Altered CD4+/CD8+ T-Cell Ratios in Cerebrospinal Fluid of Natalizumab-Treated Patients With Multiple Sclerosis

Background  Treatment with natalizumab, a monoclonal antibody against the adhesion molecule very late activation antigen 4, an α4β₁ integrin, was recently associated with the development of progressive multifocal leukoencephalopathy, a demyelinating disorder of the central nervous system caused by JC virus infection.

Objective  To test the effect of natalizumab treatment on the CD4⁺/CD8⁺ T-cell ratios in cerebrospinal fluid (CSF) and peripheral blood.

Design  Prospective longitudinal study.

Setting  Academic and private multiple sclerosis centers.

Patients  Patients with multiple sclerosis (MS) treated with natalizumab, untreated patients with MS, patients with other neurologic diseases, and human immunodeficiency virus–infected patients.

Main Outcome Measures  CD4⁺ and CD8⁺ T cells were enumerated in CSF and peripheral blood. The mean fluorescence intensity of unbound α4 integrin on peripheral blood CD4⁺ and CD8⁺ T cells was analyzed before and after natalizumab therapy.

Results  Natalizumab therapy decreased the CSF CD4⁺/CD8⁺ ratio of patients with MS to levels similar to those of human immunodeficiency virus–infected patients. CD4⁺/CD8⁺ ratios in peripheral blood in patients with MS progressively decreased with the number of natalizumab doses, but they remained within normal limits. Six months after the cessation of natalizumab therapy, CSF CD4⁺/CD8⁺ ratios normalized. The expression of unbound α4 integrin on peripheral blood T cells decreases with natalizumab therapy and was significantly lower on CD4⁺ vs CD8⁺ T cells.

Conclusions  Natalizumab treatment alters the CSF CD4⁺/CD8⁺ ratio. Lower expression of unbound α4 integrin on CD4⁺ T cells is one possible mechanism. These results may have implications for the observation that some natalizumab-treated patients with MS developed progressive multifocal leukoencephalopathy.

Multiple sclerosis (MS) is an inflammatory demyelinating disorder of the central nervous system (CNS) of unknown origin.¹ Natalizumab (Tysabri; Biogen Idec Inc, Cambridge, Mass, and Elan Corp, Dublin, Ireland) is a humanized monoclonal antibody designed to bind very late activation antigen 4, which is an α4β₁ (CD49d/CD29) integrin that serves as an adhesion molecule and is expressed by all leukocytes except neutrophils. Natalizumab was intended to prevent ingress of leukocytes into the CNS² and other target tissues by physically blocking the interaction of very late activation antigen 4 with its ligands, vascular cell adhesion molecule 1³ or the CS-1 fragment of fibronectin.⁴ The clinical effectiveness of natalizumab in MS, Crohn disease, and rheumatoid arthritis was evaluated in approximately 3000 patients. Based on the results of a phase 2 clinical trial⁵ and data from 2 phase 3 studies (the AFFIRM [Natalizumab Safety and Efficacy in Relapsing Remitting Multiple Sclerosis] monotherapy trial⁶ and the SENTINEL [The Safety and Efficacy of Natalizumab in Combination With Interferon Beta-1a in Patients With Relapsing Remitting Multiple Sclerosis] add-on trial with interferon beta-1a [Avonex; Biogen Idec Inc]⁷), the Food and Drug Administration approved natalizumab for the treatment of relapsing forms of MS on November 23, 2004 (http://www.fda.gov/bbs/topics/news/2004/NEW01141.html). On February 28, 2005, the manufacturers of natalizumab announced the voluntary withdrawal of this agent from the market, and the use of natalizumab in clinical trials was halted after 2 patients with MS and 1 patient with Crohn disease who had received natalizumab in clinical trials were diagnosed as having progressive multifocal leukoencephalopathy (PML), a demyelinating disorder of the CNS caused by infection with the human JC polyomavirus.^8-10

This study examined cerebrospinal fluid (CSF) and peripheral blood CD4⁺/CD8⁺ ratios in patients treated with natalizumab compared with controls (including patients infected with human immunodeficiency virus [HIV]). Specifics regarding absolute numbers of different lymphocyte phenotypes in CSF and peripheral blood of natalizumab-treated patients are reported elsewhere and do not include HIV-infected controls.ext-link ext-link-type="bibr" rid="REF-NOC60065-11"/ We demonstrate that natalizumab therapy significantly alters the CD4⁺/CD8⁺ ratio in CSF, reducing it to levels comparable with those in HIV-infected individuals, a patient group known to be at high risk for PML. In addition, we show that the expression of unbound α4 integrin on T cells decreases with natalizumab therapy and is significantly lower on CD4⁺ vs CD8⁺ T cells.

