A Molecular Epidemiological Analysis Of Programmed Cell Death Ligand-1 (PD-L1) Protein Expression, Mutations And Survival In Non-Small Cell Lung Cancer

Figures & data

Table 1 Patient Characteristics By PD-L1 Expression Level^a And By TMB

Covariate	Overall	PD-L1		P-Value^b	TMB^c		P-Value^b
N=136	PD-L1 <25% n=104 (76.4%)	PD-L1 ≥25% n=32 (23.5%)	<Median (n=43)	≥Median (n=44)
Age, mean (SD)	61.9 (9.3)	61.3 (6.4)	63.4 (8.5)	0.137	63.1 (8.6)	60.8 (9.7)	0.296
Gender, n (%)
Female	66 (48.5)	52 (50.0)	14 (43.8)	0.543	22 (51.2)	22 (50.0)	0.999
Male	70 (51.5)	52 (50.0)	18 (56.3)		21 (48.8)	22 (50.0)
Smoking History, n (%)
Smoker	113 (83.1)	87 (83.7)	26 (81.3)	1.000	36 (83.7)	37 (84.1)	0.999
Non-smoker	17 (12.5)	13 (12.5)	4 (12.5)		5 (11.6)	4 (9.1)
Missing/unknown	6 (4.4)	4 (3.8)	2 (6.3)		2 (4.7)	3 (6.8)
Race, n (%)
Asian	3 (2.2)	2 (1.9)	1 (3.1)	0.661	0 (0.0)	0 (0.0)	0.020
Black/African American	5 (3.7)	5 (4.8)	0 (0.0)		0 (0.0)	5 (11.4)
Other	3 (2.2)	2 (1.9)	1 (3.1)		1 (2.3)	2 (4.5)
Unknown	2 (1.5)	1 (1.0)	1 (3.1)		0 (0.0)	2 (4.5)
White	123 (90.4)	94 (90.4)	29 (90.6)		42 (97.7)	35 (79.5)
Ethnicity, n (%)
Hispanic/Latino	9 (6.6)	8 (7.7)	1 (3.1)	0.366	2 (4.7)	2 (4.5)	0.801
Non-Hispanic/Non-Latino	117 (86.0)	90 (86.5)	27 (84.4)		38 (88.4)	36 (81.8)
Missing/unknown	10 (7.3)	6 (5.8)	4 (12.5)		3 (7.0)	6 (13.6)
Histology, n (%)
Adenocarcinoma	97 (71.3)	77 (74.0)	20 (62.5)	0.422	32 (74.4)	33 (75)	0.999
Large cell carcinoma	3 (2.2)	1 (1.0)	2 (6.3)		1 (2.3)	1 (2.3)
Neuroendocrine	1 (0.7)	1 (1.0)	0 (0.0)		0 (0.0)	0 (0.0)
Non-small cell carcinoma	16 (11.8)	12 (11.5)	4 (12.5)		5 (11.6)	5 (11.4)
Squamous cell carcinoma	17 (12.5)	12 (11.5)	5 (15.6)		4 (9.3)	5 (11.4)
Missing/unknown	2 (1.5)	1 (1.0)	1 (3.1)		1 (2.3)	0 (0.0)
Tumor stage, n (%)^d
I	13 (9.6)	9 (8.7)	4 (12.5)	0.309	4 (9.3)	7 (15.9)	0.324
II	34 (25.0)	30 (28.8)	4 (12.5)		18 (41.9)	13 (29.5)
III	36 (26.5)	26 (25.0)	10 (31.3)		12 (27.9)	9 (20.5)
IV	53 (39.0)	39 (37.5)	14 (43.8)		9 (20.9)	15 (34.1)
ECOG PS, n (%)
0	72 (52.9)	57 (54.8)	15 (46.9)	0.667	29 (67.4)	26 (59.1)	0.132
1	45 (33.1)	32 (30.8)	13 (40.6)		11 (25.6)	15 (34.1)
2	9 (6.6)	7 (6.7)	2 (6.3)		3 (7.0)	0 (0.0)
3	2 (1.5)	2 (1.9)	0 (0.0)		0 (0.0)	1 (2.3)
Missing/unknown	8 (5.9)	6 (5.8)	2 (6.3)		0 (0.0)	2 (4.5)
Site radiotherapy, n (%)
None	41 (30.1)	31 (29.8)	10 (31.3)	0.795	10 (23.3)	17 (38.6)	0.139
Brain only	16 (11.8)	13 (12.5)	3 (9.4)		10 (23.3)	3 (6.8)
Chest/lung plus others	54 (39.7)	40 (38.5)	14 (43.8)		17 (39.5)	15 (34.1)
Others	25 (18.4)	20 (19.2)	5 (15.6)		6 (14.0)	9 (20.5)
Comorbidities, n (%)
Hypertension	33 (24.3)	23 (22.1)	10 (31.3)	0.281	12 (27.9)	12 (27.3)	0.831
None	54 (39.7)	45 (43.3)	9 (28.1)		14 (32.6)	17 (38.6)
Other	49 (36.0)	36 (34.6)	13 (40.6)		17 (39.5)	15 (34.1)
Site of metastatic disease, n (%)
None	29 (21.3)	20 (19.2)	9 (28.1)	0.158	13 (30.2)	10 (22.7)	0.018
Adrenal gland	11 (8.1)	6 (5.8)	5 (15.6)		1 (2.3)	9 (20.5)
Bone	33 (24.3)	29 (27.9)	4 (12.5)		15 (34.9)	6 (13.6)
Brain	20 (14.7)	16 (15.4)	4 (12.5)		7 (16.3)	7 (15.9)
Heart/abdomen/lymph	29 (21.3)	21 (20.2)	8 (25.0)		4 (9.3)	5 (11.4)
Lung/liver	14 (10.3)	12 (11.5)	2 (6.3)		3 (7.0)	7 (15.9)
Number of therapy lines, n (%)
0/1	9 (6.6)	6 (5.8)	3 (9.4)	0.465	2 (4.7)	4 (9.1)	0.058
2	45 (33.1)	37 (35.6)	8 (25.0)		18 (41.9)	14 (31.8)
3	43 (31.6)	30 (28.8)	13 (40.6)		17 (39.5)	10 (22.7)
4	39 (28.7)	31 (29.8)	8 (25)		6 (14.0)	16 (36.4)
EGFR, n (%)
Wild-type	81 (77.9)	58 (74.4)	23 (88.5)	0.178	31 (72.1)	38 (86.4)	0.117
Mutant	23 (22.1)	20 (25.6)	3 (11.5)		12 (27.9)	6 (13.6)
KRAS, n (%)
Wild-type	74 (73.3)	56 (74.7)	18 (69.2)	0.612	30 (69.8)	34 (77.3)	0.479
Mutant	27 (26.7)	19 (25.3)	8 (30.8)		13 (30.2)	10 (22.7)
ALK, n (%)
Negative	91 (98.9)	67 (98.5)	24 (100)	1.000	43 (100)	43 (97.7)	0.999
Positive	1 (1.1)	1 (1.5)	0 (0.0)		0 (0.0)	1 (2.3)

