Matched Analysis of Detailed Peripheral Blood and Tumor Immune Microenvironment Profiles in Bladder Cancer

Figures & data

Figure 1. Data processing and cell distribution in tumor microenvironment and peripheral blood.

(A) Data processing schematic. DNA was extracted from tumor tissues and matched blood samples of 88 bladder cancer patients. After serial preprocessing steps, DNA methylation profiles were applied to estimate cell-type proportions in (B) tumor tissues and (C) blood using HiTIMED and FlowSorted.Blood.EPIC methods respectively.

NLR: Neutrophil-to-lymphocyte ratio.

Table 1. Characteristics of the study subjects.

	NMIBC (n = 75)	MIBC (n = 13)	All (n = 88)
Age
Median (Q1, Q3)	67 (56, 71)	69 (64, 76)	67 (57, 72)
Sex
Male	57 (76%)	8 (62%)	65 (74%)
Female	18 (24%)	5 (38%)	23 (26%)
Tumor grade
Low	56 (75%)	3 (23%)	59 (67%)
High	19 (25%)	10 (77%)	29 (33%)
Tumor stage
0a	61 (81%)	0 (0%)	61 (69%)
I	14 (19%)	0 (0%)	14 (16%)
II	0 (0%)	8 (62%)	8 (9%)
III	0 (0%)	2 (15%)	2 (2%)
IV	0 (0%)	3 (23%)	3 (4%)
Smoking status
Never	12 (16%)	2 (15%)	14 (16%)
Former	40 (53%)	6 (46%)	46 (52%)
Current	21 (28%)	5 (39%)	26 (30%)
Missing	2 (3%)	0 (0%)	2 (2%)
BCG immunotherapy
No	65 (87%)	13 (100%)	78 (89%)
Yes	10 (13%)	0 (0%)	10 (11%)

BCG: Bacillus Calmette–Guérin; NMIBC: Non-muscle-invasive bladder cancer; MIBC: Muscle-invasive bladder cancer.

Figure 2. Cell profiles of tumor microenvironment in non-muscle-invasive and muscle-invasive bladder cancer patients.

Differences in cell-type proportions between NMIBC and MIBC patients were evaluated using the Wilcoxon rank sum test.

MIBC: Muscle-invasive bladder cancer; NMIBC: Non-muscle-invasive bladder cancer.

Figure 3. Tumor and blood cell profiles distribution of two groups assigned by the proportion of antitumor immune infiltration.

The high immune infiltration group (High) consisted of the 50 patients who had the sum of B, CD8 T, CD4 T, natural killer and dendritic cell proportions >5% in the tumor microenvironment. The low immune infiltration group (Low) consisted of the 38 patients who had the sum of B, CD8 T, CD4 T, natural killer and dendritic cell proportions ≤5% in the tumor microenvironment. Differences in cell-type proportions between two groups were evaluated using the Wilcoxon rank sum test.

Figure 4. Tumor and blood cell profiles distribution of two groups assigned by consensus clustering algorithm using tumor immune profiles as input.

Group 1 and group 2 consisted of 43 and 45 patients, respectively. Differences in cell-type proportions between two groups were evaluated using the Wilcoxon rank sum test.

Table 2. The distribution of subject characteristics within groups derived from consensus clustering.

	Group 1 (n = 43)	Group 2 (n = 45)	Fisher’s exact test; two-tailed
			Odds ratio (95% CI)	p-value
Tumor grade			14.24 (4.32–46.89) ^†	8.50E-7
Low	18 (42%)	41 (91%)
High	25 (58%)	4 (9%)
Tumor stage			17.68 (4.74–65.99) ^‡	4.46E-7
0a	19 (44%)	42 (94%)
I–IV	24 (56%)	3 (6%)
Muscle invasive			17.03 (2.1–137.9) ^§	0.0007
No	31 (72%)	44 (98%)
Yes	12 (28%)	1 (2%)
BCG treatment			4.91 (0.98–24.65) ^¶	0.047
No	35 (81%)	43 (96%)
Yes	8 (19%)	2 (4%)
Immune infiltration			82.0 (16.62–404.6) ^#	7.22E-14
Low	2 (5%)	36 (80%)
High	41 (95%)	9 (20%)

† Group 1 patients had 14.24-times the odds of having a high tumor grade compared with patients in group 2.

‡ Group 1 patients exhibited 17.68-times the odds of having an invasive tumor stage (stage I–IV vs 0a) compared with patients in group 2.

§ Group 1 patients had 17.03-times the odds of having muscle-invasive tumors compared with patients in group 2.

¶ Group 1 patients had 4.91-times the odds of receiving BCG treatment compared with patients in group 2.

# Group 1 patients had 82-times the odds of having high antitumor immune infiltration compared with patients in group 2.

BCG: Bacillus Calmette–Guérin.

Data sharing statement

Datasets generated and analyzed in the present study are publicly available in Gene Expression Omnibus (GEO) at GSE237100.