β2-Microglobulin is a subunit of the human leukocyte antigen class I molecule that is shed by activated lymphoid cells and has been found to be raised when there is increased cell division, such as in several malignant diseases (1) and HIV infection (2). Previous reports have shown that β2-microglobulin levels, usually measured at arbitrary points in patients' infection, can predict the rate of development of AIDS over the ensuing few years (2-17). In this study we aimed to assess whether β2-microglobulin measured close to HIV seroconversion could predict the course of infection over the following 10-15 years.
PATIENTS AND METHODS
We studied 63 men with haemophilia who had been seen every 3-6 months for clinical review by one of us (CAL) at the Haemophilia Centre, Royal Free Hospital, London. The use of frozen serum enabled us to estimate the date of seroconversion to within an average maximum error of ±4.9 months in 63 men with haemophilia infected with HIV-1 between 1979 and 1985. These men are part of a larger cohort of 111 men previously reported (3,18), and represent those with the most precisely known dates of seroconversion. The estimated date of seroconversion is calculated as the midpoint of the last date when they were known to be anti-HIV negative and the first date when they were known to be anti-HIV positive. The patients were aged 2-77 years (centiles: 10%-14; 25%-19; 50%-25; 75%-38; 90%-54) at seroconversion. AIDS was defined according to the CDC 1987 definition (19).
Treatment of AIDS-free patients began in some patients in 1988 with zidovudine as part of the MRC/ANRS Concorde trial. From February 1989, patients had been given prophylaxis against Pneumocystis carinii pneumonia with pentamidine and, more recently, co-trimoxazole. Currently, patients are treated with zidovudine/ddl, PCP prophylaxis, and Candida prophylaxis (with fluconazole) when the CD4 lymphocyte count declines <200/mm3 and clarithromycin or rifabutin as prophylaxis against MAC infection when the CD4 lymphocyte count declines <100/mm3.
CD4 lymphocyte counts have been measured since 1982. Absolute CD4 counts were calculated from the lymphocyte count and CD4 percent values or, more recently, counted directly. The absolute lymphocyte count was determined by an automated whole blood counter (Ortho ELT 800 with differential screen). Between 1982 and 1986, the percentages of CD4 lymphocytes were counted in Ficoll-Hypaque separated blood mononuclear cell suspensions using an EPICS V flow cytometer (Coulter Electronics). Between 1986 and 1993 a whole blood lysis method was used, and percentage of CD4 lymphocytes was analysed by flow cytometry, using either an EPICS V or an FACScan (Becton Dickinson). A monoclonal CD4 antibody, RFT4, to the p55 CD4 antigen was used in double concentration with a monoclonal CD3 antibody (OKT3 or UCHT1). Flow cytometer quality control was monitored using QC beads, and in the U.K. NEQAS external quality assurance scheme. Since 1994, absolute CD4 counts have been measured on a Cytoron absolute (ORTHO Diagnostics).
β2-Microglobulin was measured on the first available serum sample. In 57 (90.5%) cases, this was taken at the same time as the first positive anti-HIV test, and in all cases was within 3 months of this date. A commercial radial immunodiffusion (RID) method was used on all samples (NanoRID, Binding Site Ltd.) and monitored by QC controls and the U.K. NEQAS quality assurance scheme.
Kaplan-Meier estimation and the Cox proportional hazards model were used to assess the association between baseline β2-microglobulin and progression to severe immunodeficiency/AIDS. Cutoffs of β2-microglobulin (2 and 3 mg/L) were chosen on the basis that they are easily remembered round numbers. Survival times were calculated as the time from seroconversion to the date when the patient reached a CD4 count <50/mm3 or developed AIDS, whichever occurred earlier. The strong relationship between severe immunodeficiency and the development of AIDS in this cohort has previously been described (18). Survival times were right-censored if patients died or reached January 1, 1995, without having previously reached one of these endpoints. All p values are two-sided, and those from the Cox model refer to the Wald test. The proportional hazards assumption was checked by fitting the interaction with the log of time. When β2-microglobulin and age were fitted as continuous variables in the proportional hazards model, the validity of the assumption of a linear relationship with the log relative hazard was examined by assessing the effect of various translations (e.g., log, square root) and by first fitting variables as categorical variables. The interaction of the effects of β2-microglobulin and age was investigated by fitting the product of the two values as an additional covariate, after setting them to mean zero to avoid collinearity problems.
