged from 0% to 42.9%, which shows great divergence among different studies. The pooled concordant rate was 80.5% and the pooled discordant rate was 19.5%. Among the patients with divergent responses in CNS and lung, nine (10.3%) had intracranial lesion response only, while eight (9.2%) had extracranial lesion response only, and the ratios were nearly the same. Thus, most patients have BMs with the same sensitivity to ICI as primary lesions, while a small proportion of patients have BMs that are more sensitive to ICI, and another subset of patients of the same proportion have primary lesions that are more sensitive to ICI. Different responding patterns of ICI in NSCLC patients with BM shed light on the possible mechanisms of ICI for the treatment of NSCLC BMs.
Rationale for ICI Treatment in BMs
Because of the existence of the BBB, the brain has been considered an immune-privileged organ for a long time. Until the discovery of subdural cerebral lymphatic vessels in 2015, it was confirmed that immune cells can traffic between the neuroimmune system and peripheral immune system under physiological conditions, albeit rarely. When BMs are formed, disturbed BBB/BTB allows for activated immune cell entry into the CNS. Many clinical studies have demonstrated that BMs have been infiltrated by immune cells before receiving ICI, which emphasizes one possible mechanism that ICI may be capable of provoking immunity in situ with regard to its access to the CNS through the BBB/BTB. Conversely, expanded immune cells activated by ICI in the primary site may migrate into the CNS, exerting antitumor effects in the brain. A panoramic view of potential mechanisms underlying ICI treatment in BMs is visualized in [Figure 1].
ICI Intracranial Efficacy may be Dependent on BBB/BTB Penetration
BBB/BTB permeability of ICI
The BBB consists of three major components: endothelial cells, pericytes, and astrocytes. Vascular endothelial cells maintain barrier integrity by the tight junction (TJ) and specialized basement membrane forming connections with pericytes and astrocytic endfeet. The formation of BMs infers defects in paracellular connections between endothelial cells, and metastatic cancer cells can destroy TJ by secreting proteases that degrade the anchor molecule jam2. With the development of BMs, the BBB is partially remodeled by modifying endothelial cells from the transcriptome to a phenotype that deprives them of normal barrier function. Neoanginegeosis induced by tumor growth contributes to the formation of BTB, which replaces part of the BBB as a new barrier. BTB is characterized by the loss of TJs and inherent heterogeneous permeability, which disrupts intracranial homeostasis. Defects in BTB structural integrity are also embodied by aberrant pericyte distribution and decreased astrocyte endfeet. It can be implied that components of the BBB and interactions among them are destroyed in the development of BMs, which theoretically confers the ability of ICIs to penetrate brain metastatic lesions.
Drug delivery across the BBB/BTB relies on paracellular and transcytotic pathways, and antibodies as macromolecular agents prefer to penetrate the brain through transcytosis. Figueria et al found that as BMs develop, the endothelial cell TJ protein claudin-5 presented an irregular and discontinuous distribution, resulting in increased BBB permeability; conversely, a sustained increase was observed in vesicular transport protein caveolin-1 expression, which may enhance transcellular transmigration and thereby increase penetration of ICI into the BMs.
Pluim et al measured the concentration of PD-1 inhibitors in the cerebrospinal fluid (CSF) using enzyme-linked immunosorbent assay, and the serum/CSF ratios ranged from 52 to 299. The results indicated that the PD-1 inhibitor is capable of migration through the BBB/BTB into the CNS despite substantial interpatient variability. However, the detailed mechanisms underlying transportation are unclear, and more studies investigating the association between BBB permeability and efficacy are warranted.
Association between penetration of ICI across the BBB/BTB and efficacy
Mittapalli et al established a mathematical modeling approach to calculate interendothelial pore size on the BTB by two preclinical models. The pore diameter of the glioma vasculature was found to be large enough to allow monoclonal antibody (mAb) diffusion into tumors. The pore size of a BM tumor was ten-fold smaller than that of a primary brain tumor, indicating much less vascular permeability. Based on this observation, it can be inferred that PD-1 inhibitors have higher BBB penetration in glioblastoma than brain metastatic cancer. However, no existing clinical trials have validated that PD-1 inhibitors are beneficial for patients with glioblastoma. For example, in a Checkmate143 phase III randomized clinical trial, nivolumab monotherapy showed no improvement in OS (9.8 vs. 10.0 months) or ORR (7.8% vs. 23.1%) for recurrent glioblastoma compared with bevacizumab. PD-1/PD-L1 inhibitors generally fail to treat glioblastoma, which may be attributed to its relatively low tumor mutant burden and conversely significant immunosuppression. PD-1/PD-L1 inhibitors have been shown effective against BMs of other malignancies in clinical trials. Therefore, the immune microenvironment plays a crucial role in ICI activity. Considering the heterogeneity of cancer, more research is needed regarding the relationship between BBB permeability and intracranial efficacy of ICIs.
