Articles
`
`TNF neutralization in MS
`Results of a randomized, placebo-controlled
`multicenter study
`
`The Lenercept Multiple Sclerosis Study Group and The University of British Columbia
`MS/MRI Analysis Group*
`
`Article abstract—Objective: A double-blind, placebo-controlled phase II study was conducted in 168 patients, most with
`relapsing-remitting MS, to evaluate whether lenercept would reduce new lesions on MRI. Background: Tumor necrosis
`factor (TNF) has been implicated in MS pathogenesis, has been identified in active MS lesions, is toxic to oligodendrocytes
`in vitro, and worsens the severity of experimental allergic encephalomyelitis (EAE) in animals. Lenercept, a recombinant
`TNF receptor p55 immunoglobulin fusion protein (sTNFR-IgG p55), protects against EAE. Methods: Patients received 10,
`50, or 100 mg of lenercept or placebo IV every 4 weeks for up to 48 weeks. MRI scans and clinical evaluations were
`performed at screening, at baseline, and then every 4 weeks (immediately before dosing) through study week 24. Results:
`There were no significant differences between groups on any MRI study measure, but the number of lenercept-treated
`patients experiencing exacerbations was significantly increased compared with patients receiving placebo (p 5 0.007) and
`their exacerbations occurred earlier (p 5 0.006). Neurologic deficits tended to be more severe in the lenercept treatment
`groups, although this did not affect Expanded Disability Status Scale scores. Anti-lenercept antibodies were present in a
`substantial number of treated patients; serum lenercept trough concentrations were detectable in only a third. Adverse
`events that increased in frequency in treated patients included headache, nausea, abdominal pain, and hot flushes.
`Conclusions: Lenercept failed to be beneficial, but insight into the role of TNF in MS exacerbations was gained.
`NEUROLOGY 1999;53:457–465
`
`MS is believed to be an inflammatory autoimmune
`disorder of the CNS with unknown myelin compo-
`nents as target. A number of findings have suggested
`that tumor necrosis factor (TNF) contributes to prop-
`agating the inflammatory response and to tissue in-
`jury in MS. In autopsy specimens, TNF has been
`demonstrated within active MS foci.1 TNF has been
`shown to have a direct toxic effect against oligoden-
`drocytes and a proliferation-inducing effect on astro-
`cytes in in vitro studies.2,3 In patients with MS,
`elevated TNF levels in the serum and CSF have been
`correlated in some studies with disease progres-
`sion.4,5 Blood mononuclear cells from MS patients,
`studied just before an exacerbation, secrete greater
`amounts of TNF in response to mitogen stimulation
`than at other times.6 Blood mononuclear cells from
`MS patients with active disease express higher lev-
`els of TNF mRNA than do cells from MS patients
`with stable disease or healthy controls.7,8
`
`Studies of experimental autoimmune encephalo-
`myelitis (EAE) have profoundly shaped views of MS
`pathogenesis. EAE is an autoimmune disease with
`pathologic features reminiscent of those seen in MS.
`TNF treatment worsens EAE,9 and TNF neutraliza-
`tion by anti-TNF antibody treatment consistently
`protects animals from EAE.10-12 Similarly, TNF cap-
`ture by lenercept, a TNFa receptor–immunoglobulin
`G (IgG)1 fusion protein, protects in EAE.13 The above
`indicates that TNF functions in EAE as a proinflam-
`matory mediator and suggests that TNF depletion
`might be protective in MS. The hypothesis that neu-
`tralization of TNF may reduce or halt MS progres-
`sion was evaluated in a phase II randomized,
`multicenter, placebo-controlled study of three doses
`of lenercept (sTNFR-IgG p55). Lenercept is a dimeric
`recombinant protein molecule built from two copies
`of the 55 kDa TNF receptor extracellular domain
`fused to a fragment of the human immunoglobulin
`
`See also pages 444 and 466
`
`*See the Appendix on page 464 for a listing of members of The Lenercept Multiple Sclerosis Study Group and The University of British Columbia MS/MRI
`Analysis Group.
`Funded by F. Hoffmann-LaRoche Ltd., Basel, Switzerland.
`Received September 11, 1998. Accepted in final form April 8, 1999.