Patients were followed up for as long as 15.2 years (median 8.8 years) from seroconversion. Forty (64%) of the 63 patients reached a CD4 lymphocyte count of 50/mm3 or developed AIDS over this time. Figure 1 shows the cumulative risk of severe immunodeficiency/AIDS over time according to three groups of patients categorized by the initial β2-microglobulin level (<2 mg/L, n = 13; 2-3 mg/L, n = 36; ≥3 mg/L, n = 14). There was a clear tendency for patients with higher β2-microglobulin levels at seroconversion to progress more rapidly (p = 0.009 log rank test). When fitting β2-microglobulin as a continuous variable in a Cox proportional hazards model (table), there was a 1.68 increase in risk for every 1 mg/L increase in β2-microglobulin (p = 0.0004). The age at seroconversion, which was correlated with β2-microglobulin (Spearman correlation coefficient 0.42), also predicted the rate of development of AIDS. There was a relative risk of 1.31 per 10 year increase in age at seroconversion (p = 0.0009). When both β2-microglobulin and age were fitted together in the Cox model, the two effects were seen to act independently (p = 0.002 and p = 0.04 for β2-microglobulin and age, respectively). The above results were similar when AIDS alone was used as the endpoint (age-adjusted relative risk for β2-microglobulin 1.44; p = 0.05) and when CD4 count cutoffs of 100/mm3 (relative risk 1.50; p = 0.01) or 0/mm3 (relative risk 1.46; p = 0.04) were used to define severe immunodeficiency instead of 50/mm3. We also separately considered the predictive value of β2-microglobulin in individuals aged >18 at seroconversion. The relative risk after adjustment for age was 1.74 (p = 0.01), whereas the relative risk for age itself was 1.22 per 10-year increase (p = 0.013). If only caucasians were considered, the results were similar (relative risk 1.69; p = 0.002 per mg/L increase in β2-microglobulin), as they were if we considered only data from before 1989, when pre-AIDS zidovudine was introduced (relative risk 1.61; p = 0.02).
These results show that the serum β2-microglobulin concentration measured within a few months of HIV-1 seroconversion predicts development of severe immunodeficiency/AIDS more than 10-15 years later. This provides further evidence that the long-term prognosis of patients infected with HIV may be already predictable to some degree soon after primary infection. Fixed cofactors such as age (20,21) and HLA type (22,23) as well as severity of seroconversion illness (24-27), HIV-specific IgM and IgA response (28), anti-p24 titre around the time of seroconversion (29), and plasma viral load (30) have all been shown to be associated with the subsequent rate of progression to clinical disease. Other parameters measured at this early stage have been shown to relate to the rate of CD4 cell decline (31-38). These results, together with our own, indicate that studies of AIDS pathogenesis ideally should identify individuals around or before the time of primary HIV infection and follow them up for several years. Such studies can help us determine which viral/immune parameters most strongly predict patients' prognoses, and hence develop our understanding of which parameters are likely to be central to the pathogenic process. There is some preliminary evidence, based on quantification of CD38 expression on CD8 cells, that measures that could reflect immune activation in HIV infection predict clinical disease independently of the plasma viral load and the CD4 count (39). One could speculate, for example, that this indicates that the prevalence of activated, dividing (and hence susceptible to infection) CD4 cells is of importance for determining the level at which there is quasi-equilibrium between cell destruction and production, and therefore is as strong a marker of infection state as the plasma viral load. Besides the implications for understanding pathogenesis, this suggests that serological activation markers, which are generally cheaper to measure than the plasma viral load, may have an important continued role in patient monitoring.
Our findings have potential implications for treatment decisions, suggesting that it is possible to identify a group of patients very early in HIV infection in whom the prognosis is particularly poor—as poor, in fact, as some groups of patients with CD4 lymphocyte counts as low as 200/mm3; i.e., the 5-year risk of developing AIDS is as high in recent seroconverters with a high β2-microglobulin level as in those with a CD4 count of 200/mm3 but a low β2-microglobulin value (data not shown). This represents a group that requires close monitoring and consideration for early antiretroviral therapy, as any benefit of such treatment is more likely to out-weigh any risks than in other patients. Any future trials of early intervention with antiretroviral therapy should perhaps focus on this group. We have previously found that high immunoglobulin A levels and CD8 lymphocyte counts in the first 5 years after HIV infection also help to define such a high-risk group (35). We cannot, however, directly compare the predictive value of these markers with that of β2-microglobulin, because the latter has been measured very close to seroconversion, before the others were available.