Preclinical studies on the ICI treatment mechanism in BMs
By using anti-PD-1 mAb in medulloblastoma-bearing mice, Pham et al found that, PD-1 blockade was present only in peripheral lymphocytes but not tumor-infiltrating lymphocytes (TILs) in the brain. Nevertheless, peripheral PD-1 mAb caused an influx or expansion of PD-1 negative T cells into tumors, which further extended the median survival of treated animals. This animal model suggested that PD-1 inhibitors could increase TILs within the brain tumor microenvironment by binding to peripheral lymphocytes without BBB/BTB penetration. Taggart et al established a melanoma tumor transplantation model and found that the intracranial efficacy of ICI could be achieved only when extracranial tumors were present. ICI induced peripheral expansion of effector T cells and recruitment of CD8+T and natural killer (NK) cells to the brain, the subtypes of which are required for ICI intracranial efficacy. For clinical animal models, intracranial efficacy relies on systematic immune responses when administered with ICIs. Given the structural differences in the BBB/BTB between rodent models and humans, no evidence has shown that ICI can migrate through mouse BBB/BTB so far, while the concentration of PD-1 antibody has been measured in human CSF. We can conclude that rodent models cannot interpret the panorama of mechanisms underlying the ICI effect on brain metastatic tumors, which highlights the need for more investigations on craniotomy specimens or even in situ.
Comparison of Immune Microenvironments Between Primary Lung Cancer and Paired BMs
Promising biomarkers of ICI efficacy in patients with lung cancer with CNS involvement
Biomarkers predicting better efficacy could provide clues to understand the mechanisms underlying ICI treatment for BM. At present, the most studied biomarker for predicting the efficacy of ICI on lung cancer is PD-L1 expression of tumor cells or immune cells, which makes mechanistic sense. At the beginning stage to explore the efficacy of ICI for lung cancer, the clinical trials enrolled only patients with high PD-L1 expression and consequently confirmed that ICI vs. chemotherapy could prolong both PFS and OS for lung cancer patients with high PD-L1 expression. As the expansion of enrollment, survival benefit was observed in NSCLC BM patients with negative PD-L1 expression who received ICI albeit with a much lower likelihood of benefit compared to patients with high PD-L1 expression. Furthermore, this benefit was more significant in the regimen of ICI plus chemotherapy. Consequently, the combination regimen received FDA approval for the expanded patient populations regardless of PD-L1 expression. Subsequent studies on PD-L1 expression as a biomarker of ICI efficacy seem to conclude that higher PD-L1 expression of either primary or metastatic lesions predicts longer OS and PFS,[50-52] meanwhile patients with negative expression of PD-L1 could still benefit from ICI treatment. Some evidence has shown the prognostic and predictive significance of TILs in ICI treatment. For NSCLC BMs, an abundance of total TILs, helper T cells, CD3+ TILs,[53,54] CD8+ TILs,[54-56] CD45RO+ TIL, and lower PD-1+ TILs[57,58] have been correlated with improved survival, and the prognostic value of CD4+ T cell density was inconclusive. There have also been studies showing that higher PD-L1 expression of either primary or metastatic lesions predicts longer OS and PFS,[50,51] excluding the intracranial control rate.
Association between PD-L1 expression and TIL in BM
PD-L1 and TIL in BM as the two most promising biomarkers of ICI efficacy on NSCLC BM have been studied for whether there is an association between them. Berghoff et al investigated TIL subsets in 116 BM specimens and found that TILs were prevalent in BM. High TIL density was most frequently observed in CD3+ TILs and least frequently in PD-1+ TILs. Of the BM specimens, 28.4% exhibited evident PD-L1 expression, whereas the expression level was unrelated to TIL density. A previous study showed that PD-L1-positive tumor cells and PD-L1-positive TILs belong to different subpopulations. Paired primary lung cancer and BM seem to have a different distribution of immune cells within and peripheral to the tumor despite similar expression of PD-L1, suggesting that BM may develop its own immune environment. In contrast, another study including 12 cases showed a positive correlation between the infiltration of PD-1+ TIL and PD-L1 expression on tumor cells in BM. For the wide use of P values to determine significance, some clinically meaningful but statistically insignificant associations may be omitted.
Comparison of PD-L1 expression between primary lung cancer and paired BMs
Some studies revealed that BM has statistically equivalent levels of PD-L1 expression compared with lung primary tumors.[56,57,61,62] Conversely, some other studies observed that brain metastatic tumors have higher[55,63] or lower[50,58] PD-L1 expression than lung primary lesions. Compared with BM from other solid tumors, BM originating from the lung seemed to have a higher expression of genes belonging to the PD-1 axis, indicating that lung cancer BM is a more favorable target for ICI therapy. Mansfield et al analyzed the immune microenvironment of 146 paired primary lung lesions and BMs and found agreement of PD-L1 expression by tumor cells in 86% of cases, most of the remaining cases with discordant expression tended to lose PD-L1 expression in BMs, and only individual cases showed the opposite.
Comparison of TILs between primary lung lesions and paired BM
Regarding TIL, total TIL,[65-67] CD8+ TIL,[55,56,63,67] and PD-1+ TIL within BM are significantly lower than paired primary lung tumors, which may be attributed to downregulated pathways, including dendritic cell maturation, leukocyte adhesion, and extravasation signaling in BMs. According to immune gene expression profiles, BMs tend to exhibit prevalent immune suppression compared to primary lesions.