`Address correspondence and reprint requests to Dr. Barry G.W. Arnason, University of Chicago, Department of Neurology, 5841 S. Maryland Ave., MC 2030,
`Chicago, IL 60637; e-mail: barnason@drugs.bsd.uchicago.edu
`
`Copyright © 1999 by the American Academy of Neurology 457
`
`Biogen Exhibit 2087
`Mylan v. Biogen
`IPR 2018-01403
`
`Page 1 of 9
`
`

`

`IgG1 heavy chain.14,15 In accordance with recently
`published recommendations, efficacy was assessed
`by means of MRI.16
`
`Methods. Patients. A total of 168 patients with clini-
`cally definite or laboratory supported definite MS were
`enrolled in a double-blind, placebo-controlled study. The
`study was approved by the institutional review boards of
`the participating centers, and all subjects gave informed
`consent. At enrollment, patients were between the ages of
`18 and 55 years and had an Expanded Disability Status
`Scale (EDSS) score #5.5. For patients with an EDSS score
`#3, the history of MS was limited to a maximum duration
`of 10 years. All patients had at least two exacerbations
`within the previous 2 years, but were clinically stable for 4
`weeks before the screening MRI and during the 4 weeks
`between screening and study entry. With the exception of
`glucocorticoids, any prior administration of agents with a
`putative effect on MS (including interferons, cyclophospha-
`mide, or azathioprine) led to exclusion. Treatment with
`glucocorticoids was not permitted within a 4-week period
`before the screening visit or between screening and base-
`line. Other exclusion criteria included the diagnosis of pri-
`mary progressive MS and inability to undergo MRI
`scanning. A randomization list with treatment blocks (four
`patients per block) was computer generated by Hoff-
`mann–La Roche (Basel, Switzerland) for each investiga-
`tion site. During the conduct of the study, the rando-
`mization list was available only to the Safety Review
`Board (SRB) members (see below). A limited number of
`Roche staff were unblinded at the time of the first analysis
`of efficacy as defined in the protocol. On study termination,
`each investigational site was provided with the site-
`specific randomization code.
`Eligible patients were randomized to 10, 50, or 100 mg
`of lenercept or to placebo, administered IV every 4 weeks.
`Study duration was 48 weeks, consisting of a 24-week,
`double-blind treatment period and a 24-week follow-up pe-
`riod. Of the 168 patients randomized to treatment, one
`patient (randomized to placebo) was identified as ineligible
`prior to the baseline visit; this patient did not receive
`treatment, have a baseline MRI scan, or return for follow-
`up. For the 167 patients who received treatment, compli-
`ance to treatment and study procedures was excellent.
`During the first 24 weeks, 99% (991/1002) of all planned
`doses were administered and 98% (1303/1336) of all MRI
`scan sets were performed.
`During the follow up period (study weeks 25–48), pa-
`tients could continue double-blind treatment on a volun-
`tary basis and 130 elected to do so. Those patients who
`opted not to continue treatment remained in the study and
`were followed on an intent-to-treat basis. For the full study
`duration, 10 doses (median) were administered to each
`treatment group.
`For safety purposes, three cohorts of up to 16 patients
`were enrolled in an ascending-dose design at approxi-
`mately 6-week intervals. The first cohort was randomized
`to placebo or 10 mg of lenercept whereas subsequent co-
`horts were randomized to placebo or 50 mg and finally to
`placebo or 100 mg of lenercept. An independent SRB eval-
`uated the unblinded study data before each dose escalation
`during the ascending dose phase of the study. Following
`these evaluations, the remaining patients were recruited.
`
`458 NEUROLOGY 53 August (1 of 1) 1999
`
`The SRB reviewed data at 3-month intervals throughout
`the study. This review included the MRI safety data but
`did not include a review of the MRI efficacy data.
`Magnetic resonance imaging. MRI scans were per-
`formed according to a predefined MRI protocol at screen-
`ing, baseline, and every 4 weeks (before each dose)
`throughout the first 24 weeks of the study for a total of
`eight scanning time points. At each time point, three scans
`with a slice thickness of 5 mm were obtained: 1) proton
`density/T2-weighted scan, 2) T1-weighted unenhanced
`scan, and 3) T1-weighted gadolinium (Gd)-enhanced scan 5
`minutes after the administration of Gd-DTPA 0.1 mmol/
`kg. All scans were analyzed according to a prospectively
`defined MRI analysis plan by the UBC MS/MRI Analysis
`Group in Vancouver, Canada. After comparison of each
`MRI follow-up scan with the prior scan, the number of
`newly active lesions was ascertained by summing the new,
`recurrent, or enlarging T2 lesions, and the new or recur-
`rent Gd-enhancing lesions. Newly active lesions identified
`on both the enhanced T1 scan and the T2 scan were
`counted as single lesions. The primary efficacy measure
`was the cumulative number of newly active lesions identi-
`fied on the six treatment scans. Definitions of new, re-
`current, and enlarging lesions have been reported
`previously.17,18 Persistently enhancing or enlarging lesions
`were separately identified as persistently active lesions, a
`secondary measure of efficacy. In this way, new lesions
`could easily be separated from other types of activity.
