Hepatic metastases of colorectal cancer are rather homogeneous but differ from primary lesions in terms of immune cell infiltration

Figures & data

Figure 1. Analytical steps. (A) Workflow for immune cell quantification, cytokine analysis and subsequent generation of a prognostic profile. (B) Analysis of opposite sides of a metastatic lesion. (C) Analysis of serial sections of a metastatic lesion.

Figure 2. Convergence of the average cell number with increasing surface tissue area. The circles on the vertical lines indicate the possible resulting average cell number. Increasing the number of analyzed 1 mm² fields (as indicated on the x-axis) results in a stabilization of the average cell number for a given tissue section.

Table 1. Cell densities in primary tumors and metastases

N°	Location	CD3^+	CD8^+	Granzyme B+	Profile^+	Immune score LiverMet*
1	Primary tumor	102,67	13,00	9,33	−
	Liver metastasis	778,50	396,00	197,39		+
2	Primary tumor	17,00	8,00	13,50	−
	Liver metastasis	971,00	581,00	80,02		+
3	Primary tumor	925,60	127,60	108,80	+
	Liver metastasis	364,67	307,20	156,66		+
4	Primary tumor	406,50	74,50	20,00	−
	Liver metastasis	814,00	407,00	45,00		+
5	Primary tumor	619,67	187,67	22,00	+
	Liver metastasis	1034,00	187,33	202,77		+
6	Primary tumor	316,00	96,50	15,33	+
	Liver metastasis	1252,67	609,00	71,33		+
7	Primary tumor	757,20	112,50	85,00	+
	Liver metastasis	479,75	394,00	22,08		+
8^#	Primary tumor	230,40	20,80	13,60	−
	Liver metastasis	858,33	505,00	72,30		+
9	Primary tumor	314,00	135,50	12,00	+
	Liver metastasis	1019,33	500,00	95,11		+
10	Primary tumor	428,00	67,00	18,50	−
	Liver metastasis	388,67	251,00	22,00		−
11^#	Primary tumor	670,50	92,21	31,50	+
	Liver metastasis	667,00	335,00	87,15		+
12^#	Primary tumor	210,67	166,00	22,62	−
	Liver metastasis	432,33	32,00	3,79		−
13	Primary tumor	693,23	305,91	55,21
	Liver metastasis	888,20	213,64	32,75		+
14^#	Primary tumor	789,05	278,20	49,30	+
	Liver metastasis	101,81	97,02	15,2		−
15	Primary tumor	821,10	198,20	101,10	+
	Liver metastasis	161,66	127,77	9,25		−
16	Primary tumor	991,71	295,91	71,01	+
	Liver metastasis	130,82	91,02	8,91		−

⁺ According to Galon et al.; *according to Halama et al., +, good profile; −, bad profile; #, patient data depicted in Figure 3.

Figure 3. Immune cell infiltration of primary colorectal carcinoma lesions and hepatic colorectal carcinoma metastases. (A–D) Triangles represent the amount of CD3⁺,CD8⁺ and granzyme B⁺ cells detected in primary tumors and the corresponding metastasis. All axes show cell density (cells/mm²) and have the same scale.

Figure 4. Prognostic signature of immune cell infiltration in primary colorectal carcinoma lesions and hepatic colorectal carcinoma metastases. CD3⁺,CD8⁺ and granzyme B⁺ cell densities were translated into a favorable (good) or unfavorable (bad) prognostic signature as described by Galon et al. and Halama et al. Each of presented doublets (P, primary tumor; M, metastasis) represents the samples of a single patient. Green, favorable signature; Red, unfavorable signature.

Table 2. TNM classification and prognostic profile (based on immune cell densities)*

TNM classification (primary tumor)	Primary tumor vs. metastasis classification	Timing of diagnosis of metastatic lesions
pT3 pN2 (8/36) M1 GII	Concordant	Synchronous
pT3 pN2 (13/28) L1 M1 GIII	Concordant	Synchronous
pT4 N0 M0 GIII	Concordant	Metachronous
pT4 pN2 V1 pM1 GIII	Discordant	Synchronous
pT3 pN1(2/13) M1 GII	Concordant	Synchronous
pT2 pN1 (1/13) M1 L1 GII	Concordant	Synchronous
pT3 pN2 (15/15) M1 GII	Discordant	Synchronous
pT4 pN1 M1	Discordant	Synchronous
pT2 N0 M1 GII	Concordant	Synchronous
pT3 pN1 (3/16) M1 GII	Discordant	Synchronous
pT4 pN0 M0 GII	Discordant	Metachronous

* Calculated comparing primary lesions and liver metastases. Average age is 57.9 y (range 40–73).

Figure 5. Correlation between cytokine profiles detected on opposite sides of hepatic metastases. (A–D) Scatterplots of four different hepatic metastases from four different patients Units displayed on both axes are pg/mL (logarithmic scale) and each axis represents a different side of the lesion, with each spot representing a single analyte. Spearman’s correlation coefficients (R_s) were calculated for each of the four data sets, yielding the following statistics. (A) R_s = 0.882, p = 0.0001; (B) R_s = 0.922, p = 0.0001; (C) R_s = 0.9, p = 0.0001; (D) R_s = 0.975, p = 0.0001. Samples (A) and (C) were analyzed for 21 analytes, samples (B) and (D) for 23 (for the detailed list of analytes, please see the Results section).

Figure 6. (A) Heterogeneity of synchronous hepatic metastases of colorectal carcinoma. Overview image of two synchronous hepatic metastases of colorectal carcinoma stained for CD3⁺ cells, showing a liver section with two round metastases (marked with an asterisk and a triangle, magnification 1.5×). (B) Magnification (10×) of the right metastatic lesion in (A) (triangle). CD3⁺ cell infiltrates are clearly discernible at the invasive margin and in the adjacent liver. (C) Magnification (10×) of the left metastatic lesion in (A) (asterisk). CD3⁺ cell infiltrates at the invasive margin are scarce, tumor cells accumulate at the same location, and the core of lesion exhibits large arterial vessels but mainly consists of stromal components.

Figure 7. Homogeneity of synchronous hepatic metastases of colorectal carcinoma. (A and B) Overview image (2.5× magnification) of synchronous hepatic metastases of colorectal carcinoma stained for CD3⁺ cells. In this case, CD3⁺ cell infiltration in both metastases is comparable.

Figure 8. Homogeneity of synchronous hepatic metastases of colorectal carcinoma. (A and B) Overview image (2.5× magnification) of synchronous hepatic metastases of colorectal carcinoma stained for CD3⁺ cells. In this case, CD3⁺ cell infiltration in both metastases is comparable.

Table 3. TNM classification and cytokine profile*

TNM classification	Cytokine profile
ypTx ypN0 M0	A
pT3 pN1 (3/18) M1 GIII	B
pT3 pN1 (2/29) M1 GII	C
pT2 N1 M1	D

* See also Figure 4.