It remains unclear whether genetic mutations and TILs in BM originate from primary lesions. Jiang et al observed that a median of 8.3% of genetic mutations were shared by paired lesions through phylogenetic analysis, suggesting that BM has genetically diverged from primary lesions at an early stage and experienced clonal evolution in parallel with primary tumors. Song et al investigated 33 cases and found that oligometastatic BM had a lower tumor mutation burden than primary tumors. However, Mansfield et al found a higher mutation burden and a lower number of unique T cell clones in BM than in primary lesions, which was thought to be correlated with ICI efficacy. Meanwhile, the overlap in T-cell receptor sequences of TILs between paired lesions was minor. Conversely, Kudo et al found that T cell clones were highly conserved in BM with primary tumors as a reference. Consistently, all these studies confirmed the expansion of unique T cells in BM. As an upstream signaling molecule of TIL, the incongruent expression of human leukocyte antigen (HLA) class one may explain the partial difference in T cell clones. Approximately a quarter of the patients exhibit disagreement in HLA class one positivity between paired lesions. Most of them were BM positive only, while a small proportion had positive primary tumors only. Because the median interval between acquisition of primary and metastatic specimens was approximately one year, temporal heterogeneity may bring about confounders.
Role of Innate Immunity in ICI's Effect on BMs
In addition to adaptive immunity, PD-1/PD-L1 inhibitors also regulate innate immunity. Fan et al found that pembrolizumab plus beta-glucan (an innate immune activator) could provoke innate immune activity in lung cancer BM in vitro, accompanied by tissue damage. According to the immunophenotypic investigation, the proportions of lymphocytes tended to decrease in BM compared to primary lung tumors; however, markers of NK cells and tumor-associated macrophages (TAM) and the ratio of M1 to M2 macrophages and NK to T cells were higher in BM. A study demonstrated that the presence of peritumor mononuclear cells, which is called the mononuclear ring in lung adenocarcinoma BM, indicated a high density of intratumor mononuclear cells and predicted better survival after BM surgery. Human TAMs can also express PD-1, and the majority of PD-1+ TAMs originate from circulating leukocytes. Among TAM, the M2 macrophage population expressed significantly more PD-1 than the M1 macrophage population. PD-1 expression on TAM attenuated its phagocytic potency against tumor cells, and PD-1/PD-L1 antagonism could in vitro relieve this type of function arrest and augment phagocytosis of PD-1+ TAM, further reducing tumor growth and prolonging survival. Genome-wide transcriptomic profiling showed that endothelial cells in the BM vasculature upregulated the expression of leukocyte adhesion molecules vs. primary tumors, which may result in increased NK cell infiltration in BM. In the melanoma microenvironment after administration of PD-1/PD-L1 inhibitor, high frequency of NK cells aroused activated dendritic cells, stimulating downstream cytotoxic T cells that drived better response to anti-PD-1 therapy, and increased OS of patients. It can be inferred that innate immunity also plays an important role in the treatment of lung cancer BM by PD-1/PD-L1 inhibitors.
Conclusions and Prospects
In this review, we summarized ICI efficacy on NSCLC BM in clinical trials and reviewed ICI responses in paired NSCLC and BM lesions. The interesting response pattern of NSCLC patients with BM which may indicate the existence of multiple mechanisms underlying the activity of ICI on NSCLC BM aroused our interest. Consequently, we proposed possible mechanisms and summarized corresponding evidence, and finally mentioned innate immunity which was rarely noted in the discussion of ICI treatment. In substantial phase III clinical trials, ICI has shown therapeutic effects in patients with NSCLC BM, and the pooled intracranial ORR is comparable to that of extracranial ORR. However, since most patients with BM were analyzed as subgroups in these studies, the number of cases was restricted. In the future, the emergence of large-scale multicenter clinical trials involving patients with lung cancer BM as the main cohort may provide stronger evidence for the use of ICI as a first-line treatment for NSCLC with BM. Currently, many real-world retrospective studies have explored the efficacy of ICIs in patients with lung cancer BM. However, due to the great heterogeneity of interstudy subjects, relatively lenient inclusion criteria, and great variance in efficacy outcomes, multicenter studies with larger sample sizes are needed. Most existing studies considered PFS and OS as outcomes, and a minority of them independently assessed primary and metastatic lesions. In terms of limited clinical data, the response rate showed no difference between intracranial and extracranial lesions after ICI treatment. However, the remission rate was not completely consistent in paired lesions, roughly 80% of cases were consistent, and 10% of cases had BMs or primary lesion response alone. This phenomenon indicates that the mechanism underlying ICI activity in BMs cannot be explained merely by the extension of peripheral immune cells into the brain.
