The ABO blood group polymorphism has been associated with different diseases, cancer included (1). The cancer developments are variegated processes associated with aging. Protection from cancer and atherosclerosis is the main longevity reason. Long-survivors are an important group for the evaluation of genetic markers in cancer pathogenesis (2).

The ‘histo-blood system’ was first proposed for the ABO system characterized by the presence or absence of the carbohydrate antigens A and B on the erythrocyte membrane and blood plasma regular antibodies (anti-A, anti-B) (3). The A and B alleles encode a specific glycosyl-transferring enzyme. The blood groups are characterized by the presence of carbohydrate sugars on the red blood cell surface: N-acetylgalactosamine for the A antigen and D-galactose for the B antigen, both for the blood type AB and neither for the phenotype O. The A, B and AB-related carbohydrate sugars are situated on the H antigen, and the unmodified H antigen defines the blood group O (4). The A and B blood groups differ mainly by four amino acid substitutions (R176G, G235S, L266M, and G268A) from four common variants (rs7853989:C>G, rs8176743:G>A, rs8176746:C>A, and rs8176747:G>C), while the blood group O is characterized by a single nucleotide deletion (rs8176719(−;G)) which results in early protein termination (5,6).

Population studies have demonstrated that the ABO group phenotypes frequencies vary widely from one ethnicity to another: The O blood group is common in South American and Africans natives, A and B being common in Asian and North European countries, respectively, while AB is more frequent in Korea, China and Japan (7).

The ABO (ABO, 9q34.1) genes control the expression of part of the carbohydrates by the epithelial cells in the respiratory, genitourinary and gastrointestinal systems; the carbohydrate variability acts as a potential receptor for non-pathogenic and pathogenic microorganisms influencing immune responses (8,9).

The first report on the relation between the ABO blood group and cancer was published by England and Scotland researchers. A study of the frequency of the ABO blood groups in patients with stomack cancer indicated the A blood group to increase the risk, while the O blood type was protective (10). Thereafter, the correlation between the ABO blood type and other malignancies, such as gastric (11) and pancreatic (12–15), has been continuously reported. A total of 1.6 million healthy blood donors were followed in Denmark and Sweden: The A, B and AB blood groups were associated either with the increased or decreased risk of cancer at 13 anatomical sites as compared with the O blood group (16). Multiple mechanisms have been indicated to explain the blood type role in cancer progression, including altered immune response, inflammation and cellular adhesion (6,17–19).

The role of the specific genetic differences contributing to life expectancy is hardly known. Several genome sequencing and GWAS studies compared the total number of disease variants in centenarians and controls, indicating that there are some evidences that centenarians harbor the anti-aging polymorphisms which protect them from diseases, although long-survivors may show numerous disease variants at a rate similar to normal people, but they are protected from their effects (20,21). The SNP (rs8176719) defining the most common allele responsible for the O blood group is related with longevity; the centenarians are more likely than controls to have the O blood group. In the GWAS analysis, this SNP has shown that the O blood group is associated with numerous protective effects (22). Genetic association studies identified that the polymorphism located near APOE was found to be consistently associated with extreme longevity (23). The longevity genetic markers in the Lithuanian population were found to be the low levels of apolipoprotein B (apoB) and the apoE and the low apoB/apoA-1 ratio (24,25). The Gc1 allele of the vitamin D-binding protein demonstrated a marked decrease in the oldest-old cohort, and the Gc2 allele frequency is expressed in the opposite way in the Lithuanian population (26).

Long-survivors as the cohort are important for evaluating the role of genetic determinants in cancerogenesis and the related mortality. The further change in mortality patterns will increase success in reducing the number of cancer-related mortalities (27). In our study, we evaluated the ABO genetic variation in prostate, renal cell and bladder cancer patients, because the frequency of kidney and prostate cancer and mortality from these cancer types in the Lithuanian population is the higest in Europe (28,29). The second task was to determine if the ABO blood type could be one of the multifactorial determinants in relation with the longevity as the resistance to cancer. The cluster analysis of the ABO blood group system in the Lithuanian population was selected to compare the structure of the analyzed data, their mathematical similarity and to form groups (clusters) in the population of Lithuania comparing the groups of prostate, kidney, bladder cancer patients with the healthy blood donors, long-survivors. The cluster method helps to estimate differences (similarities) and genetic distances among the groups of study subjects by showing the cluster hierarchy and their interconnectedness regarding the ABO group frequency. In the literature review, we included the ABO relationship with cancer, whereas resistance to cancerogenesis is an additional important determinant for longevity.