`Other secondary efficacy measures included the percent-
`age of active scans, defined as the proportion of scans with
`one or more newly active lesions, and the burden of dis-
`ease, which was assessed as reported previously, at base-
`line, and at 24 weeks.17,18 Burden of disease was
`determined by outlining each MS lesion identified on the
`T2-weighted MRI scan. These areas were summed slice by
`slice for a total lesion area recorded as mm3. In addition,
`the total number of Gd-enhancing lesions (a measure of
`safety) was counted for each patient at each scanning time
`point on an ongoing basis to allow MRI data to be reviewed
`by the SRB.
`Clinical assessments. At the baseline visit and every 4
`weeks thereafter for the first 24 weeks, a history was
`taken, physical and neurologic examinations performed,
`and adverse events noted. Study drug was administered at
`the end of each visit. Patients were encouraged to come for
`additional visits should exacerbations occur between visits.
`During the second 24-week period, two formal visits were
`planned at weeks 36 and 48. Whenever possible, patients
`who withdrew from treatment were asked to continue all
`study procedures including all MRI scans.
`Clinical endpoints. Exacerbations were defined as the
`appearance of a new sign or symptom or the worsening of
`an old sign or symptom attributable to MS, lasting at least
`24 hours in the absence of fever, and preceded by a period
`of stability of 28 days. An exacerbation was deemed to
`have ended when signs and symptoms had begun to im-
`prove. For those patients with permanent deficits, the first
`day of a 28-day period of stability was taken as the ending
`of an attack. The Neurological Rating Scale (NRS)19 was
`completed each time the neurologic examination was per-
`formed. As in other clinical trials in MS, a decline in NRS
`score of 15 points or more was considered to reflect a se-
`vere change in the patient’s neurologic condition, whereas
`declines of 8–14 or 1–7 points were considered to reflect
`
`Page 2 of 51
`
`Page 2 of 51
`
`Page 2 of 9
`
`

`

`moderate or mild changes, respectively.17,20 A difference of
`0 points was categorized as no change; an increase in the
`score as an improvement. The EDSS as recommended by
`Kurtzke21 was scored at screening and at weeks 24 and 48
`by the study neurologist. In 120/167 (72%) of patients,
`the EDSS was performed at all time points by the same
`neurologist.
`In accordance with the protocol, a first analysis was
`undertaken after all patients had completed 24 weeks of
`double-blind treatment and after all MRI scans had been
`evaluated. An increase in the exacerbation rate was noted
`in lenercept-treated patients. This finding resulted in the
`sponsor’s decision to terminate the study and to release
`the treatment code. All study drug administration was
`stopped promptly and, after a final visit, data collection
`was discontinued. For this reason, study data through
`week 48 are incomplete. The follow-up period through
`week 48, however, was similar in all treatment groups.
`Pharmacokinetic/dynamic parameters. Serum samples
`were obtained at baseline and before dosing every 4 weeks
`for 24 weeks and at study weeks 36 and 48. The concentra-
`tion of lenercept and titers of antibodies to lenercept were
`determined. All samples were analyzed centrally. Lener-
`cept concentrations were measured using an enzyme-
`linked immunologic and biologic binding assay (ELIBA)
`developed by Roche (Hoffmann-LaRoche Ltd., Basel,
`Switzerland); antibodies to lenercept were identified by
`means of a double antigen antibody test. Samples with
`detectable anti-lenercept antibodies were further evalu-
`ated to determine the neutralizing potential of the antibod-
`ies (Medi-Lab, Medicinsk Laboratorium A/S, Copenhagen,
`Denmark).
