Circulating Fatty Objects and Their Preferential Presence in Pancreatic Cancer Patient Blood Samples

Abstract

Human cancers are often complicated with increased incidences of blood vessel occlusion, which are mostly insensitive to anticoagulation therapy. We searched for causal factors of cancer-associated embolism. A total of 2,017 blood samples was examined for visible abnormalities. Examined were peripheral blood samples from cancer patients who were about to undergo surgical treatment for genitourinary, breast, gastrointestinal or abdominal tumors. Samples from ambulatory patients being treated for recurrent or castration-resistant prostate cancers were included in the study. The lipid-rich nature was studied with lipophilic stains and lipid panel analysis, while surface membrane was assessed with specific staining and antibody detection. We identified a new entity, lipid droplet-like objects or circulating fatty objects (CFOs), visible in the blood samples of many cancer patients, with the potential of causing embolism. CFOs were defined as lipid-rich objects with a membrane, capable of gaining in volume through interaction with peripheral blood mononuclear cells in ex vivo culture. Blood samples from pancreatic cancer patients were found to have the highest CFO incidence and largest CFO numbers. Most noticeably, CFOs from many pancreatic cancer samples presented as large clusters entangled in insoluble fiber networks, suggestive of intravascular clotting. This study identifies CFO as an abnormal entity in cancer patient blood, and a contributory factor to intravascular embolism during cancer development and progression.

Keywords: biomarker; blood sample; blood vessel occlusion; cancer-associated thrombosis; circulating fatty object; hyperthrombosis; pancreatic cancer; pre-surgery cancer patient.

FIGURE 1

CFOs in clinical cancer patient blood samples. Packed blood cells from clinical cancer patients were diluted and spread into a thin layer for CFO observation under an inverted microscope. Representative results are shown. (A) When packed blood cells were subjected to 20-fold dilution, CFOs in samples could be seen as transparent dots against a thick background. CFOs can be detected at 20× magnification (upper row). In sample 4, 4 small CFOs could be seen. At 400× magnification (lower row), the relationship between CFOs and red blood cells can be appreciated. (B) When packed blood cells were subjected to 200-fold dilution, CFOs became prominent among the red blood cells.

FIGURE 2

CFOs have a lipid-rich content. Representative images are shown. (A) CFOs isolated from 4 samples of pancreatic cancer were stained with Oil Red O. Red color represents positive staining. All images were taken at 100 × magnification with the same microscope. Note the clustered CFOs entangled in fiber networks in Samples 3 and 4. (B) Representative results of the CellMask green stain of 8 CFO samples from pancreatic cancer (100×). Sample 5 represents the 6 cases in which CFO contents were stained after 30 min. Sample 6 represents the 2 cases in which only the CFO surface was stained after 30 min. Note bundled CFOs entangled with fiber networks in Sample 6. (C) Cholesterol-like crystals in CFOs are indicated with arrows (100×).

FIGURE 3

Basic CFO structure. Basic CFO structure was investigated by determining the presence of an enveloping surface membrane and absence of a cell nucleus-like structure. (A) Some CFOs displayed a wrinkled surface (Control, CFO only) suggesting the presence of an enveloping membrane. The enveloping surface membrane and absence of a nucleus-like structure were investigated with positive staining by membrane-specific CellMask green and negative staining of DNA-binding Hoechst 33,342, respectively. K562 cells were added to indicate positive staining (200×). CellTrace CFSE failed completely to stain CFOs. (B) An indirect detection method for CFO membrane was tested by detecting membrane perilipin and perilipin-2 proteins with biotin-conjugated specific antibodies, which were revealed by secondary binding of streptavidin-conjugated nanoparticles, each of which was 4.5 μm in diameter. During microscopic imaging (400×), when the focus was set on bottom of a CFO, nanoparticles binding specifically on the CFO could be seen.

FIGURE 4

CFOs gain in volume by annexing PBMC blasts. Representative images of mixed ex vivo culture of CFO and PBMCs are shown. (A) In ex vivo culture 1, images show PBMCs on the CFO surface at the beginning of the culture (Day 1), and PBMC blast interaction with CFOs in 2 weeks (Day 14). In ex vivo cultures 2 and 3, two additional examples of the interaction are shown. (B) Tracking the fate of PBMCs fused to CFOs with GFP reporter expression (100×). After hemolytic isolation, samples were placed in mixed ex vivo culture and infected with MISSION TurboGFP lentiviral particles on the first day of culture (Day 1, not shown). PBMC fluorescence was tracked for 4 weeks.

FIGURE 5

CFOs in pancreatic cancer patient blood samples. Packed blood cells from pancreatic cancer patients were subjected to hemolysis. Recovered CFOs mixed in PBMCs were subjected to ex vivo culture. CFOs in 8 representative cultures are shown. All photos were taken on the first day of culture. In the upper panels, CFOs are seen among a background of PBMCs (which have a diameter of 10 μm). In the lower panels, the samples had more CFOs than PBMCs. Many of these samples (e.g., Samples 5, 6, and 7) contained mainly CFO clusters entangled with fiber networks.

FIGURE 6

CFO incidence based on cancer type. Each data point represents estimated CFO numbers in a mixed ex vivo culture. Data points were plotted based on the patient’s cancer type.