A multicenter, randomized, double blind, placebo-controlled clinical trial of DL-3-n-butylphthalide in treatment of amyotrophic lateral sclerosis

To the Editor: DL-3-n-butylphthalide (NBP) is a synthetic drug based on a component extracted from seeds of Apium graveolens Linn. (Chinese celery). It reportedly has protective effects in amyotrophic lateral sclerosis (ALS) model mice.^[1] A randomized, double-blind, placebo-controlled trial was performed at 19 ALS clinical centers of the Chinese ALS association. All patients provided written informed consent. The Ethics Committee of Peking Union Medical College Hospital approved the study protocol (No. S621) and all patients signed informed consent. All centers followed the same research protocol.

Patients with definite or probable ALS according to the revised version of the El Escorial World Federation of Neurology criteria were recruited for the study.^[2] Included patients displayed progressive weakness onset within 18 months prior to the study. Age at onset was between 20 and 75 years. Forced vital capacity (FVC) was >70% of the predicted normal value. No patients had taken riluzole or edaravone in the 3 months before recruitment.

Exclusion criteria were as follows: Electromyography showing motor nerve conduction block and abnormal sensory nerve conduction; computed tomography or magnetic resonance imaging revealing lesions that may explain the patient's clinical presentation; dementia or mental disorder; serious heart, liver, kidney, or other related diseases; the patient's overall medical condition being such that participation may endanger their life; gastrointestinal disorders or surgery that may affect gastrointestinal absorption; a history of allergies, especially to medication; a requirement for ventilator-assisted breathing or tracheotomy; breastfeeding or pregnant patients; presentation of bulbar or thoracic symptoms/signs at onset; and difficulties taking medicine.

Patients were randomly assigned to be treated with NBP soft capsules (100 mg/tablet) or placebo (with appearance, size, color, dosage form, weight, taste, and smell as similar as possible to the experimental drug, but without the active ingredients). Patients were given two soft capsules (NBP or placebo) thrice daily for 12 months.

All patients were evaluated before treatment with NBP or placebo and followed up for 12 months. Functional evaluations were conducted every 3 months after enrollment. The primary endpoint was the Revised Amyotrophic Lateral Sclerosis Functional Rating Scale (ALSFRS-R) score after 12 months of treatment. Secondary efficacy outcomes were survival, tracheotomy, total Medical Research Council (MRC) score, percentage of predicted FVC, and the self-reported Clinical Global Impression (CGI) scale. The MRC scores were collected from 24 muscle movements; each movement was evaluated using the MRC scale (0–5) with a possible total score of 120. The CGI scale assessesd disease severity and was evaluated using a visual analog scale ranging from no symptoms to most serious symptoms (0–7).

Adverse events were observed every 3 months after treatment began and patients were instructed to report events at any time. Adverse events were also monitored for 3 months after patients terminated the study (either at the end of the study period or prematurely).

The sample size was calculated according to a previously published clinical trial of ALS patients. To provide 80% power to detect an adjusted mean difference between groups of 3 points on the ALSFRS-R score (30.2 ± 8.9 vs 33.2±8.0) and considering a dropout rate of 25% at the last visit, with a 1:1 randomization ratio between NBP and placebo group, at a two-sided alpha level of 0.05, we calculated that a sample size of 312 patients would be required approximately (156 in the NBP group and 156 in the placebo group).

Before treatment, 312 patients with ALS were recruited; 93 patients in the NBP group and 92 patients in the placebo group finished the study and were included in the analysis. The study flow chart is shown in [Supplementary Figure 1].

Baseline data were collected at the first visit before treatment. Both groups had a similar sex ratio, age at first visit, duration from disease onset, onset site, bulbar involvement at first visit, body mass index, ALSFRS-R score, predicted FVC, total MRC score, and CGI score [Supplementary Table 1].