The ABO blood phenotype frequencies vary in different ethnic/racial groups; they vary globally, where the type O frequency is most common and approaches 100% among the indigenous populations of South and Central America; the phenotype A is more common in Eastern and Central Europe, the phenotype B being more common in China, India and the AB type is more frequent in Korea, China and Japan (32). In the US Caucasians, the frequency of blood groups O, A, B and AB is 45, 40, 11 and 4%, respectively, in Hispanics the distribution being 57, 31, 10 and 3%; and in Blacks 50, 26, 20 and 4% (33); in the Lithuanian population, the blood type frequencies (in the blood donors cohort) are as follows: O-38.6%; A-37.5%, B-16.2%, AB-7.6%. The geographical and ethnic differences in the distribution of ABO blood groups may influence associations between the ABO blood group phenotypes and the risk of diseases.

The clustering analysis of our studied groups has shown that there are separate main clusters of oldest-old individuals in the analysis of pooled as well as of separate male and female cancer (prostate, bladder, kidney) in comparison with control blood donors and oldest-old groups.

The study data on the ABO blood type frequency and OR in prostate cancer patients (all cases pooled independently of the Gleason score and stage) showed a significantly higher blood group B frequency as compared with the blood donors in the Lithuanian population. Other researchers using three single nucleotide polymorphisms (rs8176746, rs505922, and rs8176704) to determine the ABO genotype in 2,774 aggressive prostate cancer cases (the Gleason score ≥8 or locally advanced/metastatic disease) found no association between the ABO blood group phenotype and aggressive prostate cancer (34). The lack of association between the ABO blood type and the prostate cancer risk or survival was reported (35). Notably, blood type O prostate cancer patients were at a significantly lower risk of biochemical recurrence as compared with the blood type A patients. The analysis of radical prostatectomy patients and their 5-year biochemical recurrence-free rate showed the blood type O to be significantly associated with a decreased risk of biochemical recurrence as compared with blood type A patients (36). A significant survival has been reported for prostate cancer blood types B and O patients treated with the therapeutic cancer vaccine PROSTVAC-VF, and the type B and O PROSTVAC-VF patients demonstrated markedly improved clinical outcomes as compared with the blood types A and AB patients, including a longer median survival and the improved overall survival. This indicates that the ABO blood group may provide a screen to pre-select patients for PROSTVAC-VF therapy (37).

The loss of the blood group A antigen expression in prostate cancer tissue and the retention of the H antigen (38) as well as the A antigen expression decrease with tumor progression have been reported (39). The lost A antigen expression on cancer cells might be related with the increased risk of biochemical recurrence of prostate cancer in patients with A blood group (36). An alternative explanation of this phenomenon could be the cross-reaction of anti-A antibodies with the tumor-associated Tn antigen (alpha-N-acetylgalactosamine-O-serine/threonine), the latter being expressed in the prostate cancer tissue (40). The Tn antigen is a marker for the design of anti-tumor vaccine studies (40,41). The anti-A antibodies can cross-react with the Tn antigen (42); this cross-reaction might be related with the decreased risk of biochemical recurrence in the blood type O patients (37). The risk of venous thromboembolic events which are the most common complication and the cause of death within 30 days after radical prostatectomy is greater in patients with a non-O blood phenotype than in those with the O blood group (43).

Our study of the OR and of the frequency of ABO blood groups shows the bladder cancer risk (in pooled males and females and men bladder cancer groups) to be significantly related with the blood group B, but no such association was determined in Lithuanian cancer female patients; also, the blood group O was found to be significantly decreasing the risk of bladder cancer for females as tested in association with the blood donors.

