Autologous hematopoietic stem cell transplantation suppresses Gd-enhanced MRI activity in MS

Materials and methods.

A phase I/II trial of ASCT in MS was initiated in July 1998 by the Italian Cooperative Group for Bone Marrow and Blood Transplantation (GITMO) and a group of neurologic teams. Five neurologic and hematologic centers participated in the study, whose main objective was to investigate MR changes after ASCT and to obtain information on clinical course, feasibility, and safety. The inclusion criteria were clinically and laboratory-confirmed secondary progressive MS with or without relapses, Expanded Disability Status Scale (EDSS) score between 5 and 6.5, documented rapid progression over the previous year unresponsive to conventional therapies (worsening of 1 point between EDSS 5 and 6, of 0.5 point between 6 and 6.5), absence of cognitive disturbances, and the presence of at least one Gd diethylenetriamine penta-acetic acid (Gd)-enhanced area on brain MRI using a triple dose (TD) of Gd (0.3 mmol/kg Gadodiamide; Omniscan; Nycomed, Oslo, Norway). Ethical committees approved the study. Patients’ cognitive functioning was measured utilizing the Mini-Mental State Examination and a battery of standard neuropsychological tests.10 All subjects scored within the standard normal ranges on these measures before the informed consent was signed.

Eighteen patients were screened and eight patients were excluded owing to the absence of enhancing lesions on brain MRI. Ten patients have now undergone ASCT, with a median follow-up of 15 months (range 4 to 30 months). Six subjects were women and four men, with a median age of 35.5 years (range 26 to 52 years), median disease duration of 12 years (range 6 to 19 years), and a secondary progressive course for a median of 4.5 years (range 1 to 8 years). Relapses were superimposed in 4 of 10 cases. Six subjects had been previously treated with interferon-β_1b and four subjects with immunosuppressive therapies only (azathioprine, CY, or mitoxantrone) or steroids, all without clinical response. At screening, the median EDSS was 6.5 (range 5.5 to 6.5) and the Scripps Neurologic Rating Scale was 59 (range 41 to 66). At baseline, just before SC mobilization, the neurologic condition had deteriorated in three patients (median EDSS 6.5, range 5.5 to 8). Only Patient 3 was treated with steroids during the run-in period (1,000 mg of methylprednisolone IV × 3 days, 500 mg × 3 days, and 250 mg × 3 days) for the occurrence at month −2 of a severe worsening of neurologic status. Patient characteristics are reported in table 1. The effect of ASCT was evaluated with serial monthly TD Gd-enhanced MRI for a pretreatment period of 3 months and compared with serial monthly TD Gd-enhanced MRI for the following 6 months. Subsequently, MRI scans were obtained every 3 months until month 30. TD was utilized rather than the standard dose of Gd because TD-enhanced MRI is more sensitive in detecting disease activity than standard dose-enhanced MRI.11 The MRI protocol was utilized in all centers according to established standards. After a three-step centering and fast para-axial sequence for repositioning control, the following para-axial sequences were acquired with 44 slices (thickness 3 mm, no gap interslice, in-plane resolution 0.98 × 0.98 mm): spin echo proton density and T2 weighted, spin echo T1 weighted, flow-attenuated inversion recovery, and postcontrast spin echo T1 weighted. The TD of Gd was injected by means of a long-line venous access before acquiring the flow-attenuated inversion recovery sequence, so that the postcontrast T1-weighted sequence was acquired about 5 to 8 minutes after the Gd injection. All images were hard copied by a laser imager, and electronic data were stored.

Images were sent to the Neuroimaging Research Unit (Scientific Institute Ospedale S. Raffaele, Milan), where they were examined and quantified by an experienced observer, blind to the period in which they were obtained, who counted the number of Gd-enhancing lesions12 and the number of T2-weighted lesions.13 The images were then sequentially ordered, obtaining the numbers of Gd-enhancing and of new T2-positive lesions before and after treatment.