We know of no evidence on how β2-microglobulin varies according to age in uninfected haemophiliacs. There is some evidence that β2-microglobulin levels tend to be slightly raised in non-HIV-infected haemophiliacs compared with nonhaemophiliacs (40), although no correlation with clotting factor use was noted. However, these differences are small compared with changes induced by HIV infection and are thus unlikely to affect the generalisability of our findings to other groups. Nonetheless, the raised β2-microglobulin levels found in non-HIV-infected injection drug users (41) would have to be borne in mind if reference is made to the absolute levels used here. Race could also be a factor; 80% of our patients are Caucasian, but all races are represented. β2-microglobulin levels have been shown to differ somewhat by race in HIV infection in one study (42), but not in another (43).
In summary, we have demonstrated that in men with haemophilia, the serum β2-microglobulin level measured soon after HIV-1 seroconversion helps to discriminate those who are more likely to develop severe immunodeficiency/AIDS most rapidly over the ensuing 10 years. This has implications for understanding AIDS pathogenesis and, potentially, for selecting patients who would benefit from very early treatment.
Acknowledgment: This study has received support from the Medical Research Council of the U.K. (Grant numbers SPG9219651 and G8924272).
1. Hansen PB. Olsen NV. Beta 2-microglobulin in medical disease. Ugeskrift for Laeger
2. Fahey JL, Taylor JMG, Detels R, et al. The prognostic value of cellular and serological markers in infection with HIV-1. N Engl J Med
3. Sabin CA, Phillips AN, Lee CA, et al. Beta-2 microglobulin as a predictor of prognosis in HIV-infected men with haemophilia: a proposed strategy for use in clinical care. Br J Haematol
4. Lacey JN, Forbes MA, Waugh MA, Cooper EH, Hambling MH. Serum beta 2-microglobulin and human immunodeficiency virus infection. AIDS
5. Alcabes P, Selwyn PA, Davenny K, et al. Laboratory markers and the risk of developing HIV-1 disease among injecting drug users. AIDS
6. Schwartlander B, Bek B, Skarabis H, Koch J, Burkowitz J, Koch MA. Improvement of the predictive value of CD4+ lymphocyte count by beta 2-microglobulin, immunoglobulin A and erythrocyte sedimentation rate. The Multicentre Cohort Study Group. AIDS
7. Rezza G, Titti F, Pezzotti P, et al. Anti-nef antibodies and other predictors of disease progression in HIV-1 seropositive injecting drug users. J Biol Reg Homeostat Ag
8. Kramer A, Biggar RJ, Hampl H, et al. Immunologic markers of progression to acquired immunodeficiency syndrome are time-dependent and illness-specific. Am J Epidemiol
9. Munoz A, Vlahov D, Solomon L, et al. Prognostic indicators for development of AIDS among intravenous drug users. J Acquir Immune Defic Syndr
10. Lifson AR, Hessol NA, Buchbinder SP, et al. Serum beta 2-microglobulin and prediction of progression to AIDS in HIV infection. Lancet
11. Chaisson RE, Taylor E, Margolick JB, et al. Immune serum markers and CD4 cell counts in HIV-infected intravenous drug users. J Acquir Immun Defic Syndr
12. Zangerle R, Fuchs D, Reibnegger G, Fritsch P, Wachter H. Markers for disease progression in intravenous drug users infected with HIV-1. AIDS
13. Morfeldt-Manson L, Bottiger B, Nilsson B, von Stedingk LV. Clinical signs and laboratory markers in predicting progression to AIDS in HIV-1 infected patients. Scand J Infect Dis
14. Zabay JM, Sempere JM, Benito JM, et al. Serum beta-2-microglobulin and prediction of progression to AIDS in HIV-infected injection drug users. J Acquir Immune Defic Syndr Hum Retrovirol
15. Cuthbert RJ, Ludlam CA, Tucker J, et al. Five year prospective study of HIV infection in the Edinburgh haemophiliac cohort. Br Med J
16. Anderson RE, Lang W, Shiboski S, Royce R, Jewell N, Winkelstein W Jr. Use of beta 2-microglobulin level and CD4 lymphocyte count to predict development of acquired immunodeficiency syndrome in persons with human immunodeficiency virus infection. Arch Intern Med
17. Moss AR, Bacchetti P, Osmond D, et al. Seropositivity for HIV and the development of AIDS or AIDS related condition: three year follow up of the San Francisco General Hospital cohort. Br Med J
18. Phillips AN, Lee CA, Elford J, et al. Serial CD4+ lymphocyte counts and the development of AIDS. Lancet