The literature data on the ABO blood group prognostic value in bladder cancer patients are conflicting. Bladder cancer patients with the O blood type seemed to have a worse clinical outcome than those with non-O blood types. A tendency for high-grade and large-sized tumours was observed in the blood groups O and B patients (44). A significantly worse 5-year recurrence-free and cancer-specific survival was associated with a non-O blood type, but no significant difference in the overall survival was found. Non-O blood types were associated with the highest cancer-specific mortality, especially in patients with the blood type A (45). In patients with nonmuscle invasive bladder urothelial carcinoma with blood type O, the recurrence and progression rates were significantly worse than those with blood types A or B (46). In contrast to findings among Europeans and Americans, in the Japan population there were no significant differences among blood groups for the stage, histological grade or survival rate of bladder cancer patients (47). In urothelial bladder carcinoma patients after radical cystectomy, the blood type B was associated with a greater likelihood of lymphovascular invasion and positive soft tissue margins, and blood type B patients showed a significantly higher urothelial bladder carcinoma-related mortality than those with the blood types O and AB (48). By other researchers, the ABO blood group was not associated with the prognosis of bladder cancer patients who underwent radical cystectomy (49).

The cancer originating in epithelial cells and the changes in blood group antigens in the cancer tissue constitute an important aspect in tumor immunology. The expression of soluble A and B antigens was decreased in patients with bladder cancer in comparison with the normal subjects (50). The reduction of A antigen expression in the bladder cancer tissue was found, while the expression of the A antigen was maintained in the normal or concomitant dysplasia urothelium, indicating that the ABO gene promoter hypermethylation and/or the gene loss is a specific marker of the bladder cancer (51). The others reported the ABO blood group antigen expression to be associated with a higher bladder tumor grade, but in adjuvant chemotherapy receiving patients the AB phenotype antigen expression was associated with a reduced bladder overall and cancer-specific survival (52).

A comparison of the frequency of ABO blood groups between blood donors and the kidney cancer group revealed no significant difference, and no OR significance of any ABO blood group type in the studied kidney cancer (polled, male and female) groups in association with the blood donors was found in our study. By others, the ABO type frequency differences according to the renal cell carcinoma (RCC) histological subtype (clear cell type and non-clear cell type) reached no statistical significance (32) or the kidney cancer risk was inversely associated with blood type AB compared with the type O (53). The follow-up cohort study has revealed the RCC incidence to be higher in non-O blood type subjects than in blood type O women (54).

There are several conflicting reports on the ABO blood type relationship with the survival of RCC patients. The ABO blood type was reported to be significantly associated with the progression-free survival and cancer-specific survival in patients with RCC following partial or radical nephrectomy: A non-O blood type (A, B or AB) is an independent prognostic factor for a worse progression-free survival outcome, and the blood type A is an independent factor associated with a worse cancer-specific survival (32). In patients undergoing surgery for loco-regional RCC, the O blood group was found to be independently associated with the better overall survival, and the non-O blood type was identified as an independent predictor of mortality (55). In other studies, the O blood group was associated with the absence or a lower level of lymph node metastases, and this finding did not generate a favourable impact on the RCC prognosis (56,57). Others have reported the ABO blood groups to be not associated with survival outcomes and not a prognostic factor in renal cancer patients after surgery (57). The risk of bilateral RCC was found to be significantly increased in relationship with the O blood type (56).

ABO antigens may be related with systemic inflammation, and chronic inflammation is associated with RCC (58). The C-reactive protein level was related to the ABO blood type: RCC patients with the blood type O had a significantly lower C-reactive protein level than those with a non-O blood type (32). There is a relationship between the ABO gene and the the tumor necrosis factor alpha (59); the ABO gene locus is linked with TNF-α serum levels (60), so blood group antigens may contribute by modifying the immune surveillance (17). Non-O blood group patients have higher levels of the von Willebrand (vWF) and VIII factors (61). Thus, non-O blood group patients have a greater tendency of hypervascularity and hypercoagulability which are typical RCC characteristics. The inquired A and/or B antigen is lost in the RCC tissue; these structural changes in the ABH antigen, occuring in the RCC, may enhance the progression of the disease or survival (62–64).

The blood group phenotype O is characterized by the presence of the H antigen and the anti-A and anti-B antibodies in the serum (natural isoagglutinins, alloantibodies) which could act as a cancer defense factor (6,64); these isoagglutinins can react to the Tn and T pancarcinoma antigens (65); the A and B antigens were suggested to be structurally related to Tn and T and could make cancer cells immunologically less recognizable in the blood groups A and B individuals (66).

