Abstract
Background: To assess the long-term effect of the lymphocyte-depleting humanized monoclonal antibody Campath 1H on MR markers of disease activity and progression in secondary progressive MS patients.
Methods: Twenty-five patients participated in a crossover treatment trial with monthly run-in MR scans for 3 months, followed (after a single pulse of Campath 1H) by monthly MR scans from months 1 to 6 and again from months 12 to 18. MR analysis was performed to provide measurements of the number and volume of gadolinium (Gd)-enhancing lesions as well as the hypointense lesion volume on a T1-weighted sequence. In addition, serial measurements of T2 brain lesion volume, brain volume, and spinal cord cross-sectional area were made over the duration of the study. The relationship between clinical and MR measures of disease evolution was also assessed.
Results: Treatment was associated with a reduction in the number and volume of Gd-enhancing lesions (p < 0.01). Despite this, a decrease in brain volume was seen in 13 patients during the 18 months post-treatment. The mean pretreatment Gd-enhancing lesion volume was predictive of subsequent reduction in brain volume (r = 0.77, p = 0.002). Reduction in brain volume also correlated with the change in T1 hypointense lesion volume after treatment (r = 0.53, p < 0.01). A reduction in spinal cord area was also seen throughout the study duration, and this correlated with an increase in disability (r = 0.65, p = 0.01).
Conclusion: Campath 1H treatment was associated with a sustained and marked reduction in the volume of Gd enhancement, indicating suppression of active inflammation. Nevertheless, many patients developed increasing brain and spinal cord atrophy, T1 hypointensity, and disability. This study highlights the potential role for novel MR techniques in monitoring the effect of treatment on the pathologic process in MS.
There is widespread agreement that advances in MRI techniques have improved clinical trial methodology in MS substantially.1,2 MRI provides an objective and sensitive tool for assessing disease progression of MS and its modification by treatment.1-3 The most established MRI outcomes in MS treatment trials are gadolinium (Gd)-enhanced T1-weighted imaging and total T2 brain lesion volume, providing measures of disease activity and progression respectively. Gd enhancement indicates blood–brain barrier breakdown—which, as documented in MS and in experimental models, is usually accompanied by active inflammation.4,5 However, a high signal on a T2-weighted sequence has a much lower level of pathologic specificity, reflecting varying degrees of edema, inflammation, demyelination, and axonal loss. A number of novel MRI tools have been developed recently, offering the prospect of greater pathologic specificity for the more destructive pathologic elements of demyelination and axonal loss. Examples include measures of both brain and spinal cord atrophy, and hypointense T1 lesion volume quantification. Hypointensity on an unenhanced T1 sequence is seen in approximately 20 to 30% of chronic MS lesions and, when irreversible, it probably indicates an important degree of axonal loss.6-8 Furthermore, recent studies6-12 have shown stronger correlations between disability and these novel MRI markers than with standard measures of T2-weighted lesion volume. These new techniques might, therefore, appear especially promising as clinically relevant tools for monitoring treatment.9,10,12,13
In parallel with these neuroimaging achievements, modern biotechnology is offering a new range of potential treatments for MS. A series of monoclonal antibodies targeting a wide range of T-cell surface molecules has been used in MS patients.14 Campath 1H is used in a variety of transplant situations and in the treatment of autoimmune disorders.15-18 It is a humanized monoclonal antibody that recognizes the CD52 antigen present on the surface of lymphocytes and monocytes. It is profoundly cytolytic and thus leads to prolonged lymphocyte depletion. A preliminary study15 involving seven MS patients showed that a single pulse of Campath 1H suppressed MR disease activity for at least 6 months.
The clinical and immunologic results of a crossover design trial evaluating Campath 1H in a cohort of MS patients are reported in full elsewhere, along with a summary of the enhancing lesion activity, brain atrophy, and proton MR spectroscopy data.17,18 We now report the longer term effect of Campath 1H in suppressing MS disease activity as expressed by a quantitative volumetric measurement of Gd enhancement in 25 of these patients. We also report the effect of treatment on quantitative measurement of T1 hypointense lesion volume, brain volume, and spinal cord cross-sectional area as potential MRI markers of structural loss (i.e., demyelination and axonal loss).
Methods.
Patients.
The inclusion criteria included 1) clinically definite MS19 that followed a secondary progressive course, 2) deterioration of 1 point on Kurtzke’s Expanded Disability Status Scale (EDSS)20 within 12 months, and 3) an EDSS score between 4.0 and 6.0 at selection. Thus patients had moderately advanced and active disease. Exclusion criteria included 1) substantial cognitive impairment that compromised the patient’s ability to give informed consent and 2) previous treatment with any previous immunosuppressive treatment other than corticosteroids. All patients gave informed written consent to enter the study, which was approved by the local research ethics committee.