Neurologic examinations (EDSS, Scripps) were scheduled at screening, baseline, before and after the mobilization regimen, after the conditioning regimen, every month for the following 6 months, and then every 3 months. Peripheral blood progenitor cells were mobilized with CY 4 g/m² followed by G-CSF 5 μg/kg/day until the completion of the cell harvests. The immunoablative regimen was carried out within 30 to 40 days after mobilization. According to the protocol previously utilized by Fassas et al.,5 the regimen consisted of BCNU 300 mg/m² at day −7, cytosine-arabinoside 200 mg/m² and etoposide 200 mg/m2 from day −6 to day −3, and melphalan 140 mg/m2 at day −2. Rabbit antithymocyte globulin (ATG; 5 mg/kg/day) was administered at +1 and +2 as in vivo T depletion. Intravenous cyclosporin A (1 mg/kg) was given during the conditioning regimen to prevent disease exacerbation due to cytokine release.

Results.

No major adverse events were recorded during or after treatment. However, fever occurred in the neutropenic postconditioning period in 9 of 10 patients, which was successfully treated with intravenous antibiotic treatment. Deterioration of the neurologic condition was not observed during the mobilization or the conditioning regimens, although in the period immediately after the conditioning procedure, a transient worsening of the previous neurologic condition was common, likely due to a concomitant fever.

Mobilization was successful in all cases, and a median number of 9.06 × 10⁶/kg of CD34+ cells were collected (range 3.51 to 26.02 × 10⁶/kg). Mobilization was generally well tolerated, with occasional nausea and headache. The most serious adverse effects were subclavian phlebitis in one patient and transient inappropriate secretion of antidiuretic hormone (ADH) in another patient (table 2). Nadir of polymorphonuclear cells (PMN) occurred 8 days after mobilization (range 7 to 10 days) and of platelets (Plt) on day 10 (range 3 to 13 days). Median days with PMN levels of <0.5 × 10⁹/L and Plt levels of <50 × 10⁹/L were 4 (range 2 to 4 days) and 0 (range 0 to 4 days). A median of 18 days (range 13 to 28 days) of hospitalization for the mobilization procedure was required. The conditioning treatment was, not surprisingly, more troublesome, considering the large amount of immunosuppressive drugs administered and the duration of treatment (from day −7 to day −2). A median number of 6.85 × 10⁶/kg of CD34+ cells were infused (range 3.01 to 16.46 × 10⁶/kg). After transplant, median days with PMN levels of <0.5 × 10⁹/L and Plt levels of <50 × 109/L were 7 (range 6 to 12 days) and 9 (range 4 to 18 days). Median days spent in the hospital for the conditioning and transplantation treatments were 25 (range 21 to 44 days). In the postconditioning period, asthenia was a constant symptom; transient increase of aspartate aminotransferase and alanine aminotransferase was observed in 2 cases, urinary tract infection in 2 cases, and neutropenia-related fever in 9 of 10 patients (see table 2). Symptomatic therapy or intravenous antibiotics were utilized with benefit in all cases. Infusion of peripheral blood progenitor cells and rabbit ATG was well tolerated with minor adverse effects. In 3 of 10 cases, cytomegalovirus (CMV) reactivation was demonstrated 15 to 35 days after transplantation. In only one patient was CMV reactivation symptomatic, with fever, arthralgias, and diarrhea, whereas the other two patients were asymptomatic, with infection being only a laboratory finding (see table 2). All patients were treated with ganciclovir, with disappearance of clinical symptoms and CMV antigen positivity in the blood. Adverse effects occurring in the 10 patients in close temporal relationship with mobilization, conditioning, and transplantation are reported in table 2. In the 2 to 3 months after ASCT, asthenia, fatigue, and fever were commonly observed. Cutaneous herpes zoster localized in the thoracic areas occurred in two cases, 6 and 9 months after ASCT, and was successfully treated with acyclovir.

No adverse effects or laboratory abnormalities were recorded with the monthly prolonged use of TD Gd. During the 3-month pretreatment period, 341 Gd-enhancing lesions were detected. The mean number of Gd-enhancing lesions per month per patient in the pretreatment period was 9 (range 1 to 38). The number of Gd-positive lesions decreased dramatically immediately after mobilization with CY and dropped to zero in 8 of 10 patients in the month after the conditioning therapy. In the fifth patient (table 3), a new enhancing area was detected at months +1, +2, and +3 after BEAM therapy, followed by complete suppression of MR activity. In the last patient, with a follow-up of only 4 months, in the first month after transplantation, two Gd-enhancing areas were still present, which, however, disappeared in the following months. The number of Gd-enhancing lesions per month before and after mobilization and after ASCT is detailed in table 3.