Through investigations into the immune microenvironment of BMs and primary lesions, a higher density of TILs in tumors is associated with better survival, and PD-L1 expression is not necessarily related to prognosis. Nevertheless, the mechanism by which ICI affects the tumor immune microenvironment remains unclear. There is a dilemma that BM specimens can be acquired only before or after ICI treatment. Therefore, the ICI effect on BM could not be directly presented, and mechanisms underlying ICI treatment for BM can only be deduced indirectly through clinical evidence. Considering that BMs generally have immune cell infiltration before ICI treatment, ICI is capable of penetrating BBB/BTB despite its uncertain penetrability. In theory, ICIs can directly penetrate the CNS, causing activation of local immune cells and expansion of effector cells to reduce tumor cells. Conversely, according to preclinical studies and genetic profiles, peripheral effector T cells can infiltrate the CNS and exert antitumor effects. These two mechanisms may coexist and function synergistically, or one of the two dominates to some extent. Given the great heterogeneity of BMs and the extremely complex immune microenvironment, it is possible that a minor subset of patients has primary lesions with a more active immune status, while another small subset of patients has BMs with a more active immune status, which indicates a higher sensitivity to ICI treatment. On the other hand, ICI may arouse immune responses by multiple mechanisms not only by attaching to lymphocytes. Both innate immunity and adaptive immunity participate in ICI activity in primary and metastatic tumors. Currently, an increasing number of clinical and preclinical studies have provided insights into the mechanisms underlying ICI treatment for BM, which lays the foundation for clarifying the mechanisms and identifying efficacy biomarkers.
This study was supported by the Youth Program of the National Natural Science Foundation of China (to YX) (No. 82003309).
Conflicts of interest
1. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2021;71:209–249. doi: 10.3322/caac.21660.
2. Soffietti R, Ahluwalia M, Lin N, Rudà R. Management of brain metastases
according to molecular subtypes. Nat Rev Neurol 2020;16:557–574. doi: 10.1038/s41582-020-0391-x.
3. Achrol AS, Rennert RC, Anders C, Soffietti R, Ahluwalia MS, Nayak L, et al. Brain metastases
. Nat Rev Dis Primers 2019;5:5. doi: 10.1038/s41572-018-0055-y.
4. Lamba N, Wen PY, Aizer AA. Epidemiology of brain metastases
and leptomeningeal disease. Neuro Oncol 2021;23:1447–1456. doi: 10.1093/neuonc/noab101.
5. Vilarino N, Bruna J, Bosch-Barrera J, Valiente M, Nadal E. Immunotherapy in NSCLC patients with brain metastases
. Understanding brain tumor microenvironment and dissecting outcomes from immune checkpoint blockade in the clinic. Cancer Treat Rev 2020;89:102067. doi: 10.1016/j.ctrv.2020.102067.
6. El Rassy E, Botticella A, Kattan J, Le Péchoux C, Besse B, Hendriks L. Non-small cell lung cancer brain metastases
and the immune system: from brain metastases
development to treatment. Cancer Treat Rev 2018;68:69–79. doi: 10.1016/j.ctrv.2018.05.015.
7. Zhou F, Zhao W, Gong X, Ren S, Su C, Jiang T, et al. Immune-checkpoint inhibitors plus chemotherapy versus chemotherapy as first-line treatment for patients with extensive-stage small cell lung cancer. J Immunother Cancer 2020;8:e001300. doi: 10.1136/jitc-2020-001300.
8. Goldman JW, Dvorkin M, Chen Y, Reinmuth N, Hotta K, Trukhin D, et al. Durvalumab, with or without tremelimumab, plus platinum-etoposide versus platinum-etoposide alone in first-line treatment of extensive-stage small-cell lung cancer (CASPIAN): updated results from a randomised, controlled, open-label, phase 3 trial. Lancet Oncol 2021;22:51–65. doi: 10.1016/S1470-2045(20)30539-8.
9. Landre T, Chouahnia K, Des Guetz G, Duchemann B, Assié JB, Chouaïd C. First-line immune-checkpoint inhibitor plus chemotherapy versus chemotherapy alone for extensive-stage small-cell lung cancer: a meta-analysis. Ther Adv Med Oncol 2020;12:1758835920977137. doi: 10.1177/1758835920977137.
10. Hellmann MD, Paz-Ares L, Bernabe Caro R, Zurawski B, Kim SW, Carcereny Costa E, et al. Nivolumab plus ipilimumab in advanced non-small-cell lung cancer. N Engl J Med 2019;381:2020–2031. doi: 10.1056/NEJMoa1910231.
11. Borghaei H, Gettinger S, Vokes EE, Chow LQM, Burgio MA, de Castro Carpeno J, et al. Five-year outcomes from the randomized, phase III trials CheckMate 017 and 057: nivolumab versus docetaxel in previously treated non-small-cell lung cancer. J Clin Oncol 2021;39:723–733. doi: 10.1200/JCO.20.01605.
12. Fehrenbacher L, von Pawel J, Park K, Rittmeyer A, Gandara DR, Ponce Aix S, et al. Updated efficacy analysis including secondary population results for OAK: a randomized phase III study of atezolizumab versus docetaxel in patients with previously treated advanced non-small cell lung cancer
. J Thorac Oncol 2018;13:1156–1170. doi: 10.1016/j.jtho.2018.04.039.
