Natalizumab for Multiple Sclerosis: A Case in Point for the Impact of Translational Neuroimmunology

The last two decades have witnessed a revolution in the treatment landscape for relapsing forms of multiple sclerosis (MS), with the introduction of mAb-based therapies and novel oral agents. This has been possible through translational neuroimmunology approaches in MS research and the advent of modern tools in neuroimmunology, neurobiology, and neuroimaging (1). The discovery and bench-to-bedside journey of natalizumab, as the first therapeutic mAb approved for the treatment of relapsing forms of MS, is a case in point.

MS is a chronic inflammatory, autoimmune, demyelinating, and neurodegenerative disease of the white and gray matter within the CNS. Affecting ∼2.5 million individuals worldwide, MS is the most common cause of neurologic disability in young adults (2). Approximately 85% of patients present with relapsing–remitting MS (RRMS) at diagnosis, which is characterized by alternating episodes of neurologic disability (relapses) followed by complete or partial recovery from symptoms (3). A major pathologic hallmark of MS is chronic inflammation (4). Therefore, inhibition of leukocyte trafficking across the inflamed blood–brain barrier should provide an effective anti-inflammatory therapy for MS (5). Natalizumab is a mAb that blocks the α₄-integrin–mediated leukocyte–endothelial interaction, resulting in inhibition of trafficking of lymphocytes from the blood into the CNS (6). Natalizumab was approved by the U.S. Food and Drug Administration (FDA) for the treatment of RRMS only 12 y after the discovery of its target molecule in 1992 and 7 y after the beginning of its first clinical trial, a time line that is rather fast for development of new drugs (7). This was based on convincing clinical and radiological benefits in phase III trials (6). However, a few months after its FDA approval, natalizumab was taken off the market because of a life-threatening complication: progressive multifocal leukoencephalopathy (PML). Natalizumab was later reinstated with required adherence to a strict monitoring program.

Although the bench-to-bedside journey of natalizumab highlights the value of a molecular targeted approach in developing new therapeutics for MS, natalizumab-associated PML emphasizes the importance of continued long-term postmarketing safety surveillance, sometimes warranting a bedside-to-bench switch. In this article, we review the translational journey of natalizumab along with the lessons learned from postmarketing studies.

Leukocyte trafficking across the blood–brain barrier involves overlapping sequential steps, including tethering and rolling, chemoattraction, cell adhesion, and proteolytic degradation of biological membranes (5). Rolling on endothelial cell walls allows leukocytes to identify the proper array of chemoattractants and integrin ligands (5). Natalizumab (Tysabri; Biogen Idec, Cambridge, MA) is a humanized rIgG4 mAb that blocks the α₄-integrin–mediated leukocyte–endothelial interaction, resulting in inhibition of trafficking of lymphocytes from the blood into the CNS (6). Natalizumab binds to α4-integrin on lymphocytes, as well as on myeloid cells, thereby inhibiting their migration into the CNS tissue (8).

The discovery of the target molecule for natalizumab dates back to 1992 (9); however, the interaction between endothelium and lymphocytes has been intriguing scientists since the mid-1960s, when it was first demonstrated in rats that lymphocytes enter lymph nodes through specialized postcapillary blood vessels called high endothelial venules (10). Several studies focused on mapping of molecules involved in the process of lymphocyte migration highlighted the role of integrins and selectins, particularly the cellular adhesion molecules, including ICAM and VCAM (11–13). The integrin very late activating Ag 4 (VLA-4; α_4β1-integrin) was identified as a receptor for VCAM-1, and the VCAM-1/VLA-4 ligand–receptor pair was suggested to play a major role in the recruitment of mononuclear leukocytes to inflammatory sites in vivo (14). In the early 1990s, scientists at Stanford University and Athena Neurosciences collaborated in an attempt to further identify the adhesion receptors involved in lymphocyte homing to inflamed brain endothelium, using the animal model of MS, experimental autoimmune encephalomyelitis (EAE) (9). Using an in vitro adhesion assay on frozen sections of EAE CNS tissue, they showed that leukocytes selectively bind to the lumen of the vessels, a finding consistent with an endothelial interaction. Interestingly, they found that this binding was blocked by Abs against the α₄β₁-integrin molecule and not by Abs against several other alternative adhesion receptors (9). Based on findings from immunohistochemical studies of EAE brains, it also was shown that the administration of Abs to α₄-integrin to rats with EAE improved the paralysis and prevented the accumulation of leukocytes in the CNS (9). This was the first study, to our knowledge, to provide evidence in 1992 that blocking α₄-integrins may be useful in treating inflammatory diseases of the CNS, such as MS (9). Later, the expression of adhesion molecules VLA-4 and VCAM-1 was shown in autopsied brains of patients with MS using immunohistochemical methods (15). These findings encouraged the first phase I trial of natalizumab (Antegren; Athena Neurosciences, South San Francisco, CA) (16).

