The increased frequency of the HLA-A2 allele among metastatic ovarian and prostate cancer patients compared to the healthy population in Sweden is an interesting finding. The northern countries in Europe are already over represented by HLA-A2 positive individuals; nevertheless, we could detect 17%, respectively 15%, higher percentage of HLA-A2 among these two cancer groups. At the moment these observations do not answer the question if this is a selection that could depend either on the tumor type or previous oncological treatments. If the latter were the case, most of the patients that survive to receive repeated lines of therapy would be HLA-A2 positive. This would suggest that HLA-A2 might be a prognostic factor. Wang et al. have presented similar observations in cervical cancer (7). They show a progressive change in the frequency pattern of a number of HLA alleles from healthy controls to patients with low-grade to high-grade dysplasia or invasive cancer. In order to answer the question if HLA-2 could be a risk factor for ovarian and prostate cancer, we have planned to perform a more extensive analysis on a number of patients that have newly detected these cancer forms.
Cavalli-Sforza et al. showing a pattern of progressive decrease in expression of HLA-A2 and -A3 alleles from north to south confirm the higher frequency of HLA-A2 among the normal Swedish population, compared to the other European countries (6). Furthermore, the correlation between HLA-A2 gene distribution in 19 European countries and their geographical position correlated as well to the distribution of age-adjusted mortality rates for ovarian and prostate cancer. As far as we know, these three variables have not been considered causal. The correlations we have described above could not be found in other gynecologic, urinary, lung, and breast tumors, corroborating the hypothesis that HLA-A2 might be associated with ovarian and prostate cancer in a more dependent way.
The major histocompatibility complex (MHC) is the direct link between peptides and specific immune response. At a population level, the control on non-self antigens depends on the HLA haplotype more frequently expressed. Moreover, different HLA “motifs” in the MHC system play different roles in the escape from T-cell-mediated recognition, as seen, for example, in HIV-1 infection (8,9).
Our findings show that HLA-A2 was the most represented allele among our ovarian and prostate cancer patients. The frequency of this allele is overrepresented compared to the general Swedish population. The observation that mortality rates for ovarian and prostate cancer correlate with HLA-A2 gene frequency, suggests that HLA-A2 might be an important prognostic marker. An extended retrospective analysis on a greater number of cases is necessary to corroborate this hypothesis. In order to ensure that the segregation of patients with these alleles is not due to the natural course of the disease or to treatment selection, a prospective analysis is mandatory.