(18)F-misonidazole PET imaging of hypoxia in micrometastases and macroscopic xenografts of human non-small cell lung cancer: a correlation with autoradiography and histological findings

Abstract

The objective of this study was to determine whether (18)F-misonidazole could detect hypoxia in macroscopic and microscopic tumors in mice. In nude mice, subcutaneous xenografts and peritoneal metastases were generated utilizing human non-small cell lung cancer A549 and HTB177 cells. Animals were co-injected with (18)F-misonidazole, pimonidazole and bromodeoxyuridine, and tumor perfusion was assessed by Hoechst 33342 injection. The intratumoral distribution of (18)F-misonidazole was determined by micro-PET scan and autoradiography. Pimonidazole, bromodeoxyuridine and Hoechst 33342 were detected by immunohistochemistry on the autoradiography sections. Submillimeter micrometastases found to be severely hypoxic. In both peritoneal metastases and subcutaneous xenografts models, PET images displayed significant (18)F-misonidazole uptake, and its distribution was non-uniform in these macroscopic subcutaneous tumors. In frozen sections, digital autoradiography and immunohistochemistry revealed similar distributions of (18)F-misonidazole, pimonidazole and glucose transporter-1, in both microscopic and macroscopic tumors. Bromodeoxyuridine stained-positive proliferative regions were well perfused, as judged by Hoechst 33342, and displayed low (18)F-misonidazole accumulation. (18)F-misonidazole uptake was low in tumor stroma and necrotic zones as well. Microscopic non-small cell lung cancer metastases are severely hypoxic. (18)F-misonidazole PET is capable to image hypoxia noninvasively not only in macroscopic tumors but also in micrometastases growing in mice. Accordingly, (18)F-misonidazole may be a promising agent to detect the burden of micrometastatic diseases.

Keywords: 18F-misonidazole; Micrometastasis; PET; autoradiography; hypoxia.

Figure 1

A. PET coronal slices of intraperitoneal distribution of ¹⁸F-misonidazole in a peritoneal disease–free mouse (left) and a mouse with peritoneal metastases after A549 cells inoculation (middle) with necropsy-confirmed peritoneal wall carcinomas (right). High radioactivity accumulation was found on the abdominal wall as indicated and the overall background in the peritoneal cavity (excluding guts) was apparently higher than control. B. ¹⁸F-misonidazole (¹⁸F-FMISO) distribution in A549 i.p. tumors in mice (n=5). C. Multiple individual lesions attaching to the peritoneal wall where PET imaging had revealed a high level of ¹⁸F-misonidazole: Hematoxylin and Eosin stain showing multiple individual micrometastases (generally less 1mm in diameter) which had little Hoechst 33342 uptake, associated with a high fraction of pimonidazole binding and high ¹⁸F-misonidazole accumulation by autoradiography. D. Single cells or ascites tumors harvested from ascites of the mouse were stained-positive for pimonidazole, Hematoxylin and Eosin stain provides as a reference. All scale bars are as indicated.

Figure 2

The relationship between ¹⁸F-misonidazole uptake and hypoxia, proliferation and perfusion in A549 peritoneal micrometastases. High levels of ¹⁸F-misonidazole uptake, pimonidazole binding and glucose transporter-1 expression are co-localized with each other and with low proliferation (by bromodeoxyuridine) and perfusion (Hoechst 33342). All scale bars are 1mm.

Figure 3

Intratumoral distribution of ¹⁸F-misonidazole in a macroscopic HTB177 subcutaneous xenograft by PET and autoradiography and its relationship to tumor microenvironment. Autoradiography, pimonidazole, glucose transporter-1 expression, and Hoechst 33342 images were obtained from the same frozen tissue section, Hematoxylin and Eosin stain from an adjacent section. Necrosis (N) and stroma (as arrow indicated) are associated with low ¹⁸F activity. All scale bars are 2 mm.

Figure 4

Intratumoral distribution of ¹⁸F-misonidazole in a macroscopic A549 subcutaneous xenograft by PET and autoradiography and its relationship to tumor microenvironment. Autoradiography, pimonidazole, glucose transporter-1 expression, Hoechst 33342 and Hematoxylin and Eosin stains were obtained from the same frozen tissue section. Stroma and pimonidzole-negative cancer cells associated with low ¹⁸F-misonidazole accumulation. Glucose transporter-1 expressing regions are much wider than those positive for pimonidazole. All scale bars are 1 mm.

Figure 5

Comparing intratumoral distribution of ¹⁸F-FDG and ¹⁸F-misonidazole by PET in the same tumor. Intratumoral distribution of ¹⁸F-FDG and ¹⁸F-misonidazole were roughly similar, though mismatched regions were apparent. Images were obtained 24 h apart; animals were fast overnight for ¹⁸F-FDG study.