After a 3-month period of monthly Gd-enhanced imaging, during which it was required that at least one enhancing lesion was seen, 25 patients received a single-pulse course of anti-CD52 (Campath 1H). Four other patients, who also exhibited at least one enhancing lesion during the run-in scans, preferred to act as open control subjects because of concern over possible adverse events. These patients were observed clinically and radiologically alongside the treated cohort.
Treatment dosage and administration.
The humanized anti-CD52 monoclonal antibody was prepared for clinical use in the Therapeutic Antibody Centre in Oxford. The therapeutic grade antibody was made in Chinese hamster ovary cells and purified on protein A. The treatment regime consisted of a 20-mg IV daily infusion of Campath 1H given over 3 to 4 hours for 5 consecutive days. Patients were hospitalized and followed daily during Campath 1H treatment. Neurologic symptoms and adverse events were documented and the EDSS score was recorded.
Patient clinical evaluation.
A neurologic assessment was performed every 3 months by one observer (A.C.), and the EDSS score was recorded. A sustained change in EDSS was defined as an increase of 1 point in EDSS if less than 6 at entry, but 0.5 points with an EDSS score ≥ 6.0 that was confirmed over at least 6 months.21
MR acquisition.
MR scans were performed on a 1.5-T system (General Electric Signa, Milwaukee, WI) using a standard quadrature head coil. Patients were imaged monthly from 3 months before treatment (month −3) to 6 months after treatment, and then again monthly from months 12 to 18.
The following sequences were performed monthly at these time points: axial proton density- (PD) and T2-weighted fast spin-echo sequence (repetition time [TR], 3,500 msec; echo time [TE], 18/90 msec; with an echo train length of 8) and axial T1-weighted spin-echo (TR, 600 msec; TE, 20 msec), obtained 10 minutes after IV injection of 0.1 mmol/kg Gd diethylene-triamine penta-acetic acid. Contiguous axial slices were obtained using a 5-mm thickness, 24-cm field of view, and 256 × 256 image matrix. Repositioning was performed based on standardized anatomic landmarks.22
In addition to the monthly protocol, all patients had axial spinal cord MR scans at months −3, 0 (baseline), and then at months 6, 12, and 18 post-treatment. An axial, single, 5-mm-thick slice was obtained using a gradient echo sequence (TR, 300 msec; TE, 75 msec; flip angle, 15 deg) through the cord at the mid C5 vertebral level. An orientation perpendicular to the cord was defined from a sagittal localizer.
MR analysis.
All MR measurements were performed blinded to the patients’ clinical status. Gd-enhancing lesions were identified and marked on hard copy by two expert clinicians (D.H.M., M.G.-C.). A third rater (A.P.) subsequently measured the volume of marked lesions on electronic data. The primary MR outcomes were the number and volume of all enhancing lesions before and after treatment. For post hoc statistical analysis, the study duration was divided into five periods of 3 months duration (months −2 to 0, 1 to 3, 4 to 6, 13 to 15, and 16 to 18).
The total lesion volume on the PD-weighted image was measured at month −3, month 0, and then at 6, 12, and 18 months after treatment. The number and volume of hypointense lesions on T1-weighted images were also measured at these time points. Each of the hypointense lesions selected had a corresponding hyperintensity on T2-weighted imaging.
The mean T1-to-T2 ratio (defined as the ratio between T1 hypointense lesion volume and T2 hyperintense lesion volume) at months 0 and 18 after treatment was calculated for the whole brain and for the infratentorial and supratentorial regions separately. All lesion volume measurements were performed with a semiautomated local thresholding technique23 by one observer (A.P.) who was blind to both the order of the scans and the clinical details.
Measurement of brain volume.
Brain extraction was performed on the Gd-enhanced, T1-weighted images at months −3, 0, 6, 9, 12, and 18 using an algorithm integrated in a window-based image analysis package (eXKalp @D.S. Yoo, Department of Medical Physics and Bio-Engineering, University College, London, UK).12 Prior to brain extraction, the scans were examined by an experienced observer (A.P.) who was blind to the clinical details to ensure that repositioning was adequate by using current, accepted treatment trial criteria.1 The observer selected four contiguous slices from each scan with the most caudal at the level of the velum, and atrophy was expressed as change in milliliters per year. Substantial change in brain volume was regarded as having occurred if the measured change exceeded the 95% CIs of measurement variation for the technique.12
Measurement of spinal cord area.