`Autoantibodies. Serum samples were obtained at base-
`line and at study weeks 24 and 48 and assayed for IgM–
`rheumatoid factor (RF), antinuclear antibodies (ANA)
`(Hep 2), and antibodies to dsDNA (DAKO; Carpinteria,
`CA) in a central laboratory (A. Wiik, Statens Seruminsti-
`tut Copenhagen, Denmark).
`Statistical analyses. The cumulative number of newly
`active lesions was tested with a closed test procedure
`based on an analysis of variance (ANOVA) of the Ln(x 1 1)
`transformed sum of the lesions. The protocol required that
`the analysis of the primary efficacy criterion be performed
`after imputation of the median number of lesions at a
`specific time point so as to compensate for missing values
`at that time point. Of the 1,008 expected values, 34 were
`missing, resulting in data imputation as noted above. The
`results of this analysis showed no differences among the
`treatment groups ( p 5 0.417) or between the pairs of
`treatment groups. Data imputation was not performed for
`the analyses presented herein.
`A closed tests procedure, based on an ANOVA with the
`factor “treatment” of Ln(1 1 x), where x denotes the cumu-
`lative number of newly active lesions, was used to compare
`the cumulative number of newly active lesions between the
`treatment groups. The closed tests procedure was first
`used to compare the means among all four treatment
`groups (global test)22; then, to compare the means among
`all combinations of three of the four treatment groups; and
`finally, to compare the means of each lenercept treatment
`group with that of the placebo group. For all comparisons,
`F tests were performed at the same significance level (a 5
`0.05); however, adjusted p values are provided. The mean
`
`of a treatment group is regarded as significantly different
`from that of the placebo group when all comparisons that
`include the two relevant treatment groups result in a p
`value #0.05; i.e., the adjusted p value is the maximum of
`the p values of these comparisons. The procedure stops
`early if the global test is nonsignificant. This procedure
`guarantees a multiple a 5 0.05. The closed tests procedure
`described above was also used to assess the cumulative
`number of persistently active lesions. Center effects were
`assessed using descriptive methods.
`The Kruskal-Wallis test was used to compare the mean
`change in EDSS scores, change in the burden of disease,
`and percent of active scans. To assess the influence of
`baseline imbalances in MRI activity among the treatment
`groups, covariance analyses using the corresponding trans-
`formed baseline MRI values as covariate were performed.
`Survival analysis methods (Kaplan-Meier estimates;
`i.e., product-limit estimates and logrank tests)23 were ap-
`plied to analyze the time to first exacerbation and the
`duration of exacerbations because of right censoring at the
`end of the observation period. Logrank tests, with a Bon-
`ferroni adjustment of the significance level (a 50.017),
`were used for the multiple comparisons among the three
`lenercept treatment and the placebo treatment Kaplan-
`Meier curves. As an exacerbation duration can only be
`observed in the presence of an exacerbation, we investi-
`gated the conditional distribution of their durations. After
`inspection of the exacerbation data, we assumed a count-
`ing process model according to Anderson and Gill24 with
`independent increments because the process is slow. For
`the same reason, exacerbation durations were assumed to
`be independently and identically distributed between pa-
`tients and within patients for those patients who had more
`than one exacerbation.
`A chi-square was used to evaluate the number of pa-
`tients with no, one, two, three, or four exacerbations in
`each treatment group after 24 weeks of treatment and at
`the end of the study (through week 48). Because of small
`frequencies in some cells, the table was collapsed to count-
`ing patients with and without exacerbations to allow a
`valid chi-square test. For the multiple comparisons the
`unadjusted p values should be compared with the Bonfer-
`roni adjusted a of 0.017 (0.05 4 3). Chi-square tests were
`also used to evaluate the NRS and the rate of RF or ANA
`among the treatment groups. Cox regression analysis was
`performed to assess potential predictive factors for the oc-
`currence of exacerbations. The data and analyses were
`performed on all data available, i.e., through week 48,
`unless otherwise stated in the text or tables. Two-tailed
`analyses were used throughout.
`
`Results. Figure 1 depicts the trial profile. The treatment
`groups were comparable at entry on all baseline disease
`characteristics and demographics (table 1). The protocol
`permitted enrollment of both relapsing-remitting and sec-
`ondary progressive patients, and from 4 to 10 patients
`with secondary progression were enrolled in each group
`(see table 1). Prestudy MRI characteristics were likewise
`comparable among the treatment groups (table 2) although
`there was a tendency (nonsignificant) for the higher lener-
`cept dose groups to have more MRI activity (median).