13. Crinò L, Bronte G, Bidoli P, Cravero P, Minenza E, Cortesi E, et al. Nivolumab and brain metastases
in patients with advanced non-squamous non-small cell lung cancer
. Lung Cancer 2019;129:35–40. doi: 10.1016/j.lungcan.2018.12.025.
14. Alessi JV, Ricciuti B, Jimenez-Aguilar E, Hong F, Wei Z, Nishino M, et al. Outcomes to first-line pembrolizumab in patients with PD-L1-high (≥ = 50%) non-small cell lung cancer
and a poor performance status. J Immunother Cancer 2020;8:e001007. doi: 10.1136/jitc-2020-001007.
15. Paz-Ares L, Ciuleanu TE, Cobo M, Schenker M, Zurawski B, Menezes J, et al. First-line nivolumab plus ipilimumab combined with two cycles of chemotherapy in patients with non-small-cell lung cancer (CheckMate 9LA): an international, randomised, open-label, phase 3 trial. Lancet Oncol 2021;22:198–211. doi: 10.1016/S1470-2045(20)30641-0.
16. Rodriguez-Abreu D, Powell SF, Hochmair MJ, Gadgeel S, Esteban E, Felip E, et al. Pemetrexed plus platinum with or without pembrolizumab in patients with previously untreated metastatic nonsquamous NSCLC: protocol-specified final analysis from KEYNOTE-189. Ann Oncol 2021;32:881–895. doi: 10.1016/j.annonc.2021.04.008.
17. Enright TL, Witt JS, Burr AR, Yadav P, Leal T, Baschnagel AM. Combined immunotherapy and stereotactic radiotherapy improves neurologic outcomes in patients with non-small-cell lung cancer brain metastases
. Clin Lung Cancer 2021;22:110–119. doi: 10.1016/j.cllc.2020.10.014.
18. Shaverdian N, Lisberg AE, Bornazyan K, Veruttipong D, Goldman JW, Formenti SC, et al. Previous radiotherapy and the clinical activity and toxicity of pembrolizumab in the treatment of non-small-cell lung cancer: a secondary analysis of the KEYNOTE-001 phase 1 trial. Lancet Oncol 2017;18:895–903. doi: 10.1016/S1470-2045(17)30380-7.
19. Rounis K, Skribek M, Makrakis D, De Petris L, Agelaki S, Ekman S, et al. Correlation of clinical parameters with intracranial outcome in non-small cell lung cancer
patients with brain metastases
treated with Pd-1/Pd-L1 inhibitors as monotherapy. Cancers (Basel) 2021;13:1562. doi: 10.3390/cancers13071562.
20. Qian JM, Martin AM, Martin K, Hammoudeh L, Catalano PJ, Hodi FS, et al. Response rate and local recurrence after concurrent immune checkpoint therapy and radiotherapy for non-small cell lung cancer
and melanoma brain metastases
. Cancer 2020;126:5274–5282. doi: 10.1002/cncr.33196.
21. Scoccianti S, Olmetto E, Pinzi V, Osti MF, Di Franco R, Caini S, et al. Immunotherapy in association with stereotactic radiotherapy for non-small cell lung cancer brain metastases
: results from a multicentric retrospective study on behalf of AIRO. Neuro Oncol 2021;23:1750–1764. doi: 10.1093/neuonc/noab129.
22. Ahmed KA, Kim S, Arrington J, Naghavi AO, Dilling TJ, Creelan BC, et al. Outcomes targeting the PD-1/PD-L1 axis in conjunction with stereotactic radiation for patients with non-small cell lung cancer brain metastases
. J Neurooncol 2017;133:331–338. doi: 10.1007/s11060-017-2437-5.
23. Singh C, Qian JM, Yu JB, Chiang VL. Local tumor response and survival outcomes after combined stereotactic radiosurgery and immunotherapy in non-small cell lung cancer
with brain metastases
. J Neurosurg 2019;132:512–517. doi: 10.3171/2018.10.JNS181371.
24. Lau SCM, Poletes C, Le LW, Mackay KM, Fares AF, Bradbury PA, et al. Durability of CNS disease control in NSCLC patients with brain metastases
treated with immune checkpoint inhibitors plus cranial radiotherapy. Lung Cancer 2021;156:76–81. doi: 10.1016/j.lungcan.2021.04.006.
25. Jessurun CAC, Hulsbergen AFC, de Wit AE, Tewarie IA, Snijders TJ, Verhoeff JJC, et al. The combined use of steroids and immune checkpoint inhibitors in brain metastasis patients: a systematic review and meta-analysis. Neuro Oncol 2021;23:1261–1272. doi: 10.1093/neuonc/noab046.
26. Fang L, Zhao W, Ye B, Chen D. Combination of immune checkpoint inhibitors and anti-angiogenic agents in brain metastases
from non-small cell lung cancer
. Front Oncol 2021;11:670313. doi: 10.3389/fonc.2021.670313.