After 12 months of treatment, the changes in endpoint parameters between baseline and 12 months were analyzed. There were no significant differences between the two groups in ALSFRS-R, total MRC, or CGI scores [Table 1]. The detailed changes in all parameters at each visit are shown in Supplementary Tables 2–5 and Supplementary Figures 2–5. There was a mild trend toward a smaller decrease in percentage of predicted FVC in the NBP group than in the placebo group. This result suggests that NBP might slow the decrease of respiratory function. Although NBP was given after disease onset, the treatment was started before respiratory muscles were involved, which is similar to how treatments are generally started before disease onset in ALS model mice. However, our results should be interpreted cautiously because more than half of patients were unable to complete the FVC test at the 12-month visit because of bulbar muscle weakness. These patients might also have had more serious respiratory muscle weakness, in addition to bulbar muscle weakness. Therefore, percentage of predicted FVC might not be a good outcome parameter for evaluating respiratory muscle function in ALS patients. Although we only recruited patients with limb symptoms at onset, most patients developed bulbar muscle weakness during the follow-up. Thus, nasal sniff may better assess respiratory function in ALS patients because it can be completed with bulbar muscle weakness.^[3]

Table 1.

Analysis of differences in changes in ALSFRS-R, total MRC, %FVC, and CGI from baseline to the last visit between DL-NBP group and placebo group.

			Between-group difference
Efficacy endpoint (n1/n2)	DL-NBP (mean ± SD)	Placebo (mean ± SD)	LS (mean ± SE)	95% CI	P-value
ALSFRS-R (76/84)	11.24 ± 8.03	11.93 ± 6.97	−0.69 ± 4.01	−3.03, 1.65	0.56
Total MRC score (55/63)	25.05 ± 16.30	25.16 ± 15.03	−0.10 ± 158.25	−5.82, 5.61	0.97
%FVC (43/39)	11.40 ± 17.83	18.73 ± 18.09	−7.34 ± 4.28	−15.24, 0.57	0.07
CGI (67/71)	−1.54 ± 1.09	−1.27 ± 1.15	−0.27 ± 0.27	−0.65, 0.11	0.27

%FVC: Percentage of predicted FVC; ALSFRS-R: Revised Amyotrophic Lateral Sclerosis Functional Rating Scale; CGI: Self-reported Clinical Global Impression scale; CI: Confidence interval; FVC: Forced vital capacity; LS: Least-squares; MRC: Medical Research Council; n1: Number of patients in the NBP group; n2: Number of patients in the placebo group; NBP: DL-3-n-butylphthalide; SD: Standard deviation; SE: Standard error.

During the study, 13 of 93 patients in the NBP group died and two underwent tracheotomy, while 6 of 92 patients in the placebo group died and none underwent tracheotomy (χ² = 4. 24, P = 0.04). However, these results should be interpreted cautiously because 88.6% (164/185) of patients were alive at the last visit; the findings may have been caused by the short disease duration before patient recruitment. Thus, a longer follow-up of survival and tracheotomy incidence should be performed to avoid sample bias.

Adverse events were reported in 56.5% (52/92) of patients in the placebo group and 68.8% (64/93) of patients in the NBP group (χ² = 2.99, P = 0.08). No severe adverse events related to treatment were reported. Adverse events with an incidence rate >1% included increased creatinine kinase, alanine aminotransferase, and aspartate transaminase, gastrointestinal symptoms, lower limb edema, dizziness, respiratory symptoms, and rash [Supplementary Table 6].

In the present clinical trial, the ALSFRS-R score was the primary endpoint, as in other clinical trials.^[4] However, ALSFRS-R scores might not be sufficiently sensitive to reflect mild deterioration in ALS patients. In most studies, ALSFRS-R scores slowly decrease during follow-up among all patients; however, in one study, 25% of 3132 patients did not exhibit any decline >6 months.^[5] Thus, a more accurate scale for evaluating changes in ALS patients should be used for future clinical trials.

The most serious study limitation was the high dropout rate at the last follow-up; furthermore, half of all patients were unable to complete the FVC test because of bulbar palsy.

In summary, we found that NBP is safe but did not improve ALSFRS-R scores in ALS. The identified difference in the percentage of predicted FVC between the two groups should be interpreted cautiously.

Conflicts of interest

This work was supported by CSPC NBP Pharmaceutical Co. Ltd.