`MRI results. The results of the cumulative number of
`newly active MRI lesions, the percentage of persistently
`active lesions, the percentage of active scans, and the
`
`August (1 of 1) 1999 NEUROLOGY 53 459
`
`Page 3 of 51
`
`Page 3 of 51
`
`Page 3 of 9
`
`

`

`Figure 1. Profile of the lenercept MS
`clinical trial.
`
`change in burden of disease over 24 weeks of treatment are
`shown in figure 2 and table 3. There were no significant
`differences between the treatment groups for any measure.
`The results of the analyses of the primary efficacy criterion
`according to the protocol specifications were similar to
`those presented here. Because of the tendency for higher
`activity at baseline in the high-dose groups (see table 2),
`covariance analyses using the corresponding transformed
`baseline MRI values as covariate were performed, but
`
`these, too, failed to show a significant difference between
`the groups.
`Exacerbations. The number of
`Clinical endpoints.
`patients who developed exacerbations by week 24 and
`through study week 48 were both increased in the lener-
`cept groups as compared with the placebo treatment group
`as shown in table 4. A center effect was not present. Over
`the entire study period, a total of 36 exacerbations was
`reported in patients taking placebo as compared with 38,
`
`Table 1 Demographic and baseline characteristics of patients entered into the lenercept MS trial
`
`Characteristics
`
`% Female
`Age, y, mean (range)
`% White
`No. with SPMS
`Mean (range) no. exacerbations
`in prior 2 years
`EDSS, mean (range)
`NRS, mean (range)
`
`Placebo,
`n 5 44
`
`66
`36.5 (21–50)
`98
`10
`2.7 (2–5)
`
`2.45 (0–5.5)
`83.2 (51–100)
`
`10, n 5 44
`
`80
`34.6 (23–51)
`100
`5
`2.8 (2–8)
`
`2.52 (0–5.0)
`83.7 (44–100)
`
`Lenercept, mg
`
`50, n 5 40
`
`78
`35.1 (19–47)
`100
`10
`2.8 (2–8)
`
`2.83 (1.0–5.5)
`81.8 (54–100)
`
`100, n 5 40
`
`73
`34.9 (21–51)
`93
`4
`3.0 (2–6)
`
`2.55 (0–5.5)
`83.0 (57–99)
`
`SPMS 5 secondary progressive MS; EDSS 5 Expanded Disability Status Scale; NRS 5 Neurological Rating Scale.
`
`Table 2 MRI measurements at baseline of patients entered into the lenercept MS trial
`
`Lenercept, mg
`
`Variable
`
`Placebo, n 5 44
`
`10, n 5 44
`
`50, n 5 40
`
`100, n 5 40
`
`Newly active lesions, mean
`Median (range)
`
`Persistently active lesions, mean
`Median (range)
`
`1.8
`0 (0–16)
`
`1.2
`0 (0–15)
`
`1.5
`0 (0–55)
`
`0.5
`0 (0–6)
`
`2.1
`1 (0–11)
`
`1.6
`0 (0–22)
`
`1.9
`1 (0–14)
`
`0.6
`0 (0–7)
`
`Burden of disease, mean
`Median (range)
`
`2,459.9
`1,626.0 (26.2–9,888.7)
`
`2,236.2
`1,147.4 (0–14,377.4)
`
`3,757.2
`2,142.4 (67.3–16,319.0)
`
`2,707.2
`1,365.3 (58.7–10,884.8)
`
`460 NEUROLOGY 53 August (1 of 1) 1999
`
`Page 4 of 51
`
`Page 4 of 51
`
`Page 4 of 9
`
`

`

`Figure 2. (A) Number of cumulative
`newly active lesions as determined by
`MRI (see Methods) over the first 24
`weeks of the lenercept MS trial:
`placebo –—, lenercept 10 mg - - - -, le-
`nercept 50 mg — - — -, lenercept 100
`mg – – – –. The vertical bars give the
`standard error at the four weekly inter-
`vals at which MRI scans were per-
`formed. (B) The mean number of
`gadolinium (Gd)–positive lesions every
`4 weeks over the first 24 weeks of the
`trial. Vertical bars give the standard
`error. (C) Mean anti-lenercept antibody
`titers at four weekly intervals.
`
`57, and 49 exacerbations in patients taking 10, 50, and 100
`mg of
`lenercept, respectively. Exacerbation duration
`showed a tendency to increase with lenercept treatment,
`but this did not reach statistical significance (see table 4).