The hypothesis encompasses the enzymatic activity dysregulation of the ABH glycosyltransferases involved in the processes of cellular adhesion, membrane signaling and immune response (17,67,68). The A/B glycosyltransferases modulate the circulating plasma levels of vWF (69,70); the vWF has been recently found to be an important modulator of angiogenesis and apoptosis which are tumorigenic processes (71). The A, B or H antigens are involved in controlling cell proliferation, adhesion and motility as these antigens are present on receptors for the epidermal growth factor, cadherins, CD44 and integrins (17,72–74). The ABO blood group carbohydrates on the surface of metastatic cancer cells are functioning as cell adhesion molecules (75). There are reports on the association between circulating levels of tumor necrosis factor-alpha and the ABO gene locus polymorphisms (60), the latter being associated with the soluble intercellular adhesion molecule (ICAM-1) (76–78), E-selectin (79,80) and P-selectin (78). All these molecules are immune cell recruitment mediators in chronic inflammation which may have a directly linking ABO blood group and tumor initiation and spread (81,82). Compared with the blood group O, the expression of ICAM-1, which inhibits lymphocyte attachment to endothelial cells, is significantly reduced in patients with a non-O blood group (blood group A in particular) (6,83). The decreased levels of the soluble ICAM in non-O blood group patients may be related to tumor spread (84). This suggestion has been supported with a number of studies showing a decreased survival in non-O blood group and a favorable prognosis in cancer patients with the O blood group (85,86). With the A1 allele, the expression of P-selectin and ICAM-1 is associated (78). During inflammation, E-selectin increases leukocyte accumulation, and P-selectin enables leukocytes to interact with endothelial cells or platelets. Inflammatory cytokines rapidly upregulate ICAM-1 and can inhibit lymphocyte adhesion to endothelial cells (83). Compared with non-O blood groups, monocytes of O blood group individuals produced higher levels of interleukin-6, nitric oxide and tumor necrosis factor-alpha (87).

The ABH antigens are expressed on the surfaces of several normal and tumour tissues, including the kidneys and renal cell carcinoma lines (17,88), bladder (50–52) and prostate tissue cells (36,38,39). Compared to normal cells, aberrant ABH antigens have been found on tumor tissue cells, such as A or B epitope loss and H antigen accumulation, or an incompatible expression of tumor A antigens is notable in blood type O subjects (17). The loss or diminished expression of ABH antigens has been reported in both solid and hematological malignancies. There have been occasional case reports of ABO blood group antigen change in malignant conditions, and the antigens were re-expressed when the patients attained remission. There are two possible mechanisms for the weakening of ABO antigens expression during malignancy: 1) Inactivation of A and B transferases (there is a decreased expression of the A and B antigens with a concurrent increase in the H antigen), 2) the inactivation of H transferase, related to a decreased H substance, and a resulting decrease in A and/or B substance (89). An altered expression of ABO antigens may contribute to a malignant phenotype via increased cell motility or enhanced evasion of apoptosis (18). The loss of A or B glycosyltransferase enhances malignancy in human carcinoma cell lines (90) and is associated with a less favorable prognosis (91). In the non-A individuals, tumor cells can express the A antigen, while glycosylation can lead to conformational changes in proteins such as the epidermal growth factor receptor, or alter the immune recognition of natural killer cells which are the conditions favoring tumorigenesis (92). The underlying mechanisms by which the ABO blood group may interplay with tumorigenesis need further research (81). A deletion of A or B antigens in non-O blood group patients leads to the upregulation of the H precursor and Lewis factor, both of them stimulating neoangiogenesis (93).

In our study, a comparison of the oldest-old and blood donor groups has revealed the blood group A to be significantly more frequent, and the blood type B was significantly rarer in the oldest-old group; the blood type A frequency was significantly lower and the blood group B was significantly more frequent in the prostate cancer group as compared with the oldest-old group, and the study of OR has shown that the blood type B significantly reduces the oldest-old age likelihood in prostate and bladder cancer patients, while the A blood type has shown a significant opposite association in prostate cancer patients.

