Introduction: The tumor immune microenvironment (TIME) plays a key role in tumor progression and response to immune checkpoint inhibitors (ICIs). Additional analysis of the TIME across sites of disease may help inform future clinical trial designs and identify unique processes to exploit through novel combinatorial approaches. Here, we investigated differences in the TIME of primary and common metastatic sites in patients with NSCLC.
Methods: We retrospectively analyzed de-identified next-generation sequencing data from unique patients with metastatic NSCLC (n=6,534) in the Tempus Database. Tumors from primary lung (n=3,823), liver (n=937), CNS, brain and meninges (n=1,035) and bone (n=739) were sequenced with the Tempus xT DNA (648-gene panel) and xR whole transcriptome RNA assays. Demographics, clinical characteristics, PD-L1 expression, tumor mutational burden (TMB), and proportions of B, T (CD4+, CD8+), NK cells and macrophages were compared across all 4 sites. Pairwise comparisons were performed using Pearson’s Chi-squared, Kruskal-Wallis rank sum, and Dunn’s tests, and the FDR method was used to correct for multiple comparisons.
Results: The cohort (median age at metastatic diagnosis=67, IQR 61-75, Male=50%) comprised a diverse population (White, 79%, Black/African American, 12%, Asian, 4.0%, Other, 5.8%). The immune biomarker composition varied in primary lung tumors as compared to metastatic sites. Liver, CNS, and bone sites had lower percentage of total immune cells, CD8+ T cells, CD4+ T cells and B cells, and higher percentage of macrophages as compared to primary lung tumors (Table 1) (q-values <0.001). PD-L1 positive status was lower in liver, CNS, and bone compared to lung (q<0.001, q<0.01, q<0.01 respectively). When compared to lung, TMB-H status was modestly higher in patients with CNS metastasis (q<0.001), while TMB-high status was unchanged in patients with liver or bone metastasis (q>0.05).
Conclusions: In this large, real world analysis of patients with metastatic NSCLC, we observed a less immunogenic TIME in liver, CNS, and bone sites of disease compared to primary lung tumors. Our findings are hypothesis generating and further research is warranted to understand how the immune composition of metastatic sites alters the efficacy of ICI in patients with NSCLC.
Table 1. Immune cell quantification in primary and metastatic NSCLC
| Characteristics |
Lung (L)N=3,8231 |
Liver (LV)N=9371 |
CNSN=1,0351 |
Bone (B)N=7391 |
L vs. LVq-value2 |
L vs. CNS q-value2 |
L vs. B q-value2 |
| % IC of all cells |
21 (13, 28) |
14 (6, 21) |
17 (8, 26) |
13 (6, 21) |
<0.001 |
<0.001 |
<0.001 |
| % B cells of all IC |
18 (10, 28) |
7 (3, 14) |
11 (4, 24) |
6 (2, 14) |
<0.001 |
<0.001 |
<0.001 |
| % CD4 cells of all IC |
23 (17, 29) |
18 (12, 26) |
19 (11, 26) |
16 (10, 23) |
<0.001 |
<0.001 |
<0.001 |
| % CD8 cells of all IC |
7 (2.3, 10.9) |
1.6 (0, 8.8) |
4.5 (0, 10.9) |
0 (0, 7.9) |
<0.001 |
<0.001 |
<0.001 |
| % Macrophages of all IC |
35 (26, 45) |
53 (43, 64) |
42 (30, 57) |
57 (43, 68) |
<0.001 |
<0.001 |
<0.001 |
| % NK cells of all IC |
12 (7, 18) |
12 (8, 17) |
14 (9, 20) |
12 (7, 20) |
>0.05 |
<0.001 |
>0.05 |
| IC = immune cells
1Median (IQR)
2Dunn’s test with FDR correction |
VIEW THE PUBLICATION (EP.03G.01)