`This assessment was limited to exacerbations with an on-
`set date within the first 24 weeks of the study as exacerba-
`tion resolution dates were available in all but four
`exacerbations (one per treatment group).
`Exacerbation rate. The overall exacerbation rate in
`patients treated with placebo was approximately one exac-
`erbation/patient/year (the expected placebo rate). The ex-
`
`acerbation rate was increased over the placebo rate by 2%,
`68%, and 50% in patients treated with lenercept at doses
`of 10, 50, and 100 mg, respectively (see table 4). To control
`for a possible effect of unequal follow-up of patients be-
`tween treatment groups, exacerbation rates were deter-
`mined for each treatment group by 4-week intervals. The
`mean 4-week exacerbation rates were then converted to
`annual rates as presented in table 4.
`There was a dose-dependent decrease in the time to
`first exacerbation as shown in figure 3 (logrank test:
`global, p 5 0.0006; 10 mg versus placebo, p 5 0.498; 50 mg
`
`Table 3 MRI measurements over the first 24 weeks of the lenercept MS trial
`
`Lenercept, mg
`
`Variable
`
`Placebo, n 5 43
`
`10, n 5 44
`
`50, n 5 40
`
`100, n 5 40
`
`Newly active lesions, mean
`Median (range)
`Persistently active lesions, mean
`Median (range)
`Percent of active scans,† mean
`Median (range)
`Percent change in burden of disease,† mean
`Median (range)
`
`* Analysis of variance of Ln(1 1 x).
`† Kruskal-Wallis test.
`
`8.9
`4.0 (0–92)
`6.2
`1.0 (0–100)
`45.6
`50 (0–100)
`6.0
`22.2 (228.8–280.9)
`
`8.2
`3.0 (0–55)
`3.3
`1.0 (0–49)
`40.6
`33 (0–100)
`9.9
`0.0 (230.5–251.0)
`
`15.5
`5.5 (0–124)
`9.6
`1.0 (0–128)
`51.6
`50 (0–100)
`4.3
`1.4 (235.2–62.8)
`
`12.0
`4.5 (0–102)
`3.3
`1.0 (0–24)
`49.2
`55 (0–100)
`4.9
`20.2 (236.0–81.4)
`
`p
`Value
`
`0.43*
`
`0.36*
`
`0.58†
`
`0.74†
`
`August (1 of 1) 1999 NEUROLOGY 53 461
`
`Page 5 of 51
`
`Page 5 of 51
`
`Page 5 of 9
`
`

`

`Table 4 Number, duration, and annual rate of exacerbations during the lenercept MS trial
`
`Exacerbations
`
`Patients with at least one exacerbation through week 24, n
`Patients with at least one exacerbation through week 48, n
`Exacerbations with onset # week 24
`Duration (days) of these exacerbations, mean
`Median (range)
`Annualized exacerbation rate
`
`Placebo,
`n 5 43
`
`15
`22
`22
`28.3
`28 (1–91)
`0.98
`
`Lenercept, mg
`
`10, n 5 44
`
`50, n 5 40
`
`100, n 5 40
`
`21
`26
`28
`38.6
`31 (6–189)
`1.00
`
`28
`32
`37
`41.6
`31 (6–201)
`1.64
`
`27
`32
`33
`42.0
`25 (4–261)
`1.47
`
`p
`Value
`
`0.003*
`0.007†
`
`0.62‡
`
`* Chi-square tests: global.
`† Kruskal-Wallis test.
`‡ Kaplan-Meier (KM) (means and medians are estimated from the KM curves).
`
`versus placebo, p 5 0.0006; and 100 mg versus placebo,
`p 5 0.006). The comparisons with unadjusted p values for 50
`and 100 mg of lenercept as presented above remain signifi-
`cant when Bonferroni adjusted (a of 0.017, i.e., 0.05 4 3).
`The severity of exacerbations was assessed indirectly by
`computing the difference between the best pretreatment
`NRS score (highest value at baseline or screening) and the
`worst score recorded at any time during the study, i.e.,
`through week 48. Although a trend for increasing severity
`with lenercept treatment can be perceived (table 5), this
`difference did not reach statistical significance.
`Exacerbations did not appear to occur more frequently
`immediately before or after study drug administration, but
`this may reflect no more than difficulties in determining
`exacerbation onset dates with precision.