A phase I, randomized, placebo-controlled study evaluating the pharmacokinetics and safety of five-level dose escalation of a single i.v. administration of natalizumab (with doses ranging from 0.03 to 3.0 mg/kg) supported the safety and tolerability of all tested doses and justified further trials (16). Later, in an open-label safety and drug interaction trial of natalizumab in combination with IFN-β1a in 38 patients with MS, further safety and tolerability data were provided, encouraging efficacy trials of this combination drug therapy (17). Aimed at evaluating the effect of natalizumab on magnetic resonance imaging (MRI) lesion activity in MS, a phase II randomized double-blind placebo-controlled trial in 72 patients with active relapsing MS showed a significant reduction in the number of new active lesions on MRI over the first 12 wk of the study but no significant difference in the second 12 wk of the trial, emphasizing the need for studies aimed at evaluating the long-term treatment effects (18). Another phase II trial targeted at assessing the effect of a single dose of i.v. natalizumab (versus placebo) soon after the onset of relapses in 180 patients with MS showed that natalizumab did not accelerate clinical recovery after relapse; however, a significant reduction in the volume of enhancing lesions at 1 and 3 wk after treatment was found (19). Around the same time, another phase II trial aimed at assessing the efficacy of two doses of monthly natalizumab (3 and 6 mg/kg) versus placebo for 6 mo, followed by a 6-mo observation in patients with relapsing MS, showed that both of the tested doses resulted in fewer inflammatory brain lesions and fewer relapses over a 6-mo period (20). Later, a phase II randomized double-blind placebo-controlled trial, Glatiramer Acetate and Natalizumab Combination Evaluation, compared the safety and tolerability of combination therapy with natalizumab plus glatiramer acetate (GA) versus GA only (21). Both groups had a similar incidence of infection and infusion reactions, and no hypersensitivity reactions were reported. Therefore, it was concluded that the combination of natalizumab and GA seemed safe and well tolerated during a 6-mo treatment period (21).

The FDA approval of natalizumab was based on two pivotal phase III trials in patients with relapsing forms of MS: the Natalizumab Safety and Efficacy in Relapsing-Remitting Multiple Sclerosis (AFFIRM) and Safety and Efficacy of Natalizumab in Combination with IFN-β1a in Patients with Relapsing-Remitting Multiple Sclerosis (SENTINEL) trials (22, 23). The AFFIRM trial was aimed at evaluating the efficacy and safety of natalizumab monotherapy in patients with RRMS, whereas the SENTINEL study was an add-on trial of natalizumab with i.m. IFN-β1a (Avonex). At the time of the pivotal phase III trials of natalizumab, IFN-β products and GA were already approved as first-line disease-modifying drugs for RRMS. Therefore, combination therapy seemed an attractive option to maximize the treatment effect.

The AFFIRM trial included a total of 942 patients randomly assigned in a 2:1 ratio to receive i.v. natalizumab 300 mg (627 patients) or placebo (315 patients) every 4 wk for up to 116 wk (22). The study inclusion criteria included the diagnosis of RRMS, age 18–50 y, a baseline expanded disability status scale (EDSS) score ≤ 5, at least one relapse in the year prior to the study, and MRI lesions consistent with MS. The primary end points were defined as the rate of clinical relapses at 1 y and the rate of sustained disability progression based on EDSS at 2 y. Patients had a median disease duration of 5 y. The results of the AFFIRM trial showed that natalizumab monotherapy significantly reduced the rate of clinical relapse at 1 y by 68% and the risk for sustained disability progression at 2 y by 42% compared with placebo. Natalizumab therapy also decreased the accumulation of new or enlarging hyperintense T2 lesions at 2 y by 83% and the number of gadolinium-enhancing lesions by 92% at 1 and 2 y. The proportion of relapse-free patients at 2 y in the natalizumab group was also significantly higher than in the placebo group (67 versus 41%, p < 0.001) (22).

