Discussion
Cytomegalovirus is wreathed in mystery, a Gordian knot of biological knowledge and understanding [82]. Cytomegalovirus had evolved a peculiar virus-host symbiosis by a taciturn tactic that co-opts or escapes immune pathways in order to arrange its persistence in the host. A strong inverse statistical correlation between the incidence of cancer and CMV seroprevalence covers quite a swath of cancer spectra and territorial ranges. Although correlation does not categorically infer causation, this one may underscore a causal meaning. Besides the benefits of vaccine against CMV [53], it may suggest an impeded carcinogenesis by generating the CMV-induced tumoricidal T cells. CMV-based vaccine antagonistic to tumorigenesis would possibly result in a global regression of cancer. Comparing non-Whites to Whites, Cannonet al. [2] evaluated that CMV seroprevalence is consistently 20-30 percent points higher in the former than in the latter (summary PR=1.59, 95% CI=1.57-1.61). Some ethnic groups had CMV seroprevalence ~ 100% (Fig. 4A in [2]). In agreement with the current work, Fowler et al. [32] discounted the importance of ethnicity per se as a risk factor for the CMV infection. Similar conclusions were reached by Lantos et al. [54], Rook [55], and Dowd et al. [56,57].
Elimination of health disparities as a consequence of racial and ethnic differentials has been earlier recognized as important [47,48,83]. Socioeconomic disparities across race/ethnicity categories impact the level of CMV seroprevalence [83]. These may have biased lower estimates of cancer incidence at the time. Rather than the race/ethnic divides of themselves, we propose CMV infection as an oncopreventer both of some hematologic malignancies at our clinic and of cancers worldwide [13]. Although CMV prevalence and education level, SES, and household income are associated with race and ethnicity, we speculate here that latent CMV infection is a fundamental cause underlying disparities in cancer incidence among race/ethnic groups in the U.S. and worldwide. The NHANES III data (the U.S. 2011-2012) report CMV prevalence (race/minority) in children 1-5 years of age as 37% among non-Hispanic other/multiracial, 31% among Hispanic, 15.9% among non-Hispanic Black, and 10.6% among non-Hispanic White ethnicities [33]. This data is inversely proportional to the incidence of all cancers (combined) in populations of these youths, indicative of a higher risk of cancer in CMV seronegative populations in the U.S. Also, unlike Hispanos, cancer rates in Cubans were comparable to non-Hispanic Whites, and Puerto Ricans and Cubans in Florida had rates of some solid cancers similar to non-Hispanic Whites despite the rates of these cancers being significantly lower in their countries of origin [22]. High incidence of cancer despite a high prevalence of CMV in the Inuit (Eskimos) may be a consequence of deficient T cell immunity in this ethnic population [23,24]. This is to be expected and, indeed, we found no global correlation between the incidence of Kaposi’s sarcoma (mostly diagnosed in HIV-positives with compromised T cell immunity) and the prevalence of CMV (Fig. 5 ). CMV does not exert protection unless T cell immunity is functional.
Immigrants to U.S. experience decreasing incidence rates of cancer of infectious origin (hepatitis B virus, Helicobacter pylori , human papillomavirus) which are prevalent in their countries of origin. On the contrary, incidence rates of lung, breast, colorectal and prostate cancer have been on the rise despite remaining relatively low in the host nations [57-59].
The statistical analysis indicates an inverse correlation between CMV pervasiveness and the race-specific cancer incidence, a valuable hint at a possible oncoprotective effect of the pathogen. Previously, we speculated that CMV may confer a protection against B cell dyscrasias [13]. For example, there is a highly significant inverse link (Fig. 1 . Spearman’s ρ = -0.754;p <0.001) between all invasive cancers combined (both genders, all ages) and the country specific CMV prevalence profile across the mainland and sea-coastal regions of 73 countries (Table 3 ). We draw attention to a possible protective effect of the CMV infection as an unappreciated factor which may subvert oncogenesis.
We sought to grasp a better understanding of fluctuating incidence of new cancers among immigrant and established indigenous populations by consulting relevant reports. We envision that dilution of the prevalence of CMV, due to a progress in economic opportunity and an open access to competent medical patronage, may have resulted in increased cancer rates, indeed the epidemics of neoplasms [60]. Societal development aside, poverty remains a considerable medical concern [30,33,62,72]. Higher prostate and lung cancer rates are reported in established immigrant enclaves from Japan and China in the U.S. than are found in Japan and China, probably because improved SES in the U.S. (i.e. a better hygiene and improved medical care) with a consequent decline of CMV-mediated oncoprotection in the U.S. Also, individuals with low SES have higher antibody titers to CMV [31] and presumably more protection against cancer.
Crucially, CMV induces a specific, CMV-determined, T cell mediated antitumor effect in immunocompetent persons but fails in patients with inoperative T cell immune surveillance, like Kaposi sarcoma (Fig. 5 ). The role of CMV acquisition and the consequent T lymphocyte-specified inhibition of tumorigenesis may prove of importance in pre-empting various malignancies, particularly those that can be detected at an early stage, such as breast and colon cancers.
Disaggregation (decomposition) of ethnic data in racial/multiethnic studies of cancer vs. rates of cancer cases combined and the CMV prevalence may, while quite information-rich, also be a source of bias due to unmeasured confounders or mediators [79]. Disaggregation may mask true biological linkages which would be more accented if global bulk of data were analyzed. Hoshiba et al. [61] examined the long-term (1980-1998) dynamics of CMV seroprevalence in pregnant women in Japan. Complement-fixing antibody and specific IgG antibody, as measured in sera, decreased gradually from 93.2% to 66.7%. CMV-IgG seropositive rates were 87.4% in 1985, and 75.2% in 1996 to 1997. This provides a separate hint at a possible protective effect of CMV in this population. Of note, the incidence of cancer increased in parallel to decrease of CMV seropositivity in this Japanese population.