Elevated levels of soluble CD44 precede the development of AIDS-associated non-Hodgkin's B-cell lymphoma

CD44 is a cell-surface adhesion molecule involved in inflammatory cell function as well as tumor cell growth and metastasis ^[1]. There are many isoforms of CD44 consisting of the predominant or standard form [CD44(std)] and a number of variants. CD44 is cleaved from the cell surface, becoming a soluble form (sCD44) detectable in the circulation. Increased sCD44 levels are seen in association with malignancies and immune activation, whereas decreased levels are associated with immune deficiency. In AIDS-associated non-Hodgkin's B cell lymphoma (AIDS-NHL), sCD44 levels may reflect malignancy, immune activation, or immune deficiency.

In an earlier letter to AIDS, Navarro et al. ^[2] reported that mean serum sCD44(std) levels were increased in relatively healthy HIV-infected subjects (CD4 cell number > 500 cells/mm³) compared with uninfected healthy blood donors, but were not further elevated after a diagnosis of AIDS-NHL. The authors also reported that HIV-infected subjects considered to be immunodeficient (CD4 cell number < 500 cells/mm³) had lower sCD44(std) levels compared with other HIV-infected subjects, similar to levels seen in uninfected blood donors. We determined the serum levels of sCD44(std) at a timepoint close to, but before the diagnosis of AIDS-NHL, as an indicator of possible increases in immune system activation, which might create a favorable environment for the development of this malignancy. We have observed that sCD44(std) is significantly elevated in those individuals who went on to develop lymphoma.

All study subjects were homosexual men in the Los Angeles center of the Multicenter AIDS Cohort Study (MACS), who were followed at 6-month intervals since the mid-1980s to examine the natural history of HIV infection and AIDS ^[3]. Frozen archived serum samples were obtained from HIV-infected MACS individuals who developed non-Hodgkin's B cell lymphoma (AIDS-NHL, n = 49), at the study visit closest to but preceding lymphoma diagnosis; the sample date preceded the lymphoma diagnosis by 1-30 months (mean 8.0 months). The mean absolute CD4 T-cell count for AIDS-NHL patients was 191 cells/mm³. Visit-matched control sera from three groups of University of California at Los Angeles MACS individuals were tested in parallel with the lymphoma sera. AIDS controls (n = 44) had an AIDS diagnosis ^[4], no reported malignancy, and were matched by absolute CD4 T-cell number (± 140 cells/mm³), with a mean absolute CD4 T-cell count of 177 cells/mm³. HIV-positive controls (n = 46) were HIV infected but relatively healthy, with absolute CD4 T-cell counts greater than 500 cell/mm³; HIV-negative controls were non-HIV-infected MACS participants (n = 42). Serum sCD44(std) levels were determined by enzyme-linked immunosorbent assay (BenderMed Systems, San Bruno, California, USA), and were calculated using the KC Junior enzyme-linked immunosorbent assay analysis software (BioTek, Inc., Winooski, Vermont, USA). Clinical data and dates were obtained from the MACS database.

The mean serum level of sCD44(std) was significantly elevated in HIV-infected subjects who went on to develop AIDS-NHL (mean ± SEM = 544 ± 20 ng/ml), not only when compared with HIV-negative controls (401 ± 21 ng/ml, P < 0.0001), but also when compared with matched AIDS controls with a similar extent of immunodeficiency and no other malignancy (469 ± 18 ng/ml, P = 0.01), or with relatively healthy HIV-positive controls (479 ± 23 ng/ml, P = 0.02; Fig. 1). In addition, both AIDS and HIV-positive controls had elevated mean sCD44(std) levels compared with HIV-negative controls (P = 0.02). Among the AIDS-NHL cases, there was no difference in mean serum sCD44(std) between NHL subtypes (P > 0.1, data not shown). Using the 95th percentile sCD44 value of the HIV-negative controls (580 ng/ml) as a reference, there were twice as many lymphoma patients with elevated levels of sCD44 (41%) compared with the AIDS (20%), or HIV-positive (20%) controls (P ≤ 0.04).

Fig. 1

Although changes in CD44 cell expression have been reported in HIV infection and AIDS-NHL ^[5-8], there are few published data available on serum sCD44. In contrast to the report of Navarro et al. ^[2], we showed that serum levels of sCD44(std) are elevated in HIV-positive, AIDS, and AIDS-NHL subjects, in spite of varying levels of HIV-induced immunodeficiency. Furthermore, the greatest elevation in sCD44 levels was seen preceding the AIDS-NHL diagnosis. In light of the data of Navarro et al. ^[2], it is possible that pre and post-AIDS-NHL sCD44 levels may differ, but no post-diagnosis sera were available in our study to address this possibility. We suggest that sCD44(std) be considered an indicator of immune activation in HIV disease, particularly in those patients who are in the process of developing a malignancy of the immune system. Our studies demonstrating elevated levels of other markers of immune activation and B-cell stimulation in these AIDS-NHL cases and controls have previously been published or presented ^[9-13]. The addition of sCD44(std) further strengthens the hypothesis that an environment characterized by B-cell hyperstimulation, and conducive to lymphomagenesis, exists preceding the diagnosis of AIDS-NHL.

Acknowledgements

The authors would like to thank the MACS participants, who have made this and many other studies possible, and Xin Huang for assistance with statistical analyses.

Sponsorship: This work was supported by grants from the Leukemia and Lymphoma Society (6155) and the National Institutes of Health (CA73475, CA96888, CA57152, AI28697). The Multicenter AIDS Cohort Study is funded by the National Institute of Allergy and Infectious Diseases, with additional supplemental funding from the National Cancer Institute [UO1-AI-35042, 5-MO1-RR-00722 (GCRC), UO1-AI-35043, UO1-AI-37984, UO1-AI-35039, UO1-AI-35040, UO1-AI-37613, UO1-AI-35041].