`Predictors of exacerbations. A Cox regression was per-
`formed. None of the following was identified as predictive
`of exacerbations: age, sex, MS category, number of exacer-
`bations within 2 years before the study, baseline EDSS
`score, or baseline number of newly active lesions.
`
`Figure 3. Proportion of patients remaining exacerbation
`free over the course of the lenercept MS trial: placebo –—,
`lenercept 10 mg - - - - , lenercept 50 mg - – - –, lenercept
`100 mg – – – –. Times of termination for individual pa-
`tients are shown as vertical bars.
`
`462 NEUROLOGY 53 August (1 of 1) 1999
`
`EDSS. There was no difference in EDSS score changes
`between the treatment groups after 24 weeks of treatment
`or at the last assessment (table 5). Sixty-seven percent of
`patients in the placebo group reported new or worsening
`MS symptoms during the study as compared with 77%,
`90%, and 93% of patients treated with lenercept in the 10-,
`50-, and 100-mg dose groups, respectively. Symptoms that
`increased in frequency or severity with lenercept treat-
`ment included sensory complaints, limb weakness, visual
`impairment, fatigue, vertigo, and spasm (table 6).
`Safety. Six patients withdrew from treatment during
`the study as a result of an adverse event. Depression wors-
`ened in one patient in the placebo group; development of a
`rash after the first injection led to the withdrawal of two
`patients, one in the 10-mg dose group and the other in the
`100-mg dose group. Three additional patients withdrew
`from the 100-mg dose group, one due to the occurrence of a
`transient episode of flushing, dyspnea, and gastralgia after
`the fifth dose and two because of exacerbation-related
`symptoms after the fifth and seventh doses, respectively.
`Ninety-five percent of patients treated with placebo had at
`least one adverse event reported as compared with 87%,
`90%, and 95% of patients treated with 10, 50, and 100 mg
`of lenercept, respectively. Adverse events that increased in
`frequency with active treatment included headache, hot
`flushes, nausea, and abdominal pain (see table 6). A tran-
`sient episode of dyspnea associated with the administra-
`tion of lenercept at doses of 50 and 100 mg occurred in six
`patients.
`Laboratory. RF (IgM) and ANA were present in 1%
`(2/162) and 20% (33/161) of patients at baseline, respec-
`tively. During the study, five patients developed a positive
`IgM-RF (all in the lenercept treatment groups) and 15
`patients developed ANA, 14 of whom were receiving lener-
`cept. One patient on lenercept developed both antibodies.
`Thus, new occurrences of RF or ANA were more frequent
`in lenercept-treated than in placebo-treated patients (18/
`124 versus 1/43; chi-square p 5 0.03). Clinical mani-
`festations of rheumatoid arthritis or systemic lupus
`erythematous were not observed, and dsDNA was consis-
`tently negative in all patients in whom ANA positivity was
`detected.
`Total serum IgM had increased in a dose-dependent
`fashion in lenercept-treated patients after 24 weeks of
`treatment by an average of 0.6, 0.9, and 1.0 g/L in the 10-,
`
`Page 6 of 51
`
`Page 6 of 51
`
`Page 6 of 9
`
`

`

`Table 5 Change in the NRS and the EDSS during the lenercept MS trial
`
`Lenercept, mg
`
`Rating scale
`
`Placebo, n 5 43
`
`10, n 5 44
`
`50, n 5 40
`
`100, n 5 40
`
`p
`Value
`
`0.37*
`
`0.24†
`
`NRS, weeks 1–48
`No change or improved
`Mild deterioration
`Moderate deterioration
`Severe deterioration
`EDSS, weeks 1–24, n
`Mean change (range)
`EDSS, weeks 1–48
`Mean change (range)
`
`* Chi-square.
`† Kruskal-Wallis test.
`
`14
`21
`3
`5
`43
`20.131 (22.0 to 2.0)
`
`13
`18
`10
`3
`44
`20.295 (23.0 to 5.5)
`
`9
`14
`8
`9
`40
`20.05 (22.0 to 2.0)
`
`9
`18
`7
`6
`40
`20.075 (112.0 to 4.0)
`
`0.014 (21.0 to 2.0)
`
`20.024 (23.0 to 5.5)
`
`20.026 (22.0 to 2.0)
`
`0.171 (22.0 to 4.0)
`
`0.71†
`
`NRS 5 Neurological Rating Scale; EDSS 5 Expanded Disability Status Scale.