The images obtained at different times were coded and ordered randomly. The cross-sectional area of the cord was measured by tracing the circumference of the cord image manually. Measurements were recorded on two separate occasions and the mean of these two measurements was considered to be the cord area.
Efficacy end points.
The primary MR end point was the cumulative number and volume of Gd-enhancing lesions. Secondary MR outcomes were as follows: 1) change of T2 total lesion volume, 2) change in hypointense T1 lesion volume, 3) change in cerebral volume, and 4) change in spinal cord area during the study. We also evaluated the relationships between MR parameters and EDSS score at baseline and during follow-up.
Statistical analysis.
Statistical analysis was performed using the SPSS/PC+ (version 7.5.1; SPSS, Chicago, IL) package. For the MR data, either Wilcoxon’s matched-pairs signed rank test or Friedman’s two-way analysis of variance was used as appropriate. The crossover analysis, comparing the total enhancing lesion volumes before and after treatment, was performed with a two-tailed signed rank sum test on the logarithm of the ratio between treatment and baseline block.
The correlations between change in disability and both the EDSS score at entry and the MR findings were calculated using Spearman’s rank correlation coefficient.
Results.
Clinical data.
Initially, 33 patients were included in the study. Of these, one was lost to follow-up and three had been followed for less than 6 months at the time of MR analysis. Therefore, the final study cohort comprised 29 secondary progressive MS patients (25 treated with Campath 1H and 4 untreated open control subjects, 14 men and 15 women). At study entry (month −3), the mean age was 38.6 years for treated patients and 40.6 years for untreated control subjects. Mean disease duration was 12.6 years for the treated patients and 6.8 years for the untreated control subjects. Mean EDSS scores were 5.4 for the treated patients and 4.8 for the untreated control subjects. No significant differences between the two groups were observed in terms of baseline clinical characteristics. By study termination (month 18), the mean EDSS increase was 0.62 in the treated patients and 0.63 in the untreated control subjects. Sixteen of 29 patients had a sustained increase in EDSS score (3 untreated, 13 treated). Four patients received corticosteroids in the 18 months after treatment, and two in the pretreatment period. When corticosteroids were given, MRI was delayed such that no image was acquired within a month of administration.
Immunologic markers.
Campath 1H depleted peripheral lymphocyte counts to a similar extent in all patients. Immediately after treatment, very few lymphocytes could be detected, but by month 3 the B-cell numbers had returned to normal. T-lymphocyte numbers were slower to recover, remaining at 30 to 40% of pretreatment levels by month 18. The extent of lymphopenia did not correlate with suppression of disease activity on MRI. However, the suppression of in vitro mitogen-induced interferon-γ secretion from patients’ peripheral blood mononuclear cells did correlate with the degree of reduction of new Gd-enhancing lesions.18
MRI data.
Gd enhancement.
Table 1 shows the mean number and volume of Gd-enhancing lesions per scan for the different 3 month periods. As shown, compared with the 3 months of pretreatment, both the mean number and the volume of Gd-enhancing lesions was markedly reduced during all 3-month periods over the 18 months of follow-up in the treated group (p < 0.01). This effect was observed in every patient. The reduction in MRI activity was significant for each post-treatment block (p < 0.01). In contrast, the mean volume of the Gd-enhancing lesions was similar over all time periods in the control group (see table 1).
Mean number and volume of gadolinium-enhancing lesions per scan over the study duration
T2 lesion volume.
In the treated group there was an increase in T2 lesion volume in the 3-month pretreatment period (p < 0.001). An additional increase between baseline (month 0) and month 6 was observed, but in the following two periods (months 6 to 12 and months 12 to 18) no significant change in lesion volume was seen (table 2). A similar pattern was seen in the untreated control group (see table 2).
Changes in T2 lesion load during the study
Hypointense T1 lesion volume.
In the treated group an increase in T1 hypointense lesion volume was observed between month −3 and month 0 (baseline), months 0 to 6, and also months 6 to 12 (p < 0.01), but no difference was seen between months 12 and 18 (table 3). A similar trend was seen in the control group. The mean supratentorial T1-to-T2 ratio at baseline (month 0) was 0.31 (range, 0.21 to 0.47), and by month 18 there had been an increase to 0.41 (range, 0.28 to 0.56; p < 0.01). The T1-to-T2 ratio was higher for supratentorial regions than for infratentorial regions at both baseline (0.43 versus 0.31; p < 0.001) and month 18 (0.54 versus 0.41; p < 0.001).