The SENTINEL add-on trial included a total of 1171 patients who, despite IFN-β1a therapy, had at least one relapse during the 1-y period prior to the study. Inclusion criteria were similar to the AFFIRM study. The eligible patients were randomly assigned in a 1:1 ratio to receive continued i.m. IFN-β1a therapy in combination with natalizumab 300 mg (589 patients) or placebo (582 patients) i.v. every 4 wk for up to 116 wk (23). The definitions of primary and secondary end points were similar to the AFFIRM study. The results of the SENTINEL trial showed that combination therapy with natalizumab resulted in a 55% relative reduction in annualized relapse rate (0.34% compared with 0.75% with IFN-β1a), an 83% reduction in the number of new or enlarging T2 lesions at 2 y, an 89% reduction in the number of gadolinium-enhancing lesions, and a 24% reduction in the relative risk for sustained disability progression at 2 y (23).

The similar inclusion criteria and end point definitions in the AFFIRM and SENTINEL trials made it possible to draw clinically useful conclusions from the two studies. FDA granted accelerated approval for natalizumab in November 2004 following priority review of the data from these two trials.

Because natalizumab inhibits trafficking of lymphocytes from the blood into the CNS, its clinical benefit can be postulated in secondary progressive MS, in which CNS-compartmentalized inflammatory processes may predominate (24). Also, previous studies suggested evidence of improved tissue integrity and enhanced axonal metabolism in some natalizumab-treated RRMS patients (25, 26). In an open-label phase IIa proof-of-concept trial of natalizumab in patients with progressive MS, results showed reductions in CSF inflammatory biomarkers, specifically osteopontin, CXCL13, and matrix metalloproteinase-9, as well as CSF markers of axonal damage and demyelination (neurofilament L chain and myelin basic protein), thereby supporting the potential benefit of natalizumab in progressive MS (27). However, the phase III ASCEND trial of natalizumab in patients with secondary progressive MS showed negative results. The study did not achieve its composite primary end point of disability progression, defined as the percentage of patients with confirmed progression of disability on one or more of the following components: EDSS, timed 25-foot walk, and nine-hole peg test (28). These results underscore our incomplete understanding of the progressive forms of MS and the different neuroinflammatory components, both spatially and qualitatively, in progressive MS compared with RRMS (29).

PML is a rare opportunistic infection of the CNS caused by a human polyomavirus, JC, named after the initials of the patient from whom it was first isolated in 1971 (30). Infection of oligodendrocytes by John Cunningham virus (JCV) results in demyelination of CNS. PML is strongly associated with, although not limited to, immunosuppressed status and predominantly occurs in patients with AIDS. Approximately half of the world’s population is estimated to have been infected with JCV at a young age; however, it becomes latent and almost never causes disease in immunocompetent individuals (31).

Only 3 mo after the first FDA approval of natalizumab, Biogen Idec and Elan voluntarily withdrew it from the market in February 2005. This was based on reports of three cases of PML following natalizumab therapy. Two cases were MS patients in the SENTINEL trial who received natalizumab in combination with IFN-β1a (32, 33); the third was a patient with Crohn’s disease who received natalizumab in a clinical trial of Crohn’s disease (34). These three cases were reported among ∼3000 patients who had participated in trials of natalizumab for MS or Crohn’s disease. Therefore, the overall risk for developing PML with natalizumab use was estimated at 1 in 1000 at that time (35). It is challenging to comment in retrospect whether it was an error to grant accelerated approval for Tysabri. Given the rarity of PML, it is hard to imagine that it would have necessarily been detected in a full approval process. Yet, the recall of natalizumab after only 3 mo raised valid concerns about the fast tracking of new drugs by the FDA.