19. Centers for Disease Control. Revision of the CDC surveillance case definition for acquired immunodeficiency syndrome. MMWR
20. Darby SC, Rizza CR, Doll R, et al. Incidence of AIDS and excess of mortality associated with HIV in haemophiliacs in the UK: report on behalf of the directors of haemophilia centres in the UK. Br Med J
21. Phillips AN, Lee CA, Elford J, et al. More rapid progression to AIDS in older HIV-infected people: the role of CD4+ T-cell counts. J Acquir Immune Defic Syndr
22. Steel CM, Ludlam CA, Beatson D, et al. HLA haplotype A1 B8 DR3 as a risk factor for HIV-related disease. Lancet
23. Klein MR, Keet IP, D'Amaro J, et al. Associations between HLA frequencies and pathogenic features of human immunodeficiency virus type 1 infection in seroconverters from the Amsterdam cohort of homosexual men. J Infect Dis
24. Lindback S, Brostrom C, Karlsson A, Gaines H. Does symptomatic primary HIV-1 infection accelerate progression to CDC stage IV disease, CD4 count below 200 × 106
/l, AIDS, and death from AIDS? Br Med J
25. Pedersen C, Orskov Lindhardt B, Lokke Jensen B, et al. Clinical course of primary HIV infection: consequences for subsequent course of infection. Br Med J
26. Keet IPM, Krijnen P, Koot M, et al. Predictors of rapid progression to AIDS in HIV-1 seroconverters. AIDS
27. Dorrucci M, Rezza G, Vlahov D, et al. Acute retroviral syndrome among injecting drug users: clinical aspects and prognostic value. AIDS
28. Simmonds P, Beatson D, Cuthbert RJG, et al. Determinants of HIV disease progression: six year longitudinal study in the Edinburgh haemophilia/HIV cohort. Lancet
29. Cheingsong-Popov R, Panagiotidi C, Bowcock S, Aronstam A, Wadsworth J, Weber J. Relation between humoral responses to HIV gag and env proteins at seroconversion and clinical outcome of HIV infection. Br Med J
30. Mellors JW, Kingsley LA, Rinaldo CR, et al. Quantitation of HIV-1 RNA in plasma predicts outcome after seroconversion. Ann Intern Med
31. Zangerle R, Fuchs D, Sarcletti M, et al. Increased concentrations of soluble tumor necrosis factor receptor 75 but not of soluble intercellular adhesion molecule-1 are associated with the decline of CD4+ lymphocytes in HIV infection. Clin Immun Immunopathol
32. Nishanian P, Hofmann B, Wang Y, Jackson AL, Detels R, Fahey JL. Serum soluble CD8 molecule is a marker of CD8 T-cell activation in HIV-1 disease. AIDS
33. Nielsen C, Pedersen C, Lundgren JD, Gerstoft J. Biological properties of HIV isolates in primary HIV infection: consequences for the subsequent course of infection. AIDS
34. Giorgi JV, Ho H-N, Hirji K, et al. CD8+
lymphocyte activation at HIV-1 seroconversion: development of HLA-DR+
cells is associated with subsequent stable CD4+ cell levels. J Infect Dis
35. Melmed RN, Taylor JM, Detels R, Bozorgmehri M, Fahey JL. Serum neopterin changes in HIV-infected subjects: indicator of significant pathology, CD4 T cell changes, and the development of AIDS. J Acquir Immune Defic Syndr
36. Schechter MT, Craib KJ, Le TN, et al. Susceptibility to AIDS progression appears early in HIV infection. AIDS
37. Phillips AN, Sabin CA, Elford J, Bofill M, Lee CA, Janossy G. CD8 lymphocyte counts and serum immunoglobulin A levels early in HIV infection as predictors of CD4 lymphocyte depletion during 8 years of follow-up. AIDS
38. Hofmann B, Wang YX, Cumberland WG, Detels R, Bozorgmehri M, Fahey JL. Serum beta 2-microglobulin level increases in HIV infection: relation to seroconversion, CD4 T-cell fall and prognosis. AIDS
39. Giorgi JV. HIV pathogenesis and prognostic markers of HIV disease progression: update from the Multicenter AIDS Cohort Study. Abstract S41. 3rd Conference on Retroviruses and Opportunistic Infections. Washington, DC, January 28-February 1, 1996.
40. Cuthbert TJG, Ludlam CA, Steel CM, Beatson D, Peutherer JF. Immunological studies in HIV seronegative haemophiliacs: relationships to blood therapy. Br J Haematol
41. Flegg PJ, Brettle RP, Robertson JR, Clarkson RC, Bird AG. Beta-2-microglobulin levels in drug users: the influence of risk behaviour. AIDS
42. Gorter RW, Vranizan KM, Osmond DH, Moss AR. Differences in laboratory values in HIV infection by sex, race and risk group. AIDS
43. Chaisson RE, Taylor E, Margolick JB, et al. Immune serum markers and CD4 cell counts in HIV-infected intravenous drug users. J Acquir Immune Defic Syndr