Notes: ^aNot all variables will sum to 136 because of missing or unknown data. ^bThe Pearson’s χ² test using Monte-Carlo estimation method was used to compare distributions of patient characteristics. ^cMedian of mutational burden was 325 on original scale (log10 scale=2.512). ^dStage at time of tissue collection. Bold P-values are statistically significant.

Abbreviations: ALK, anaplastic lymphoma kinase; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; KRAS, Kirsten rat sarcoma viral oncogene; PD-L1, programmed cell death ligand-1; SD, standard deviation; TMB, tumor mutational burden.

Table 2 OS And PFS By PD-L1 Expression Level^a

Covariate^b	PD-L1 <25%	PD-L1 ≥25%
OS
Total number of deaths	93	31
Median survival time, months	38.1	39.5
% survival at 6 months	0.98	1
% survival at 12 months	0.9	0.97
% survival at 18 months	0.87	0.81
% survival at 24 months	0.75	0.71
OS from initiation of second line
Total number of deaths	88	28
Median survival time, months	14.7	13
% survival at 6 months	0.72	0.72
% survival at 12 months	0.57	0.54
% survival at 18 months	0.39	0.4
% survival at 24 months	0.33	0.29
OS from initiation of third line
Total number of deaths	59	20
Median survival time, months	10	11.5
% survival at 6 months	0.62	0.71
% survival at 12 months	0.45	0.4
% survival at 18 months	0.35	0.2
% survival at 24 months	0.23	0.15
OS from initiation of fourth line
Total number of deaths	30	8
Median survival time, months	8.8	8.8
% survival at 6 months	0.68	0.63
% survival at 12 months	0.45	0.38
% survival at 18 months	0.38	0.38
% survival at 24 months	0.31	0.13
PFS
Total number of events	102	32
Median survival time, months	18.6	15.8
% survival at 6 months	0.83	0.81
% survival at 12 months	0.63	0.59
% survival at 18 months	0.5	0.5
% survival at 24 months	0.41	0.41
PFS from initiation of second line
Total number of events	90	29
Median survival time, months	5.3	3.6
% survival at 6 months	0.47	0.38
% survival at 12 months	0.21	0.24
% survival at 18 months	0.13	0.14
% survival at 24 months	0.11	0.07
PFS from initiation of third line
Total number of events	60	20
Median survival time, months	4.2	5.6
% survival at 6 months	0.39	0.51
% survival at 12 months	0.24	0.2
% survival at 18 months	0.08	0
% survival at 24 months	0.03	Not calculable