27. Tozuka T, Kitazono S, Sakamoto H, Yoshida H, Amino Y, Uematsu S, et al. Poor efficacy of anti-programmed cell death-1/ligand 1 monotherapy for non-small cell lung cancer
patients with active brain metastases
. Thorac Cancer 2020;11:2465–2472. doi: 10.1111/1759-7714.13557.
28. Gauvain C, Vauleon E, Chouaid C, Le Rhun E, Jabot L, Scherpereel A, et al. Intracerebral efficacy and tolerance of nivolumab in non-small-cell lung cancer patients with brain metastases
. Lung Cancer 2018;116:62–66. doi: 10.1016/j.lungcan.2017.12.008.
29. Goldberg SB, Schalper KA, Gettinger SN, Mahajan A, Herbst RS, Chiang AC, et al. Pembrolizumab for management of patients with NSCLC and brain metastases
: long-term results and biomarker analysis from a non-randomised, open-label, phase 2 trial. Lancet Oncol 2020;21:655–663. doi: 10.1016/S1470-2045(20)30111-X.
30. Skribek M, Rounis K, Makrakis D, Agelaki S, Mavroudis D, De Petris L, et al. Outcome of patients with NSCLC and brain metastases
treated with immune checkpoint inhibitors in a ‘real-life’ setting. Cancers (Basel) 2020;12:3707. doi: 10.3390/cancers12123707.
31. Hendriks LEL, Henon C, Auclin E, Mezquita L, Ferrara R, Audigier-Valette C, et al. Outcome of patients with non-small cell lung cancer
and brain metastases
treated with checkpoint inhibitors. J Thorac Oncol 2019;14:1244–1254. doi: 10.1016/j.jtho.2019.02.009.
32. Sun L, Davis CW, Hwang WT, Jeffries S, Sulyok LF, Marmarelis ME, et al. Outcomes in patients with non-small-cell lung cancer with brain metastases
treated with pembrolizumab-based therapy. Clin Lung Cancer 2021;22:58–66. doi: 10.1016/j.cllc.2020.10.017.
33. Sun C, Zhou F, Li X, Zhao C, Li W, Li J, et al. PD-1/PD-L1 inhibitor combined with chemotherapy can improve the survival of non-small cell lung cancer
patients with brain metastases
. Onco Targets Ther 2020;13:12777–12786. doi: 10.2147/OTT.S286600.
34. Zhang G, Cheng R, Wang H, Zhang Y, Yan X, Li P, et al. Comparable outcomes of nivolumab in patients with advanced NSCLC presenting with or without brain metastases
: a retrospective cohort study. Cancer Immunol Immunother 2020;69:399–405. doi: 10.1007/s00262-019-02462-1.
35. Dudnik E, Yust-Katz S, Nechushtan H, Goldstein DA, Zer A, Flex D, et al. Intracranial response to nivolumab in NSCLC patients with untreated or progressing CNS metastases. Lung Cancer 2016;98:114–117. doi: 10.1016/j.lungcan.2016.05.031.
36. Metro G, Gili A, Signorelli D, De Toma A, Garaffa M, Galetta D, et al. Upfront pembrolizumab as an effective treatment start in patients with PD-L1 ≥ 50% non-oncogene addicted non-small cell lung cancer
and asymptomatic brain metastases
: an exploratory analysis. Clin Transl Oncol 2021;23:1818–1826. doi: 10.1007/s12094-021-02588-8.
37. Louveau A, Smirnov I, Keyes TJ, Eccles JD, Rouhani SJ, Peske JD, et al. Structural and functional features of central nervous system lymphatic vessels. Nature 2015;523:337–341. doi: 10.1038/nature14432.
38. Wrobel JK, Toborek M. Blood-brain barrier remodeling during brain metastasis Formation. Mol Med 2016;22:32–40. doi: 10.2119/molmed.2015.00207.
39. Arvanitis CD, Ferraro GB, Jain RK. The blood-brain barrier and blood-tumour barrier in brain tumours and metastases. Nat Rev Cancer 2020;20:26–41. doi: 10.1038/s41568-019-0205-x.
40. Steeg PS. The blood-tumour barrier in cancer biology and therapy. Nat Rev Clin Oncol 2021;18:696–714. doi: 10.1038/s41571-021-00529-6.
41. Figueira I, Galego S, Custodio-Santos T, Vicente R, Molnar K, Hasko J, et al. Picturing breast cancer brain metastasis development to unravel molecular players and cellular crosstalk. Cancers (Basel) 2021;13:910. doi: 10.3390/cancers13040910.
42. Pluim D, Ros W, van Bussel MTJ, Brandsma D, Beijnen JH, Schellens JHM. Enzyme linked immunosorbent assay for the quantification of nivolumab and pembrolizumab in human serum and cerebrospinal fluid. J Pharm Biomed Anal 2019;164:128–134. doi: 10.1016/j.jpba.2018.10.025.
43. Mittapalli RK, Adkins CE, Bohn KA, Mohammad AS, Lockman JA, Lockman PR. Quantitative fluorescence microscopy measures vascular pore size in primary and metastatic brain tumors. Cancer Res 2017;77:238–246. doi: 10.1158/0008-5472.CAN-16-1711.
