35 

NASCIS 11,2

The original NASCIS trial compared two MP dosing regimens: what was then considered standard therapy for SCI (100 mg bolus of MP followed by 25 mg every 6 hours for 10 days) versus a regimen of a 1000 mg bolus followed by 250 mg every 6 hours for 10 days. No placebo arm was used because withholding steroids was considered to be unethical at that time. The two groups did not differ in neurologic outcome at 6 weeks, 6 months, or 1 year after injury. Although this is largely considered to be a “negative” trial, it laid the groundwork for the design and analysis of subsequent NASCIS trials.

NASCIS 2: Six-Month Report3

As NASCIS 1 progressed, new animal data suggested that higher doses of steroids administered for shorter periods might provide efficacy with fewer complications. These modifications were incorporated into NASCIS 2, a study that used a placebo group and that also included a group that received naloxone, which had proved promising in preclinical studies.

Release of Results

The 6-month outcome data from NASCIS 2 were published in the New England Journal of Medicine on May 17, 1990. In an unusual move, the Journal agreed to lift its customary embargo concerning prepublication release of a study’s findings, partly because the National Institute of Neurological Disorders and Stroke requested early release in the public interest. The results of NASCIS 2 were widely reported by the media roughly 6 weeks before their publication in the Journal.⁴ Thus, a month and a half before the actual data were available for them to review, health care workers were told that they essentially had to administer MP to SCI patients. When the publication finally appeared in print, many physicians contested the importance of the findings, which they thought had been tremendously overstated.^5,6 This bitterness about the way the NASCIS 2 results were initially disseminated probably accounts for much of the ongoing antagonism toward these trials.

Methodology

NASCIS 2 attempted to randomize SCI patients within 12 hours of injury to one of three groups: placebo, MP, or naloxone. MP was given as a bolus of 30 mg/kg of body weight over 15 minutes, followed 45 minutes later by a continuous infusion of 5.4 mg/kg/h for 23 hours. The naloxone dose was 5.4 mg/kg for the 15-minute bolus, followed 45 minutes later by 4.0 mg/kg/h for 23 hours.

Neurologic assessment was divided into motor function, sensation to pinprick, and sensation to light touch. For motor testing, 14 muscle segments (innervated by 14 different levels of the spinal cord from cervical to sacral) were each given one of six scores: 0 (no contraction), 1 (reduced contraction), 2 (active movement without antigravity, i.e., side-to-side but not upward), 3 (active movement against gravity), 4 (reduced function but active movement against resistance), or 5 (normal function). Possible scores ranged from 0 to 70.

For sensory testing, 29 segments of the spinal cord from C2 to S5 were evaluated bilaterally for pinprick and light touch according to the following scale: 1, absent; 2, decreased; or 3, normal. Possible scores ranged from 29 (absent at all levels) to 87 (normal at all levels).

The primary end point was the change in neurologic score between baseline assessment and follow-up evaluation. Additional preplanned analyses would analyze the effects of the protocol, of the time the MP dose was received, and of the degree of neurologic loss (complete or incomplete).

Each patient was placed into one of five motor and sensory categories. The five motor categories were quadriplegic, paraplegic, quadriparetic, paraparetic, and normal. Sensory categories were as follows: analgesic and anesthetic at or above T1, analgesic and anesthetic below T1, hypalgesic and hypesthetic at or above T1, hypalgesic and hypesthetic below T1, and normal.

The analysis of neurologic scores used only the right side of body, and to simplify the presentation of results, only data from the right side were shown in the initial publication. However, when the analyses were performed on the left side, essentially identical results were obtained.

Results

Atotal of 487 patients were randomized: 162 to MP, 154 to naloxone, and 171 to placebo.

Six-Week Results

At the 6-week outcome assessment of all patients in the MP and placebo groups, there was no significant difference in neurologic change scores between these groups. However, if only those patients treated within 8 hours were considered (remember that one of the a priori hypotheses was that effects of treatment would be influenced by how quickly the drug was given), the MP group showed significantly more improvement in motor and light touch scores, with a trend toward improvement in pinprick scores. The 8-hour cutoff interval was chosen because the average time from injury to bolus dose for all patients in the entire study was 8.7 hours,3 with a median time of 8.5 hours.⁷ The exact number of patients treated within 8 hours drops to 66 MP patients and 69 placebo patients (62 and 67, respectively, at 6 months) out of the 487 entered into the trial.