PML also was reported with the use of immunosuppressant agents, such as efalizumab (anti-CD11a), rituximab (anti-CD20), infliximab (anti–TNF-α), fingolimod (sphingosine-1-phosphate receptor modulator), alemtuzumab (anti-CD52), dimethyl fumarate (activator of the antioxidative transcription factor NF [erythroid-derived 2]-like 2 pathway), and mycophenolate mofetil (inhibitor of inosine monophosphate dehydrogenase), in various immune-mediated disorders (28, 29). To our knowledge, no case of PML with the use of rituximab, teriflunomide, and alemtuzumab was reported in patients with MS. Although there are many unknown factors with regard to our understanding of PML, the occurrence of natalizumab-associated PML in patients with MS may suggest that natalizumab prevents the entry of JCV cytotoxic lymphocytes into the CNS, and a reduction in the number of APCs may contribute to the development of PML (29). The release of CD19⁺ CD10⁺ pre-B cells (which can be latently infected with JCV) after natalizumab administration and alteration of transcriptional expression patterns of several genes involved in B cell activation and differentiation might be other contributing factors (29, 30).

The most common presenting symptoms for PML include changes in cognition, personality, and motor function and, less commonly, seizures (36). The median treatment duration to the onset of PML symptoms was estimated as 25 mo (range: 6–80 mo), based on a postmarketing study of natalizumab (36). Clinical vigilance is the key to an early diagnosis of PML. In case of clinical suspicion of natalizumab-associated PML, natalizumab must be discontinued, an MRI should be obtained to look for radiological features suggestive of PML, and a CSF sample should be taken to assess JCV status. Typical PML lesions on brain MRI are confluent, bilateral but asymmetric, and involve the subcortical supratentorial white matter, with a low signal on T1-weighted images and a high signal on T2-weighted and diffuse-weighted images (37). Recently, a punctuate pattern of MRI lesions was suggested as a useful imaging feature in the presymptomatic stage of PML (38).

Management of natalizumab-associated PML requires plasma exchange or immunoabsorption to accelerate the removal of natalizumab and to re-establish immune surveillance of the CNS (39). The effective clearance of natalizumab and restoration of immune surveillance may result in a paradoxical worsening of neurologic deficits and MRI lesions, a potentially life-threatening condition called immune reconstitution inflammatory syndrome (IRIS) (40), which can occur in natalizumab-associated PML after the cessation of natalizumab. Histopathologically, IRIS was reported to be characterized by a marked inflammatory infiltrate with a predominance of CD8⁺ T cells and high numbers of plasma cells within lesions, as well as neighboring gray and white matter (41). Corticosteroid therapy provides a modest benefit in natalizumab-associated PML-IRIS (40). Further studies are needed to understand the relationship among IRIS, immunoregulatory function, and cell trafficking during and after stopping natalizumab therapy (42).

In June 2006, natalizumab was reintroduced to the market and reapproved by the FDA as a monotherapy for highly active MS, but with a black box warning describing an increased risk for PML with its use and required adherence to a special program developed by the FDA known as Tysabri Outreach Unified Commitment to Health, which is only available in the United States (43). This is a restricted distribution program aimed at minimizing the risks associated with the use of natalizumab. Only prescribers and patients enrolled in the Tysabri Outreach Unified Commitment to Health program may prescribe and receive natalizumab. Also, the dispensing and infusion of natalizumab can only be done through certain pharmacies and infusion centers that are authorized by the program (43).

Three factors (alone or in combination) were suggested by postmarketing analyses to contribute to the risk for natalizumab-associated PML: positive status of anti-JCV Abs in serum, prior exposure to immunosuppressants, and longer duration of exposure to natalizumab therapy (44, 45). Quantitative JCV serology to better predict the risk for PML was the subject of multiple studies. Some studies suggested that the onset of PML can be accompanied by increasing anti-JCV Abs in serum (46). The anti-JCV Ab index (47) may help to further define the risk for PML in anti-JCV Ab–positive MS patients with no prior immunosuppressant use who are treated with natalizumab (48). Table I summarizes the estimated risk for PML associated with natalizumab use based on the three risk factors mentioned above (49, 50). Mitigation strategies need to be incorporated based on those factors and the associated risk estimates. The JCV Ab serology status generally needs to be retested every 6 mo. JCV Ab–negative patients with a previous history of exposure to immunosuppressants may need to be retested every 3 mo. A recent study estimated a seroconversion rate of 8.5–10.3%/y in natalizumab-treated patients (51). A seroconversion warrants revisiting the risk/benefit profile of natalizumab therapy.