44. Lim M, Xia Y, Bettegowda C, Weller M. Current state of immunotherapy for glioblastoma. Nat Rev Clin Oncol 2018;15:422–442. doi: 10.1038/s41571-018-0003-5.
45. Reardon DA, Brandes AA, Omuro A, Mulholland P, Lim M, Wick A, et al. Effect of nivolumab vs bevacizumab in patients with recurrent glioblastoma: the CheckMate 143 phase 3 randomized clinical trial. JAMA Oncol 2020;6:1003–1010. doi: 10.1001/jamaoncol.2020.1024.
46. Pham CD, Flores C, Yang C, Pinheiro EM, Yearley JH, Sayour EJ, et al. Differential immune microenvironments and response to immune checkpoint blockade among molecular subtypes of murine medulloblastoma. Clin Cancer Res 2016;22:582–595. doi: 10.1158/1078-0432.CCR-15-0713.
47. Taggart D, Andreou T, Scott KJ, Williams J, Rippaus N, Brownlie RJ, et al. Anti-PD-1/anti-CTLA-4 efficacy in melanoma brain metastases
depends on extracranial disease and augmentation of CD8(+) T cell trafficking. Proc Natl Acad Sci USA 2018;115:E1540–E1549. doi: 10.1073/pnas.1714089115.
48. Reck M, Rodriguez-Abreu D, Robinson AG, Hui R, Csoszi T, Fulop A, et al. Updated analysis of KEYNOTE-024: pembrolizumab versus platinum-based chemotherapy for advanced non-small-cell lung cancer with PD-L1 tumor proportion score of 50% or greater. J Clin Oncol 2019;37:537–546. doi: 10.1200/JCO.18.00149.
49. Memmott RM, Wolfe AR, Carbone DP, Williams TM. Predictors of response, progression-free survival, and overall survival in patients with lung cancer treated with immune checkpoint inhibitors. J Thorac Oncol 2021;16:1086–1098. doi: 10.1016/j.jtho.2021.03.017.
50. Hong L, Negrao MV, Dibaj SS, Chen R, Reuben A, Bohac JM, et al. Programmed death-ligand 1 heterogeneity and its impact on benefit from immune checkpoint inhibitors in NSCLC. J Thorac Oncol 2020;15:1449–1459. doi: 10.1016/j.jtho.2020.04.026.
51. Hulsbergen AFC, Mammi M, Nagtegaal SHJ, Lak AM, Kavouridis V, Smith TR, et al. Programmed death receptor ligand one expression may independently predict survival in patients with non-small cell lung carcinoma brain metastases
receiving immunotherapy. Int J Radiat Oncol Biol Phys 2020;108:258–267. doi: 10.1016/j.ijrobp.2020.04.018.
52. Negrao MV, Skoulidis F, Montesion M, Schulze K, Bara I, Shen V, et al. Oncogene-specific differences in tumor mutational burden, PD-L1 expression, and outcomes from immunotherapy in non-small cell lung cancer
. J Immunother Cancer 2021;9:e002891. doi: 10.1136/jitc-2021-002891.
53. Pocha K, Mock A, Rapp C, Dettling S, Warta R, Geisenberger C, et al. Surfactant expression defines an inflamed subtype of lung adenocarcinoma brain metastases
that correlates with prolonged survival. Clin Cancer Res 2020;26:2231–2243. doi: 10.1158/1078-0432.CCR-19-2184.
54. Berghoff AS, Fuchs E, Ricken G, Mlecnik B, Bindea G, Spanberger T, et al. Density of tumor-infiltrating lymphocytes correlates with extent of brain edema and overall survival time in patients with brain metastases
. Oncoimmunology 2016;5:e1057388. doi: 10.1080/2162402X.2015.1057388.
55. Zhou J, Gong Z, Jia Q, Wu Y, Yang ZZ, Zhu B. Programmed death ligand 1 expression and CD8(+) tumor-infiltrating lymphocyte density differences between paired primary and brain metastatic lesions in non-small cell lung cancer
. Biochem Biophys Res Commun 2018;498:751–757. doi: 10.1016/j.bbrc.2018.03.053.
56. Song Z, Yang L, Zhou Z, Li P, Wang W, Cheng G, et al. Genomic profiles and tumor immune microenvironment
of primary lung carcinoma and brain oligo-metastasis. Cell Death Dis 2021;12:106. doi: 10.1038/s41419-021-03410-7.
57. Kim R, Keam B, Kim S, Kim M, Kim SH, Kim JW, et al. Differences in tumor microenvironments between primary lung tumors and brain metastases
in lung cancer patients: therapeutic implications for immune checkpoint inhibitors. BMC Cancer 2019;19:19. doi: 10.1186/s12885-018-5214-8.
58. Song SG, Kim S, Koh J, Yim J, Han B, Kim YA, et al. Comparative analysis of the tumor immune-microenvironment of primary and brain metastases
of non-small-cell lung cancer reveals organ-specific and EGFR mutation-dependent unique immune landscape. Cancer Immunol Immunother 2021;70:2035–2048. doi: 10.1007/s00262-020-02840-0.