For analysis of CD4⁺/CD8⁺ ratios, 15 CSF and 23 blood samples from patients with MS treated with natalizumab were analyzed at study enrollment, and 13 samples were analyzed at 6-month follow-up. Control subjects included 17 patients with other, noninflammatory neurologic diseases (headache, polyneuropathy, and normal pressure hydrocephalus), 35 patients with clinically definite relapsing-remitting MSext-link ext-link-type="bibr" rid="REF-NOC60065-12"/ who did not receive natalizumab, and 16 HIV-infected patients. Natalizumab-treated patients with MS were recruited from the University of Texas Southwestern Medical Center at Dallas Multiple Sclerosis Center and at the Multiple Sclerosis Center at Texas Neurology. Controls were recruited from Heinrich Heine University, Düsseldorf, Germany.

Eight patients with clinically definite relapsing-remitting MSext-link ext-link-type="bibr" rid="REF-NOC60065-12"/ who were transitioning into the open-label phase of the natalizumab clinical trial program at the Montreal Neurological Institute, Quebec, provided peripheral blood samples for quantitation of unbound α4 integrin. All the patients were studied during treatment with the standard 300-mg monthly intravenous regimen of natalizumab. For each patient, blood samples were collected just before and 1 ½ hours after each treatment.ext-link ext-link-type="bibr" rid="REF-NOC60065-13"/ Informed consent was obtained from all patients, and the study procedures were approved by the respective institutional review boards.

Neubauer or Fuchs-Rosenthal counting chambers were used to count total cell numbers in CSF and peripheral blood mononuclear cells by means of light microscopy. Total CSF cell count was determined manually by trypan blue exclusion or methylene blue staining; the lower limit of detection was 250 cells/mL.ext-link ext-link-type="bibr" rid="REF-NOC60065-14"/

For investigation of CD4⁺/CD8⁺ ratios, peripheral blood mononuclear cells were isolated by means of density gradient centrifugation (Ficoll-Paque; Amersham Pharmacia, Uppsala, Sweden). The CSF cells were collected by means of centrifugation. Cell staining for flow cytometry was performed according to published methods.ext-link ext-link-type="bibr" rid="REF-NOC60065-15"/ A flow cytometer (FACSCalibur; BD Biosciences, San Jose, Calif) was used for data acquisition. Flow cytometry data were analyzed using a software program (CellQuestPro; BD Biosciences, or Flowjo; Tree Star Inc, Ashland, Ore). Flow cytometry data for CD4⁺/CD8⁺ ratios were reviewed by 3 of us (O.S., S.C., and B.H.).

For analysis of unbound α4 integrin (CD49) on CD4⁺ and CD8⁺ T cells, 3-color staining was performed by incubating whole blood with saturating amounts of directly conjugated monoclonal antibodies against CD49d, CD3, CD4, and CD8 (BD PharMingen, San Diego, Calif).ext-link ext-link-type="bibr" rid="REF-NOC60065-13"/ Isotype-matched antibodies were used as negative controls (BD PharMingen).ext-link ext-link-type="bibr" rid="REF-NOC60065-13"/ Flow cytometry data were acquired using a BD FACScan (Becton Dickinson, Franklin Lakes, NJ), and the mean fluorescence intensity of α4 integrin was analyzed by a blinded operator using FlowJo software.

The Mann-Whitney test was used to compare CD4⁺/CD8⁺ ratios in CSF and peripheral blood. The strength of associations between the CD4⁺/CD8⁺ T-cell ratio in peripheral blood and the number of natalizumab doses received was analyzed using the Spearman rank correlation coefficient. The α4 integrin expression levels on T cells were compared using paired t tests. P<.05 is considered statistically significant.

Participant characteristics are given in the Table. All patients with MS had relapsing-remitting disease. The duration of MS was similar in natalizumab-treated patients with MS and in controls.

The CSF CD4⁺/CD8⁺ ratios in patients with other, noninflammatory neurologic diseases and patients with MS who had not received natalizumab were comparable with published data (Figure 1A).^16-19 In contrast, CSF CD4⁺/CD8⁺ ratios in patients with MS who had received natalizumab were very low (Figure 1A) and were not significantly different from those in HIV-infected controls (P = .8). There were no statistically significant differences in CD4⁺/CD8⁺ T-cell ratios between patients enrolled in the AFFIRM trial and those enrolled in the SENTINEL trial (Figure 1B). A patient who received a single dose of natalizumab had a CSF CD4⁺/CD8⁺ ratio as low as a patient who received 41 doses. When examined 6 months after the cessation of natalizumab therapy, the CSF CD4⁺/CD8⁺ ratio had normalized (Figure 1A).