`
`50-, and 100-mg dose groups. For those patients in the 100
`mg dose group who remained on treatment beyond 24
`weeks, the mean increase reached 1.2 g/L. Total serum IgG
`concentrations had increased by an average of 0.5 g/L at 24
`weeks on treatment in patients receiving 100 mg of lener-
`cept; for those patients who remained on treatment in the
`100-mg dose group, this increase ultimately reached 1.2
`g/L. Serum IgG concentrations were not increased in the
`other groups.
`Pharmacokinetics. Trough serum concentrations of le-
`nercept were detectable in only a third of the patients in
`all dose groups and tended to relate to the anti-lenercept
`antibody profile. Thus, patients without antibodies had
`persistently detectable lenercept trough serum concentra-
`
`Table 6 Patient complaints over the course of the lenercept MS
`trial
`
`Lenercept, mg
`
`Placebo,
`n 5 43
`
`10,
`n 5 44
`
`50,
`n 5 40
`
`100,
`n 5 40
`
`67
`44
`19
`19
`14
`2
`2
`
`23
`7
`5
`5
`0
`
`77
`45
`20
`14
`14
`7
`5
`
`36
`7
`11
`18
`0
`
`90
`58
`33
`28
`25
`13
`7
`
`35
`8
`23
`23
`2
`
`93
`63
`43
`40
`28
`15
`13
`
`45
`18
`15
`8
`13
`
`Patient complaints
`
`MS-related
`Any MS symptom
`Sensory complaints
`Limb weakness
`Visual impairment
`Fatigue
`Vertigo
`Spasm
`Other
`Headache
`Hot flushes
`Nausea
`Abdominal pain
`Dyspnea
`
`Values are %.
`
`tions, whereas patients with high antibody concentrations
`had consistently undetectable trough levels. Antibodies to
`lenercept at a concentration greater than 20 ng/mL, as
`determined by the double-antigen screening assay, were
`detected in the majority of patients on treatment (5%
`[2/43] of patients on placebo, as compared with 98% [43/
`44], 88% [35/40], and 100% [40/40] of patients in the 10-,
`50-, and 100-mg dose groups, respectively.) Antibody con-
`centrations varied considerably from patient to patient and
`over time; generally, a dose-dependent peak occurred after
`the first dose, followed by a relatively stable level that
`persisted for as long as the patient continued to receive
`lenercept (see figure 2).
`
`Discussion. Lenercept treatment in the phase II
`study reported here increased MS attack frequency.
`There was also a suggestion that lenercept increased
`attack duration and worsened attack severity as
`judged by the extent of changes noted on the NRS.
`The lenercept effect was more pronounced at the two
`higher doses employed (50 and 100 mg) than at the
`lowest dose (10 mg). An increase in attack frequency
`was noted in lenercept-treated patients within the
`first month on drug and persisted throughout the
`period during which drug was given. The magnitude
`of the treatment effect on exacerbations could be bi-
`ased because the trial was stopped early. We believe
`this is unlikely as the results of the second 24-week
`study period were consistent with those found during
`the first 24 weeks.
`Despite the seemingly deleterious clinical effects
`of lenercept, no meaningful difference between treat-
`ment groups was noted in terms of worsening of the
`EDSS score. Whereas MS attack frequency and sub-
`sequent development of disability have correlated in
`large patient series studied over many years, no such
`correlation was noted in the current study, perhaps
`because of the relative insensitivity of the EDSS and
`the prompt cessation of the study following the in-
`terim monitoring analysis.
`August (1 of 1) 1999 NEUROLOGY 53 463
`
`Page 7 of 51
`
`Page 7 of 51
`
`Page 7 of 9
`
`

`

`It is important to remember that had the current
`study shown a positive treatment effect, the results
`would have had to be validated in a second trial.
`Prudence dictates that similar caution be applied to
`the detrimental clinical effect noted in the current
`trial, particularly in view of the seemingly discor-
`dant MRI data. Setting caution aside, it is reason-
`able to conclude that lenercept is probably con-
`traindicated in MS based on the internal consistency
`of the clinical results presented here and the unto-
`ward results obtained in a preliminary study of two
`patients administered a humanized mouse monoclo-
`nal anti-TNF antibody.25 The dissociation seen be-
`tween a significantly increased clinical activity and,
`at best, a slight trend toward an increase in M