59. Hendry S, Salgado R, Gevaert T, Russell PA, John T, Thapa B, et al. Assessing tumor-infiltrating lymphocytes in solid tumors: a practical review for pathologists and proposal for a standardized method from the international immuno-oncology biomarkers working group: part 2: TILs in melanoma, gastrointestinal tract carcinomas, non-small cell lung carcinoma and mesothelioma, endometrial and ovarian carcinomas, squamous cell carcinoma of the head and neck, genitourinary carcinomas, and primary brain tumors. Adv Anat Pathol 2017;24:311–335. doi: 10.1097/PAP.0000000000000161.
60. Téglási V, Pipek O, Lózsa R, Berta K, Szüts D, Harkó T, et al. PD-L1 expression of lung cancer cells, unlike infiltrating immune cells, is stable and unaffected by therapy during brain metastasis. Clin Lung Cancer 2019;20:363–369. doi: 10.1016/j.cllc.2019.05.008.
61. Moutafi MK, Tao W, Huang R, Haberberger J, Alexander B, Ramkissoon S, et al. Comparison of programmed death-ligand 1 protein expression between primary and metastatic lesions in patients with lung cancer. J Immunother Cancer 2021;9:e002230. doi: 10.1136/jitc-2020-002230.
62. Lu BY, Gupta R, Aguirre-Ducler A, Gianino N, Wyatt H, Ribeiro M, et al. Spatially resolved analysis of the T cell immune contexture in lung cancer-associated brain metastases
. J Immunother Cancer 2021;9:e002684. doi: 10.1136/jitc-2021-002684.
63. Jiang T, Yan Y, Zhou K, Su C, Ren S, Li N, et al. Characterization of evolution trajectory and immune profiling of brain metastasis in lung adenocarcinoma. NPJ Precis Oncol 2021;5:6. doi: 10.1038/s41698-021-00151-w.
64. Jiang J, Wu L, Yuan F, Ji J, Lin X, Yang W, et al. Characterization of the immune microenvironment in brain metastases
from different solid tumors. Cancer Med 2020;9:2299–2308. doi: 10.1002/cam4.2905.
65. Mansfield AS, Aubry MC, Moser JC, Harrington SM, Dronca RS, Park SS, et al. Temporal and spatial discordance of programmed cell death-ligand 1 expression and lymphocyte tumor infiltration between paired primary lesions and brain metastases
in lung cancer. Ann Oncol 2016;27:1953–1958. doi: 10.1093/annonc/mdw289.
66. Mansfield AS, Ren H, Sutor S, Sarangi V, Nair A, Davila J, et al. Contraction of T cell richness in lung cancer brain metastases
. Sci Rep 2018;8:2171. doi: 10.1038/s41598-018-20622-8.
67. Kudo Y, Haymaker C, Zhang J, Reuben A, Duose DY, Fujimoto J, et al. Suppressed immune microenvironment and repertoire in brain metastases
from patients with resected non-small-cell lung cancer. Ann Oncol 2019;30:1521–1530. doi: 10.1093/annonc/mdz207.
68. Schaffenrath J, Wyss T, He L, Rushing EJ, Delorenzi M, Vasella F, et al. Blood-brain barrier alterations in human brain tumors revealed by genome-wide transcriptomic profiling. Neuro Oncol 2021;23:2095–2106. doi: 10.1093/neuonc/noab022.
69. Riaz N, Havel JJ, Makarov V, Desrichard A, Urba WJ, Sims JS, et al. Tumor and microenvironment evolution during immunotherapy with nivolumab. Cell 2017;171:934–949. doi: 10.1016/j.cell.2017.09.028.
70. Failing JJ, Aubry MC, Mansfield AS. Human leukocyte antigen expression in paired primary lung tumors and brain metastases
in non-small cell lung cancer
. Cancer Immunol Immunother 2021;70:215–219. doi: 10.1007/s00262-020-02677-7.
71. Fan T, Higashi RM, Song H, Daneshmandi S, Mahan AL, Purdom MS, et al. Innate immune activation by checkpoint inhibition in human patient-derived lung cancer tissues. Elife 2021;10:e69578. doi: 10.7554/eLife.69578.
72. Téglási V, Reiniger L, Fábián K, Pipek O, Csala I, Bagó AG, et al. Evaluating the significance of density, localization, and PD-1/PD-L1 immunopositivity of mononuclear cells in the clinical course of lung adenocarcinoma patients with brain metastasis. Neuro Oncol 2017;19:1058–1067. doi: 10.1093/neuonc/now309.
73. Gordon SR, Maute RL, Dulken BW, Hutter G, George BM, McCracken MN, et al. PD-1 expression by tumour-associated macrophages inhibits phagocytosis and tumour immunity. Nature 2017;545:495–499. doi: 10.1038/nature22396.
74. Barry KC, Hsu J, Broz ML, Cueto FJ, Binnewies M, Combes AJ, et al. A natural killer-dendritic cell axis defines checkpoint therapy-responsive tumor microenvironments. Nat Med 2018;24:1178–1191. doi: 10.1038/s41591-018-0085-8.