Notes: ^aOS was defined as the time from index date to death due to any cause or last follow-up, whichever was later. PFS was defined as the time from index date to the date the patient changed therapy due to first documented disease progression or death due to any cause, whichever occurred earliest. Those who did not have an event during follow-up were censored at the date of last follow-up. ^bThree index dates were utilized: i) date of initiation of third-line therapy (for the subset that received third line), ii) date of initiation of second-line therapy, and iii) date of diagnosis.

Abbreviations: OS, overall survival; PD-L1, programmed cell death ligand-1; PFS, progression-free survival.

Figure 1 OS and PFS by PD-L1 expression level. (A) OS by PD-L1 expression ≥25% showing no significant association between PD-L1 expression level and OS (log-rank test, P=0.968). (B) PFS by PD-L1 expression ≥25% showing no significant association between PD-L1 expression level and PFS (log-rank test, P=0.714). (C) OS by PD-L1 expression ≥25% showing no significant association between PD-L1 expression level and OS for patients receiving second-line therapies (log-rank test, P=0.523). (D) OS from third-line therapy by PD-L1 expression ≥25% showing no significant association between PD-L1 expression level and OS for patients receiving third-line systemic therapies (log-rank test, P=0.607). X-axis is in years. The analyses are censored at 5 years.

Abbreviations: OS, overall survival; PD-L1, programmed cell death ligand-1; PFS, progression-free survival.

Figure 2 Mutational load vs PD-L1 expression level. (A) Scatter plot of log TMB and the percentage of tumor cells with PD-L1 expression demonstrating correlation between mutational load and % PD-L1 staining. PD-L1 expression, as a continuous covariate, was weakly but positively associated with the number of somatic mutations (Pearson’s correlation coefficient=0.22, P=0.05). (B) Correlation between mutational load and % PD-L1 staining in the TCGA demonstrating similar results as in Figure 2A. (C) Mutational load by high vs low PD-L1 expression level demonstrating no significant difference observed when TMB was analyzed by PD-L1 expression ≥25%.

Abbreviations: PD-L1, programmed cell death ligand-1; TCGA, The Cancer Genome Atlas; TMB, tumor mutational burden.

Table 3 High-Frequency Mutations And PD-L1 Expression

Gene	No. Of Variants	Log-rank P-Values
P-Value^a	P-Value^b	OS From Primary Diagnosis	OS From Second-Line Therapy	OS From Third-Line Therapy	PFS From Primary Diagnosis	PFS From Second-Line Therapy	PFS From Third-Line Therapy
CFTR	33	1.000	0.538	0.150	0.799	0.889	0.072	0.484	0.084
CRIPAK	67	0.223	0.051	0.682	0.834	0.370	0.479	0.286	0.700
DDHD1	43	0.205	0.290	0.580	0.195	0.235	0.167	0.664	0.331
DEFB126	68	1.000	0.763	0.766	0.636	0.774	0.380	0.681	0.048
KRAS	22	1.000	0.789	0.742	0.573	0.473	0.337	0.028	0.687
NCOR2	86	1.000	0.826	0.344	0.959	0.497	0.195	0.775	0.201
RBMX	21	1.000	0.899	0.403	0.024	0.480	0.833	0.089	0.077
TP53	37	0.308	0.745	0.002^c	0.014	0.536	0.021	0.041	0.367

Notes: ^aFisher’s exact test P-value for PD-L1 positivity/negativity vs mutation/wild-type. ^bOne-way ANOVA test P-value for PD-L1 IHC level vs mutation/wild-type. ^cP<0.01. Bold P-values are statistically significant.

Abbreviations: ANOVA, analysis of variance; IHC, immunohistochemistry; KRAS, Kirsten rat sarcoma viral oncogene; OS, overall survival; PD-L1, programmed cell death ligand-1; PFS, progression-free survival; TP53, tumor protein p53.

Data Availability

Data underlying the findings described in this manuscript may be obtained in accordance with AstraZeneca’s data sharing policy described at https://astrazenecagrouptrials.pharmacm.com/ST/Submission/Disclosure.