Change in hypointense T1 lesion load during the study
Cerebral atrophy.
The mean cerebral volume of the selected four slices in the 25 treated patients analyzed at baseline was 298 mL (range, 237 to 344 mL). The mean rate of brain volume change by the end of the study for all 29 patients (treated and untreated) was −4.1 mL/year, and volumes at month 18 were lower than at month 0 (p < 0.001). A significant reduction in brain volume beyond the confidence limits of measurement variation was identified in 13 of the 25 treated patients during the study. There was a correlation between the mean pretreatment Gd-enhancing lesion volume and subsequent reduction in brain volume after treatment (r = 0.77, p < 0.005). However, neither mean pretreatment T2 lesion volume nor T1 hypointense lesion volume predicted subsequent reduction in cerebral volume (table 4).
Correlations between change in brain volume and other MRI data for treated group (n = 25)
There was no correlation between progression in brain atrophy and mean change in T2 lesion volume (r = 0.12, p = not significant), but there was a correlation with mean hypointense T1 lesion volume increase from month 0 to 18 (r = 0.53, p < 0.01; see table 4).
The mean cerebral volume of the control subjects at month 0 was 296 mL (range, 287 to 302 mL), and the mean rate of brain volume change by the end of the study was 4.2 mL/year. Three of four control subjects developed a significant reduction in brain volume from months 0 to 18.
Spinal cord atrophy.
Fifteen of 29 patients were excluded from the analysis because of poor repositioning or movement artifact that rendered the scans uninterpretable. Therefore our analysis is confined to 14 patients (all treated) measured at months 0 and 18. The mean cord area of the patients at month 0 was 94 mm2 (range, 69 to 115 mm2). Compared with baseline (month 0), by month 18 there was a reduction in cord area with a mean change of −6.4 mm2 (range, −16.5 to +3.4 mm2; p < 0.001).
There was no significant correlation between change in spinal cord atrophy and 1) increasing brain atrophy, 2) number or volume of Gd-enhancing lesions during either the pre- or post-treatment periods, 3) T2 lesion volume, and 4) hypointense T1 lesion volume. There was, however, a correlation between reduction in spinal cord area and the increase of infratentorial T1-to-T2 ratio between months 0 and 18 (r = 0.64, p = 0.01).
Relationship between MRI findings and clinical change.
The cross-sectional and longitudinal correlations between MRI variables and EDSS score were generally not significant (table 5). Only the change in the spinal cord area and in the infratentorial T1-to-T2 ratio correlated significantly with the increase in EDSS score over the study duration.
Clinical MRI correlations for treated group (n = 25)
Further analysis was performed by stratifying the treated patients according to clinical progression. The 13 patients with a sustained change in EDSS score developed more brain atrophy than the group with no EDSS score change (−6.7 mL/year versus 0.7 mL/year, p < 0.01). Those patients with no change in EDSS score did not show significant changes in brain volume from baseline (figure 1).
Figure 1. Serial cerebral volume measurements stratified by change in Expanded Disability Status Scale (EDSS) score. Rx = treatment; squares = no change in EDSS (n = 12); diamonds = change in EDSS (n = 13).
The 13 patients with a sustained increase in EDSS score did not differ significantly from the 12 without EDSS score change according to change in T2 lesion volume. However, the group with a sustained increase in EDSS score did have a higher baseline hypointense T1 lesion volume than the group with no EDSS change (p < 0.05; figure 2). Furthermore, the group with a sustained increase in EDSS score demonstrated an increase in hypointense T1 lesion volume over the study duration (p < 0.05), whereas no significant change was seen in the group with a stable EDSS score (see figure 2). Patients with a significant sustained increase in EDSS score also had a higher pretreatment Gd-enhancing lesion volume than those with a stable EDSS score (p < 0.001; figure 3). In the control group, the three patients who developed brain atrophy showed a sustained change in EDSS score, whereas the one patient with no change in brain volume did not.
Figure 2. Serial hypointense T1 lesion load measurements stratified by change in Expanded Disability Status Scale (EDSS) score. Rx = treatment; squares = no change in EDSS; diamonds = change in EDSS.
Figure 3. Box plot shows mean pretreatment gadolinium-enhancing lesion load stratified by change in Expanded Disability Status Scale (EDSS) score. p < 0.001.
Discussion.