The number of new T2-weighted positive lesions paralleled data previously described for Gd-enhanced MRI. In 9 of 10 cases, no new T2 lesions were observed after ASCT. In the fifth patient, only one new T2 lesion was detected in each of the first 3 months after therapy. For comparison, 162 new T2-weighted lesions were detected during the 3 months of the pretreatment period. The difference between pre- and posttreatment period was highly significant (p < 0.000001, Mantel–Haenszel test), both for Gd-enhancing lesions and for new T2-weighted positive lesions.

Clinically, all patients improved slightly or remained stable. After 6 months, the median EDSS decreased to 6 (range 4 to 7) and the median Scripps scale increased to 70 (range 68 to 74). In one patient (Patient 4), after an initial improvement that lasted for 9 months, the clinical disturbances, characterized mainly by spastic paraparesis, resumed worsening. However, currently, after 15 months, the neurologic condition has been evaluated to be the same as before treatment. In the months after ASCT, the patients received only symptomatic therapies. Clinical data of patients before and after transplantation are reported in table 4.

Discussion.

ASCT completely suppressed MR-enhancing activity in 10 patients with rapidly progressive, severe MS unresponsive to conventional treatments. These results were assessed by performing serial monthly MR scans with TD Gd, which detects 70 to 80% more lesions than the standard dose.11 ASCT was well tolerated, but infections were frequent in the 2 to 3 months after transplantation, and in this period, the neurologic and general condition of patients slightly worsened, likely owing to infections and concomitant fever. Subsequently, in the following months, all patients improved and returned to the pretreatment neurologic status or to an ameliorated status. This study was designed mainly to obtain information on the effect of ASCT on a laboratory marker of disease activity such as MRI, and therefore clinical data must be interpreted with caution. After a median follow-up of 15 months (range 4 to 30 months), the neurologic condition of patients was evaluated as stable or slightly improved (see table 4). In one case (Patient 4), however, after improvement that continued until month 9, the disease resumed progression, although after 15 months after transplantation, EDSS is currently identical to baseline.

ASCT was first proposed for intractable systemic lupus erythematosus14 and was then performed as a new approach for severe autoimmune diseases unresponsive to and/or constantly relapsing after conventional immunosuppressive therapy. There are basically two ways of interpreting the therapeutic mechanism. The first postulates that an intense myelolymphosuppressive regimen, followed by ASCT with T-cell depletion either ex or in vivo, may provide what has been defined as a “window of time” free of memory T-cell influence, during which the maturation of new lymphocyte progenitors may occur without recruitment to anti-self-reactivity.15 A second and perhaps more simple explanation postulates an intense and prolonged immunosuppression, which is shown by the very long depletion of CD4+ and CD45 RA+ lymphocyte subpopulations.16

Independent of its interpretation, ASCT is considered an important new promising therapeutic procedure for refractory autoimmune disease because of lower transplant-related mortality and greater feasibility than allogeneic transplantation. However, according to the EBMT Registry, transplant-related mortality in autoimmune diseases is 9% at 1 year.9

MS is the disease that has been more frequently treated with ASCT. Recently, after the first pilot study,5 the Thessaloniki group reported the interim analysis of efficacy in the first 15 cases, who were conditioned with ATG to deplete lymphocytes in vivo, and added 9 patients on which additional CD34+ cell selection of the graft was performed.6 One patient died from aspergillosis 2 months after transplantation. This patient had received a lymphocyte-depleted graft in addition to ATG. It is likely that the terminal infection was related to profound and prolonged immunosuppression. Eighteen patients improved or stabilized after a median follow-up of 40 months; 9 of these 18 patients subsequently progressed, although their neurologic condition did not become worse than before transplantation. Eight of these 24 patients had a primary progressive clinical course, which deteriorated more frequently than secondary progressive patients. Eight severe cases of MS treated with BEAM and with in vivo or ex vivo T depletion without serious adverse events had (except in one case) an apparent stabilization of the disease.8 A report9 of six patients treated by incorporating TBI in the conditioning regimen noted improvement or stabilization of the clinical course in all cases. Recently, four patients with MS enrolled in a protocol of high-dose immunosuppression, with peripheral blood stem rescue, experienced neurologic worsening while receiving recombinant human G-CSF at the dosage of 10 to 16 μg/kg/day SC or IV for 4 consecutive days for the mobilization of stem cells. In our series of patients, the mobilization of progenitor cells was obtained with CY (4 g/m2) followed by subcutaneous G-CSF at the dosage of 5 μg/day, without any disease flare.17 Four additional cases of bone marrow transplantation in MS have just been reported.18

Other severe cases of MS worldwide have been treated with ASCT. There are now 102 patients in the EBMT Registry, with seven fatalities, due to complications related to the transplantation in five cases and disease progression in the other two cases (A. Tyndall and A. Fassas, EMBT Registry, personal communication). Data on the effect of ASCT on laboratory markers of disease activity were accordingly needed to evaluate whether this therapy merits further consideration. The total clearing of active lesions from the CNS, as demonstrated in this study, supports the further evaluation of this procedure for selected MS cases.