Our finding that the blood group A was significantly overrepresented in the oldest-old cohort deserves special attention. When comparing cancer patient groups with oldest-olds, the blood group A was found to be significantly overrepresented among urogenital cancer patients. There are two blood group A subtypes, namely A1 and A2, and the A1 allele increases the ABO protein level with the highest enzymatic activity (94). The increased susceptibility of pancreatic cancer has been reported to be possibly connected with the blood group A which is likely to be related with the A₁ subtype (15). The possible associations with the blood type A have been found in ovarian cancer: Cases with minor alleles of rs1053878 had a 50% lower risk of death and a better survival related with the blood type A and imply that this relationship may be caused by the A2 allele (95). The meta-analysis of eight studies indicated that women with genetic blood type A variants had a 9% greater ovarian cancer risk than did blood group O ones, and the increased risk was only evident for A1 genotype cases (89).

Furthermore, thromboembolism is the additional way by which the ABO blood type may influence the survival of cancer patients. Individuals with the blood type O are at a lower venous thromboembolism risk and have lower vWF levels than the ones with non-O blood groups, and the A2 blood phenotype was found to be independently related with a lower thromboembolism risk and to have the lowest vWF levels among all other ABO blood phenotypes (96,97). The better survival in relationship with the lower risk of tromboembolism among cancer patients with blood type A could be due to the A2 allele. The limitation of our and many other studies of the role of ABO blood groups in longevity is that they do not distinguish between the A1 and A2 subgroups evaluating the phenotype rather than the ABO genotype, whereas this might be of biological significance as the A1 and A2 glycosyltransferases display different catalytic activities, the A2 isoform having a higher K_m and the estimated enzymatic activity 30–50 times lower than A1 (17). The differences in the A1 and A2 allelic proportions across different populations may produce some inconsistencies across the studies. A significant increase of the A blood phenotype was observed in the population of healthy elderly males aged over 64 years in the UK (98).

Our OR analysis of the association among cancer and oldest-old groups shows that the blood type B significantly reduces the old-oldest age likelihood in all studied kidney cancer (pooled, men, and women), bladder cancer patients for the pooled and separate men and women groups; notably, the A blood group has shown a significant opposite association for the women kidney cancer, in pooled and men bladder cancer groups in the Lithuanian population. We and others have found a significant association between the blood group B and the higher prevalence of tumors, but there are no clear biological mechanisms to correlate the association of the blood type B with different tumor development (13). The ABO blood type may influence cancer patients’ survival by mechanisms related to coagulation. The ABO gene is responsible for the post-translational glycosylation of the procoagulant molecule vWF (99). Thromboembolism has an impact on the poor cancer patients’ survival, and patients with the blood type O have lower rates of thromboembolism because of the lower levels of vWF and the increased risk for those with blood type B (100,101).

Another study shows a negative relationship between the B blood type and life expectancy in a large cohort (n=28,129); its authors indicate that this evidence may be attributable to the association between the B blood group and some aging-associated conditions raising a new link between the ABO blood type and cancer (102). Others have found the percentage of patients with the blood group B and their survival to decline with age, and the blood group B may be a marker for earlier death (103). On the contrary, compared to the controls, the B group more frequently was observed in Japanese centenarians, suggesting that the group B might be associated with longevity; the group B individuals were more likely to survive age-related diseases, since 33% of the centenarians were free of such diseases, and this did not correlate with the group B. These conflicting results regarding the blood type B demonstrate that demographic evidences suggest longevity to be possibly caused by different combinations of genes in relation with the different environmental and geographical quantitative and qualitative differences in which the population-specific genetic factors play a role in the longevity (104).

The ABO blood group system is a critical player in the genomic medicine. The structural analysis of our study and other researchers’ data indicate the blood type B to be related with the risk of prostate and bladder cancer and could be evaluated as a congenital determinant in the negative relationship with longevity, and that blood types O and A could be the positive factors increasing the oldest-old age likelihood. Moreover, the further studies of the ABO blood groups evaluating the A₁ or A₂ in relationship with the ABH enzymatic activities, their role in cellular adhesion, immune response in the pathogenesis of cancer could be also important for longevity elucidation.