Peripheral blood CD4⁺/CD8⁺ ratios in patients with other neurologic diseases, untreated patients with MS, and natalizumab-treated patients with MS were within normal limits. As expected, CD4⁺/CD8⁺ ratios were low in HIV-infected patients (Figure 2A). Peripheral blood CD4⁺/CD8⁺ ratios decreased significantly with increasing numbers of natalizumab doses but remained within normal limits (Figure 2B). Six months after the cessation of natalizumab therapy, CD4⁺/CD8⁺ T-cell ratios in peripheral blood remained normal (Figure 2B).

The expression of unbound α4 integrin on CD4⁺ and CD8⁺ T cells decreased significantly after natalizumab therapy (Figure 3). CD4⁺ T cells expressed significantly less unbound α4 integrin before and after natalizumab therapy than CD8⁺ T cells (Figure 3).

The occurrence of PML in 3 of approximately 3000 natalizumab-treated patients was unexpected. Progressive multifocal leukoencephalopathy is a demyelinating CNS disorder caused by active infection of oligodendrocytes by JC virus.ext-link ext-link-type="bibr" rid="REF-NOC60065-20"/ JC virus is presumably acquired in childhood, and the seroprevalence of antibodies to the virus is 80% to 90% in adults.ext-link ext-link-type="bibr" rid="REF-NOC60065-21"/ JC virus remains latent in the host, although consensus has not been reached regarding the site of latency, and peripheral blood lymphocytes, particularly B cells, kidney, and brain, have all been suggested.ext-link ext-link-type="bibr" rid="REF-NOC60065-22 REF-NOC60065-23 REF-NOC60065-24 REF-NOC60065-25 REF-NOC60065-26 REF-NOC60065-27 REF-NOC60065-28"/ Progressive multifocal leukoencephalopathy results when the virus reactivates, and this most often occurs in patients with underlying immunodeficiency, particularly those infected with HIV.ext-link ext-link-type="bibr" rid="REF-NOC60065-29"/ It was for this reason that we compared CD4⁺/CD8⁺ ratios in CSF and peripheral blood of natalizumab-treated patients with MS with those in HIV-infected patients.

Although the patient numbers were small, the data indicate that natalizumab therapy results in a striking decrease in the CSF CD4⁺/CD8⁺ ratio. The biological significance of a decreased CSF CD4⁺/CD8⁺ ratio in the context of a cellular immune response against foreign antigens, such as JC virus, is unknown. In patients with AIDS, CD4⁺/CD8⁺ ratios have long been used as a disease marker. Our data show that CD4⁺ and CD8⁺ T cells expressed significantly less unbound α4 integrin before and after natalizumab therapy and that the effect on CD4⁺ T cells was greater than on CD8⁺ T cells. This suggests that the differential surface expression of unbound α4 integrin on CD4⁺ and CD8⁺ T cells may contribute to the differential effect of natalizumab on these lymphocyte subsets; probably only the unbound (available) fraction of very late activation antigen 4 on the cell surface will facilitate the migration of these cells into tissues. The total amount of unbound α4 integrin is one possible explanation for altered CD4⁺/CD8⁺ ratios in CSF. Another is that a threshold of total cell surface expression of unbound α4 integrin is required for effective migration of leukocytes from peripheral blood into target tissues and that it differs for CD4⁺ T cells vs CD8⁺ T cells.

Alteration of the CD4⁺/CD8⁺ ratio in CSF by natalizumab may provide information on the pathogenesis of MS. Although antigen-specific CD4⁺ T cells have long been implicated in the pathogenesis of MS, several recent studiesext-link ext-link-type="bibr" rid="REF-NOC60065-30 REF-NOC60065-31"/ demonstrated clonal and oligoclonal accumulation of CD8⁺ T cells in the CNS of patients with MS. However, the initiation and perpetuation of most antigen-specific CD8⁺ T-cell responses requires the help of CD4⁺ T cells.ext-link ext-link-type="bibr" rid="REF-NOC60065-32"/ The clinical efficacy of natalizumab and the present data may suggest that CD4⁺ T cells do indeed play a critical role in the inflammatory cascade of MS.

