Hodgkin Lymphoma: Metabolic Symbiosis Between Malignant Cells and Cancer-Associated Stromal Tissue

Sprandio, John D; Dulau Florea, Alina E.; Martinez-Outschoorn, Ubaldo

doi:10.1182/blood.V122.21.2993.2993

Background

Hodgkin lymphoma (HL) is mainly composed of reactive cells. Lymphocytes, macrophages, eosinophils, mast cells, plasma cells, and other stromal cells constitute these supportive or reactive cells. In-situ glycolytic and oxidative phosphorylation metabolism of reactive cells and cancer cells is unknown in HL. We interrogated HL specimens (N=24) to examine the metabolic compartments in-situ in this disease.

Introduction

High glycolytic activity is seen in the majority of HL on the basis of high uptake of 2-[18] fluoro-2-deoxy-glucose (FDG) on positron emission tomography (PET). The advent of FDG-PET in HL has revealed that these tumors as a unit are glycolytic and staging and assessment of response to treatment has improved. However, it is unknown if cancer cells and reactive cells share a metabolic phenotype.

Purpose

To study mitochondrial and glycolytic metabolism in HL cancer and reactive cells.

Methods/Materials

Two immunohistochemical biomarkers of metabolism were employed on HL tumor sections. Translocase of the Outer Mitochondrial Membrane 20 (TOMM20) protein expression is a biomarker of functional mitochondrial mass and oxidative phosphorylation. Monocarboxylate transporter 4 (MCT4) is a biomarker of glycolysis and lactate export. We selected 24 consecutive cases of classical HL based on morphology and immunohistochemical (IHC) staining for CD30, CD15 and CD45. We recorded FDG-PET standard uptake values (SUV) where available for all tumors studied. Immunohistochemical stains were performed on 5-micron thick, formalin-fixed, paraffin-embedded tissue sections using the horseradish peroxidase method. The primary antibodies used were TOMM20 and MCT4 (Santa Cruz). The immunostaining was graded on a scale of 0 to 2+ according to the intensity and percentage of immunoreactive cancer and reactive cells.

Results

Reed Sternberg and Hodgkin cells (RS/H) had high TOMM20 expression and low MCT4. In fact, the highest TOMM20 expression out of all cells analyzed was found in the RS/H cells in 20 of the 24 samples. In the other 4 samples, cancer and reactive lymphocytes (RL) had similar expression of TOMM20. The tumor stroma or stromal cells in proximity to cancer cells which includes histiocytes (TS/H) had low TOMM20 expression and high MCT4 expression in all 24 samples analyzed. Higher than baseline FDG-PET uptake is a measure of glycolysis and was found in all tumors where FDG-PET was performed. High MCT4 expression was not found in normal stroma or stroma at a distance from cancer cells. Histiocytes always had low TOMM20 expression but MCT4 expression was high in proximity to cancer cells and low at a distance from cancer cells.

Table 1

Patient	Stage	IPI	Max SUV		TOMM20			MCT4
				H/RS	RL	TS/H	H/RS	RL	TS/H	NS
1	IVB	4	20.9	1+	1+	0	0	0	2+	1+
2	IIIB	1	11.63	2+	1+	1+	0	0	2+	0
3	IVB	3	10.9	2+	1+	0	0	0	2+	0
4	IIIA	2	14.7	2+	1+	0	0	0	2+	0
5	IIA	3	23.16	2+	1+	0	0	0	1+	0
6	IIA	1	10.96	2+	1+	1+	0	0	2+	0
7	IIA	1	17.31	2+	1+	0	0	0	1+	0
8	IVB	4	15.87	2+	1+	0	0	0	2+	0
9	IIB	1	2.4	2+	1+	0	0	0	2+	0
10	IIA	3	19.27	1+	1+	0	0	0	2+	0
11	IIB	2	12.67	2+	1+	1+	0	0	2+	0
12	IVB	4	17.24	2+	1+	0	0	0	2+	0
13	IIB	1	14.19	2+	1+	0	0	0	2+	0
14	*	*	*	2+	1+	0	0	0	2+	0
15	IVB	4	12.6	2+	1+	0	0	0	2+	0
16	*	*	*	2+	1+	0	0	0	2+	0
17	IIA	1	3.45	2+	1+	0	0	0	1+	0
18	IIB	3	12.47	2+	1+	0	0	0	2+	0
19	IIB	1	13.47	2+	0	0	0	0	2+	0
20	IIA	1	16.8	2+	1+	0	0	0	2+	0
21	IIIB	3	8	2+	1+	0	1+	0	2+	0
22	IIIA	1	2.77	1+	1+	0	0	0	1+	0
23	IIA	1	7.04	1+	0	0	0	0	2+	1+
24	IIB	1	2.4	1+	0	1+	1+	0	2+	0

Patient	Stage	IPI	Max SUV		TOMM20			MCT4
				H/RS	RL	TS/H	H/RS	RL	TS/H	NS
1	IVB	4	20.9	1+	1+	0	0	0	2+	1+
2	IIIB	1	11.63	2+	1+	1+	0	0	2+	0
3	IVB	3	10.9	2+	1+	0	0	0	2+	0
4	IIIA	2	14.7	2+	1+	0	0	0	2+	0
5	IIA	3	23.16	2+	1+	0	0	0	1+	0
6	IIA	1	10.96	2+	1+	1+	0	0	2+	0
7	IIA	1	17.31	2+	1+	0	0	0	1+	0
8	IVB	4	15.87	2+	1+	0	0	0	2+	0
9	IIB	1	2.4	2+	1+	0	0	0	2+	0
10	IIA	3	19.27	1+	1+	0	0	0	2+	0
11	IIB	2	12.67	2+	1+	1+	0	0	2+	0
12	IVB	4	17.24	2+	1+	0	0	0	2+	0
13	IIB	1	14.19	2+	1+	0	0	0	2+	0
14	*	*	*	2+	1+	0	0	0	2+	0
15	IVB	4	12.6	2+	1+	0	0	0	2+	0
16	*	*	*	2+	1+	0	0	0	2+	0
17	IIA	1	3.45	2+	1+	0	0	0	1+	0
18	IIB	3	12.47	2+	1+	0	0	0	2+	0
19	IIB	1	13.47	2+	0	0	0	0	2+	0
20	IIA	1	16.8	2+	1+	0	0	0	2+	0
21	IIIB	3	8	2+	1+	0	1+	0	2+	0
22	IIIA	1	2.77	1+	1+	0	0	0	1+	0
23	IIA	1	7.04	1+	0	0	0	0	2+	1+
24	IIB	1	2.4	1+	0	1+	1+	0	2+	0

IPI – International prognostic index, SUV – standard uptake values, RS/H – Reed-Sternberg or Hodgkin cells, RL – reactive lymphocytes, TS/H – tumor-associated stroma and histiocytes, NS – normal reactive stroma

*

No follow-up after initial pathologic diagnosis

Conclusion

Glycolysis and lactate export occurs in cancer-associated stroma (TS/H) that is spatially linked to mitochondrial metabolism in cancer cells (RS/H) in HL. This suggests that the most FDG-PET avid cells within HL are the reactive cells and not the cancer cells. Also, normalization of FDG-PET uptake when assessing response to treatment suggests reversal to a metabolically normal stroma.

Representative images, 60x: