O6-methylguanine DNA methyltransferase and glucose transporter 2 in foregut and hindgut gastrointestinal neuroendocrine neoplasms

Abstract

Background: Streptozocin (STZ) is used for treating both pancreatic (PanNET) and gastrointestinal (GI-NET) neuroendocrine tumors but its therapeutic efficacy is relatively low in GI-NETs. Therefore, it has become pivotal to select GI-NET patients who could benefit from STZ treatment. STZ is transported via the glucose transporter 2 (GLUT2) into the cells and the loss of O6-methylguanine DNA methyltransferase (MGMT) also increases its therapeutic efficacy. Therefore, GLUT2 high and MGMT low status could be the surrogate markers of STZ.

Methods: In this study, we examined the MGMT and GLUT2 status in gastrointestinal neuroendocrine neoplasm (NEN). We studied 84 NEN cases: 33 foregut and 37 hindgut GI-NETs and 14 gastrointestinal neuroendocrine carcinomas (GI-NECs).

Results: In GI-NETs, MGMT scores of ≥2 and ≥ 3 were 77% (54/70) and 56% (39/70), respectively, and GLUT2 scores of ≥4 and ≥ 6 were 30% (21/70) and 4.3% (3/70), respectively. Methylation-specific polymerase chain reaction revealed that MGMT promoter methylation was detected only in 2/14 GI-NECs but none of the included GI-NETs. GLUT2 (GLUT2 score) and MGMT immunoreactivity (MGMT and H-scores) were both significantly correlated with Ki-67 labeling index (GLUT2 score: P = 0.0045, ρ = - 0.4570; MGMT score: P = 0.0064, ρ = - 0.4399; H-score: P = 0.0110, ρ = - 0.4135) and MGMT immunoreactivity were significantly correlated with GLUT2 immunoreactivity (MGMT score: P = 0.0198; H-score, P = 0.0004, ρ = 0.5483) in hindgut NETs, but not in foregut NETs. However, discrepancies from the above correlation between GLUT2 and MGMT immunoreactivity were detected in several GI-NET cases which could be potential candidates for STZ therapy.

Conclusion: The evaluation of MGMT and GLUT2 status could provide an important information in planning STZ therapy in GI-NET patients.

Keywords: Glucose transporter 2; Immunohistochemistry; Neuroendocrine neoplasm; O6-methylguanine DNA methyltransferase.

Fig. 1

Representative immunohistochemistry illustrations of MGMT, GLUT2, and Ki-67. 1-1 Representative illustrations before and after digital image analyses of Ki-67 LI. We analyzed nuclear immunoreactivity according to the gradients of brown color (3,3-diaminobenzidine [DAB]) spectrum immunointensity. Tumor cells with blue nuclei were negative, whereas those with yellow, orange, and red nuclei were positive for Ki-67 immunoreactivity. The arrow indicated intratumoral vessels excluded. 1-2 Representative illustrations before and after digital image analyses of MGMT immunoreactivity. MGMT was analyzed only in nuclei according to the gradients of brown color (3,3-diaminobenzidine [DAB]) spectrum intensity. The relative immunointensity of the gradient was evaluated as follows: 0, negative (blue); + 1, weak (yellow); + 2, moderate (orange); + 3, strong (red). 1-3 MGMT and GLUT2 scores

Fig. 2

Correlation among MGMT and GLUT2 scores and Ki-67 LI in foregut and hindgut NETs. 2-1 Results of MGMT and GLUT2 immunoreactivity in all the GI-NET cases examined. There was a significant positive correlation between GLUT2 and H-scores, but not between the GLUT2 score, MGMT score or H-score and Ki-67 LI. 2-2 Results of MGMT and GLUT2 immunoreactivity of hindgut NET cases. There was a significant inverse correlation between Ki-67 LI and GLUT2, MGMT, and H-scores. There was a significant positive correlation between the GLUT2 and H-score. 2-3 Results of MGMT and GLUT2 immunoreactivity of foregut NET cases. There were no significant correlations between the GLUT2, MGMT, or H-scores and Ki-67 LI

Fig. 3

Comparison of MGMT immunoreactivity with relation to individual histological grades of foregut and hindgut NETs. 3-1 Comparison of H-scores between foregut and hindgut GI-NETs. The H-scores were not significantly different between foregut and hindgut NETs. 3-2 Comparison of H-scores between foregut and hindgut G1 GI-NETs. H-scores were not significantly different between foregut and hindgut G1 NETs. 3-3 Comparison of H-scores between foregut and hindgut G2 GI-NETs. H-scores were not significantly different between foregut and hindgut G2 NETs. Data are expressed as mean ± SD (round the fourth digit)

Fig. 4

Comparison of MGMT immunoreactivity between the cases with and without lymph node metastasis in GI-NETs. 4-1 Among G1 GI-NET cases examined, MGMT immunoreactivity (H-scores) was significantly different between the cases with and without lymph node metastasis in GI-NETs. Data are expressed as mean ± SD (round the fourth digit). 4-2 Cases with lymph node metastasis were significantly correlated with H-Score by less than 158.5540 (area under the curve [AUC]: 0.89333; sensitivity: 100%; specificity: 76.67%)

Fig. 5

MGMT status in GI-NETs and GI-NECs. 5-1 H-scores were not significantly different between NETs and NECs. Data are expressed as mean ± SD (round the fourth digit). 5-2 MGMT promoter hypermethylation was detected in two (case 38, 48) of 14 GI-NEC patients. 5-3 Representative images of MGMT immunohistochemistry of case 38 and 48. Case 38 and 48 demonstrated MGMT score 0 and score 3, respectively