This study shows that a single pulse of treatment with the humanized monoclonal antibody Campath 1H is associated with a major and sustained reduction in inflammatory MS disease activity as demonstrated by the suppression of Gd-enhancing lesion formation. No such reduction in activity was seen in the four untreated control subjects. Such a small control group cannot provide definitive comparative data, and indeed may by chance have had a lower level of enhancement during the pretreatment period. However, the baseline-treatment crossover design that we used has successfully shown a treatment effect on enhancing MRI activity using another agent, interferon β-1b,24 which was subsequently confirmed using the more robust parallel-group, placebo-controlled design.25 That both the treated and control group patients were blinded to the level of MRI activity throughout the study reinforces our impression that the reduction in Gd enhancement reflects a therapeutic effect.
In this analysis we have combined conventional MRI parameters (Gd enhancement and T2 lesion volume) with new approaches (hypointense T1 lesion volume, brain volume, and spinal cord area quantification) in an attempt to identify the effect of treatment on individual pathologic elements. These results should be interpreted with some caution due to the small numbers of control patients and the potential for measurement drift over time. However, all quantitative analyses were performed by the same experienced observer, in randomized order, without knowledge of patient identity or scan order, in an attempt to minimize the potential for bias.
We found no significant correlation between T2 lesion volume changes and clinical findings. This emphasizes the low pathophysiologic specificity of focal T2 signal hyperintensity, and is concordant with the weak correlations found between T2 abnormalities and disability in other studies.26-28 Although significant correlations between hypointense T1 lesion volume have been reported elsewhere,6 we found no overall correlation between hypointense T1 lesion volume and EDSS score changes. However, those with a sustained increase in EDSS score exhibited a significant increase in hypointense T1 lesion volume compared with those with stable disability. Furthermore, a moderate longitudinal correlation between the infratentorial T1-to-T2 ratio and EDSS score was identified. There is emerging evidence that those areas of high signal on a T2 sequence that are hypointense on a corresponding T1 image represent more severe structural loss.7 Therefore progressive loss of structure in functionally eloquent areas such as the brainstem and cerebellum might be expected to result in an increase in disability, as suggested by our results.
We also found a clear relationship between sustained increase in EDSS score and the development of cerebral (p < 0.009) and spinal cord atrophy (p < 0.01), although the significance of the latter is qualified by the smaller cohort of patients who could be analyzed for methodological reasons. Furthermore, those patients with a sustained increase in disability and brain atrophy had significantly smaller cerebral volumes at baseline. This finding suggests that an ongoing atrophic process is already more established in this cohort, and that loss of brain volume is predictive of subsequent clinical progression.
Those patients with a significant increase in EDSS score showed a much higher cumulative Gd-enhanced volume during the pretreatment period than the clinically stable patients. This indicates that although Campath 1H was able to suppress new lesion formation, it did not prevent the secondary consequence of inflammatory lesions that had already developed during the immediate pretreatment period. Increasing disability has been associated with higher Gd-enhancing lesion frequencies,28,29 and it is possible that inflammation might compromise repair mechanisms within demyelinated regions and thereby expose axons to the immunologic and biological consequences of persistent demyelination. Thus, even if inflammation is stopped, previously acquired extensive areas of damage may still undergo secondary degeneration with disease progression. Severe atrophy is likely to represent the final irreversible sign of axonal loss, leading to progression of disability.10 The recent interferon-β trials also suggest that suppression of active inflammation delays but does not prevent progression of disability,25,30 perhaps indicating ongoing axonal loss in established areas of pathology. A key issue in current MS research is to identify whether earlier intervention with therapies aimed at suppressing inflammation will delay or prevent later irreversible atrophy.31
Measurements of cerebral atrophy, spinal cord area, and T1 hypointense lesion volume represent important new approaches to the assessment of therapeutic efficacy. These techniques are objective, have high serial reproducibility, and are both simple and quick to apply. They can supplement the more traditional MRI measures and provide a means of assessing the impact of therapeutic intervention on destructive pathology and tissue loss.
Acknowledgments
Acknowledgment
The authors thank Ms. E. Hughes for performing all the imaging and Dr. D. Yoo for providing the brain subtraction program.
Footnotes
The NMR Research Unit is supported by a generous grant from the Multiple Sclerosis Society of Great Britain and Northern Ireland. A.J.C. is an MRC Clinical Training Fellow and some aspects of the work were supported by a grant from Muster.
- Received November 16, 1998.
- Accepted in final form March 27, 1999.
References
Letters: Rapid online correspondence
REQUIREMENTS
You must ensure that your Disclosures have been updated within the previous six months. Please go to our Submission Site to add or update your Disclosure information.
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