Other immunosuppressive therapies currently utilized in severe and active MS have also shown a profound effect on MR activity, although not of the same magnitude as reported herein. Mitoxantrone appears to be the most promising immunosuppressive therapy at present. It reduces MR activity by at least 90%19-20⇓ (evaluated with the standard dose of Gd) and decreases disease progression in secondary progressive forms of MS.21 CY is also utilized in severe forms of MS with different regimens and with contrasting results. With pulse intravenous administration at 1 g/m2, it has a marked effect on MR activity, reducing Gd enhancement22 almost to zero (evaluated with the standard dose of Gd). However, this effect is time limited, and resumed MR activity requires that the treatment be repeated and maintained. Interestingly, in our study, after CY at the dosage of 4 g/m2 utilized for mobilization, MR enhancing activity began to decrease and dropped to zero in three patients, as judged by MRI scans performed before BEAM treatment (see table 3).

The final impact of this procedure on the natural history of the disease remains to be established, especially because a clear relationship between MR activity and disease progression has not yet been demonstrated. A high MR activity rate has been indeed associated with periods of clinical activity23 and with relapse rate,24 but there is only a trend between Gd-enhancing lesion count and later impairment after 1 or 2 years.24 Neuropathologic studies25 have demonstrated that areas of Gd enhancement correspond to areas of breakdown of the blood–brain barrier and inflammation. Progression of the disease and disability is due not only to the inflammatory process but also to axonal degeneration, loss of oligodendrocytes, and failure of remyelination. Axons can be damaged after inflammation,26,27⇓ but axons devoid of myelin can also degenerate later, possibly for a perturbation of the local balance between neurotrophic and destructive factors. Progression of disease may therefore be due to events not necessarily related to inflammation. It is possible that even treatment with such a profound impact on the inflammatory phase of the disease may have a limited effect on the progression of disability. A possible discrepancy between the effect on MR activity and clinical efficacy is exemplified by cladribine. Cladribine has a pronounced and sustained effect on the presence of Gd-enhanced T1 brain lesions on MRI28 without, however, significant clinical effects on EDSS in progressive MS cases.29 In a multicenter controlled cladribine study, 30% of patients had a primary progressive MS course and approximately only 35% of patients in each treatment arm had Gd-enhanced T1 lesions at baseline.29 Selection of patients is probably crucial for evaluating the clinical response of high-dose immunosuppressive therapies; in our opinion, only secondary progressive severe cases with clinical and MR signs of active disease should be included.

The role of ASCT in the management of refractory treatment-resistant active MS still needs to be established. The current study demonstrates that it is particularly effective in suppressing the breakdown of the blood–brain barrier and in abrogating the inflammatory phase of the disease. However, due to the high mortality risk and the lack of data on efficacy, ASCT cannot be recommended as standard therapy for severe and rapidly worsening MS. After the demonstration that ASCT is active on laboratory markers of disease activity, as clearly shown in this study, its clinical efficacy must now be evaluated in controlled trials involving hematologic departments with established and certified experience in auto- and allogeneic transplantation30 and neurologic centers with experience in clinical trials in MS. At this time, it is unclear whether ASCT offers advantages in comparison with conventional immunosuppressive therapies such as mitoxantrone or CY. A randomized trial comparing ASCT with the most effective im-munosuppressive therapies should now be considered.

Appendix

The Italian GITMO-NEURO Intergroup that has actively contributed to this study also includes the following persons: G. La Nasa, E. Cocco, and V. Cherchi (Cagliari); A. Bosi, A. Repice, and A. Konze (Firenze); O. Figari, E. Capello, and L. Dogliotti (Genova); F. Papineschi, S. Mosti, and A. Abbruzzese (Pisa); and P. Di Bartolomeo, D. Farina, C. Iarlori, and A. Tartaro (Chieti).