It is questionable whether the introduction of the risk-mitigation strategies has been a success with respect to its intended purpose, given that the rate of natalizumab-associated PML does not seem to have decreased (52). In fact, the most recently published data indicate that the absolute number of cases and the relative risk for PML have actually doubled (49, 50). The reason for this possible failure is not completely understood. It appears that many neurologists may have difficulty interpreting the risk-stratification algorithm. The primary focus may be on the JCV serostatus when assessing the overall risk for natalizumab-associated PML, while neglecting other risk factors. Also, the duration of natalizumab therapy should not be undervalued. Finally, disease reactivation following cessation of natalizumab (see below) may sometimes lead to a reluctance to discontinue natalizumab, even if the risk for PML is considered high.

Patients may need to discontinue natalizumab treatment to reduce the risk for PML or for other reasons, such as pregnancy. Some lines of evidence indicate the return of disease activity or rebound when natalizumab is discontinued (53, 54). This is while other studies suggest no dramatic increase in disease activity 6 or 14 mo after treatment cessation (20, 55), a post hoc analysis of natalizumab-treated patients in the AFFIRM, SENTINEL, and Glatiramer Acetate and Natalizumab Combination Evaluation trials prior to its suspension in 2005, including monthly follow-up for 8 mo after the cessation of natalizumab, showed that, although disease activity seemed to increase after 3 mo of natalizumab discontinuation, the average disease activity did not exceed that of the pretreatment status (56). This finding may indicate the lack of a true rebound activity; however, it remains possible that a subgroup of patients may be more susceptible to show highly active disease once natalizumab treatment is discontinued (57). Assessment of the Th17 cell/IL-17 axis was suggested recently to help predict disease reactivation after natalizumab cessation (58).

Abs against natalizumab were detected in 9% of patients who received natalizumab monotherapy in the AFFIRM study, with 3% classified as transiently positive and 6% classified as persistently positive (22). The definition of a persistent positive result was based on at least two visits that were ≥6 wk apart (22). In the SENTINEL study, 12% of patients treated with natalizumab combined with IFN-β1a developed anti-natalizumab Abs (with 6% classified as persistently positive) (23). Patients with persistently positive Abs to natalizumab were found to have loss of clinical efficacy, as measured by relapse rate, disability progression, and MRI activity (59), and an increase in infusion-related adverse events (22, 23). However, in patients with transiently positive Abs, efficacy was achieved after ∼6 mo of therapy, when patients’ Ab status became negative (59). It is recommended that natalizumab-treated patients with ongoing disease activity or persistent infusion-related adverse effects be assessed for Abs against natalizumab (59, 60).

Advances in translational immunology have generated novel approaches to immunotherapy for MS. The development of natalizumab, its bench-to-bedside journey, and the lessons learned from its postmarketing experience support the value of a molecular targeted approach in an attempt to modify the course of MS and emphasize the risks for unexpected adverse events associated with novel biological therapeutics. Although a molecular targeted approach may theoretically indicate that the remainder of the immune system would be left to do its job, thereby decreasing the risk for opportunistic infections, such as PML, there is still a long way to go toward Ag-specific therapies in MS, given the need for a better understanding of the key Ags driving disease progression.

Natalizumab remains a highly effective therapy for relapsing forms of MS. However, there is more to be learned regarding its long-term safety profile. Given the estimated risk for natalizumab-associated PML, several new cases will probably be identified in the future. This provides an important opportunity to better understand the pathobiology of JCV, especially because new cases of PML also were reported with some of the recently developed oral immunomodulatory therapies for MS (61, 62).

O.S. serves on the editorial boards of JAMA Neurology, Multiple Sclerosis Journal, and Therapeutic Advances in Neurological Disorders. He has served on data monitoring committees for Pfizer and TG Therapeutics without monetary compensation. O.S. collaborated with Medscape on educational initiatives, advised Genentech and Genzyme, participated in a Teva-sponsored meeting, and consulted for Navigant Consulting. O.S. currently receives grant support from Teva Pharmaceuticals and Opexa Therapeutics, and is funded by a Merit Review grant (Federal Award Document Number I01BX001674) from the U.S. Department of Veterans Affairs, Biomedical Laboratory Research and Development. A.S. has no financial conflicts of interest.