The decrease in CD4⁺/CD8⁺ ratios in peripheral blood with increasing doses of natalizumab may reflect a differential sequestration of CD4⁺ and CD8⁺ T cells into various tissues. The reduced CD4⁺/CD8⁺ ratios in peripheral blood across time do not correlate with CD4⁺/CD8⁺ ratios in CSF, which were unaffected by the total number of natalizumab doses. Prolonged treatment with natalizumab may affect the circulation of leukocytes from peripheral tissues into the blood and may explain this discrepancy. Although the mechanism remains undefined at present, the implications for immune control of CNS infections are potentially important and suggest that prolonged, uninterrupted natalizumab therapy may eventually alter systemic cellular immune responses.

Interpretation of these data should not lead to the conclusion that natalizumab therapy is the biological equivalent of an infection with HIV. These data point out at least 1 significant difference between natalizumab-treated patients with MS and HIV-infected patients: although CD4⁺/CD8⁺ ratios in peripheral blood may be low in all stages of HIV infection, they remain within normal limits in patients who receive a limited course of natalizumab therapy.

Treatment of patients with MS with natalizumab leads to an immediate decrease in the CD4⁺/CD8⁺ ratio in CSF but not in peripheral blood. Although patients with HIV had similarly low CSF CD4⁺/CD8⁺ ratios compared with natalizumab-treated patients, patients with HIV also had low peripheral blood CD4⁺/CD8⁺ ratios. The biological implications of a low CSF CD4⁺/CD8⁺ ratio are not defined. However, the differential effect of natalizumab on specific lymphocyte subsets in the CNS, as reflected by a differential decrease in α4 integrin on CD4⁺ compared with CD8⁺ T cells, and alteration of the CSF CD4⁺/CD8⁺ ratio could pose an increased risk of opportunistic infections in natalizumab-treated patients.

Correspondence: Olaf Stüve, MD, PhD, Neurology Section, VA North Texas Health Care System, Medical Service, 4500 S Lancaster Rd, Dallas, TX 75216 (olaf.stuve@utsouthwestern.edu).

Accepted for Publication: March 29, 2006.

Author Contributions:Study concept and design: Stüve, Marra, Cravens, Frohman, Phillips, Arendt, Hemmer, Monson, and Racke. Acquisition of data: Stüve, Marra, Bar-Or, Niino, Cravens, Cepok, Frohman, Phillips, Arendt, Cook, Grand'Maison, Hemmer, Monson, and Racke. Analysis and interpretation of data: Stüve, Marra, Bar-Or, Cravens, Frohman, Jerome, Hemmer, Monson, and Racke. Drafting of the manuscript: Stüve, Marra, Bar-Or, Niino, Cravens, Cepok, Frohman, Phillips, Jerome, Hemmer, Monson, and Racke. Critical revision of the manuscript for important intellectual content: Marra, Frohman, Arendt, Cook, Grand'Maison, and Racke. Statistical analysis: Stüve, Bar-Or, Cravens, Cepok, Cook, and Hemmer. Obtained funding: Frohman, Hemmer, Monson, and Racke. Administrative, technical, and material support: Stüve, Marra, Cravens, Frohman, Phillips, Arendt, Cook, Grand'Maison, Monson, and Racke. Study supervision: Stüve, Bar-Or, Niino, Cepok, Frohman, Cook, Monson, and Racke.

Financial Disclosure: Dr Frohman has received speaker honoraria from Biogen/Teva.

Funding/Support: This study was supported by a start-up grant from the Dallas VA Research Corporation and a New Investigator Award grant from the Veterans Administration, VISN 17 (Dr Stüve); grants NS 37513 and NS 44250 from the National Institutes of Health and grant RG 2969-B-7 from the National Multiple Sclerosis Society (Dr Racke); by grants He2386/4-1 and He2386/4-2 from the Deutsche Forschungsgemeinschaft (Drs Hemmer and Cepok); by the Adult AIDS Clinical Trials Group funded by grants AI 38858 and AI 27664 from the National Institute of Allergy and Infectious Diseases; and by grant NS 40993 from the National Institutes of Health (Dr Monson).

Additional Information: The AFFIRM and the SENTINEL trials were sponsored by Biogen Idec Inc and Elan Corp.

Acknowledgment: We thank the patients who participated in this study; Friederike Vogel, MD, Gina Remington, RN, Jane Lee, BSc, Janey Phillips, RN, Jill Fowler, RN, Nancy Perna, RN, and Subir Sinha, BSc, for assistance in data acquisition; and Lauren Tantalo, BSc, and April Colina, BSc, for technical assistance.