Another a priori hypothesis was that any treatment effects would be influenced by the severity of injury. Accordingly, patients treated within 8 hours were classified as either plegic with total sensory loss, plegic with partial sensory loss, or paretic with variable sensory loss. At 6 weeks, the first group had more improvement in motor function than placebo, with strong trends for sensory improvement. The small number of patients in the second group demonstrated no differences between MP and placebo, whereas the third group exhibited a trend toward improvement only in motor score.

Six-Month Results

At 6 months after injury, the entire MP group showed significantly more improvement than the placebo group in pinprick and light touch scores, but apparently not in motor improvement. However, among patients treated within 8 hours of injury, the MP group demonstrated significantly more improvement than placebo in all three measures: motor, pinprick, and light touch.

At 6 months postinjury, MP patients who were plegic with total sensory loss had more improvement than placebo patients in all three neurologic measures. On the other hand, patients who were plegic with partial sensory loss showed the same amount of improvement in neurologic scores regardless of group. Among those who were paretic with variable sensory loss, only the amount of motor improvement was significantly greater in the MP group than in the placebo group.

Among patients who were treated after 8 hours, there were no differences in neurologic change scores related to treatment. Likewise, patients treated with naloxone demonstrated no statistically significant differences in improvement.

The above analyses were done on all randomized patients; that is, “intent-to-treat analyses.” If only those who received drug within the protocol’s time limits are analyzed, the differences in favor of MP become larger.

If patients were classified as quadriplegic, paraplegic, quadriparetic, or paraparetic, and if they were treated within 8 hours, the odds of improving by a full category at 6 weeks tended to be higher in the MP group. The degree of such improvement was less evident at 6 months and was not seen if drug was given after 8 hours.

For MP, naloxone, and placebo patients, wound infections occurred in 7.1%, 3.3%, and 3.6% of patients, respectively, and gastrointestinal bleeding occurred in 4.5%, 2.0%, and 3.0%, respectively. These differences were not significantly different.

NASCIS 2: One-Year Report⁸

The numbers of patients who received study drug within 8 hours of injury and who were assessed 1 year after injury were 62 for MP, 56 for naloxone, and 65 for placebo.

There were no significant differences in neurologic function by treatment group. However, if only those patients who received their drug bolus within 8 hours were considered, those in the MP group demonstrated significantly greater motor recovery (17.2 versus 12.0; p =.03). Improvements in pinprick and light touch were not statistically significant.

Among all patients who were randomized more than 8 hours from injury, there were strong but statistically nonsignificant trends for the MP and naloxone groups to recover less neurologic function than the placebo group.

Analysis of changes in neurologic function scores 1 year after injury in patients who received the study drug within 8 hours demonstrated significant improvement in motor scores for patients who were plegic with total sensory loss or paretic with variable sensory loss, but not in the small group of patients who were plegic with partial sensory loss. When patients were analyzed for improvement from an abnormal neurologic category (quadriplegic, paraplegic, quadriparetic, or paraparetic) to a higher level, there was a nonsignificant trend for more improvement with MP.

The groups did not differ significantly for any of 13 complications, most of which were reported at the 6week follow-up. For all follow-up periods, only three complications approached statistical significance, and all occurred in comparisons of the naloxone group to the placebo group. The authors note that complications, even if related to the MP treatment, were manageable and were clearly outweighed by the potential benefits of improved recovery from SCI.

NASCIS 3: Six-Month Report⁹

Methodology

NASCIS 3 randomized patients into three groups: 24 hours of MP (24MP), 48 hours of MP (48MP), or 48 hours of the free radical scavenger tirilazad mesylate (TM). Because of the presumed importance of ongoing lipid peroxidation and hydrolysis as deleterious pathophysiologic sequelae of SCI, it was thought that extension of the duration of MP therapy beyond 48 hours might confer more benefit than a 24-hour period of administration. In addition, it was thought that administration of the lipid peroxidation inhibitor TM might facilitate neurologic recovery in a manner comparable to that of MP, but with fewer complications.