Amyotrophic lateral sclerosis (ALS) is a rapidly progressive neurodegenerative disease that primarily affects upper and lower motor neurons, resulting in death three to five years after symptom onset in the majority of patients (1). The pathomechanism of ALS is largely unknown, but research efforts have been focused on several hypothetical models such as glutamate excitotoxicity, oxidative damage, mitochondrial dysfunction, inflammation and apoptosis (2). Numerous clinical trials have been performed, but unfortunately all failed to demonstrate any beneficial effects on the disease progression or survival, with the exception of riluzole.

Ursodeoxycholic acid (3α, 7β-dihydroxy-5β-cholanic acid, UDCA), a hydrophilic tertiary bile acid, is a major component of black bear bile that was used in traditional Chinese medicine (3). It is the current mainstay of treatment of primary biliary cirrhosis as the only drug approved by the US FDA, and has been increasingly used for the treatment of various cholestatic liver diseases such as primary sclerosing cholangitis and progressive familial intrahepatic cholestasis (4, 5). Although the mechanisms of action is not completely understood, UDCA and its taurineconjugated derivative (TUDCA) have been shown to strongly inhibit apoptosis in different type of cells by either stabilizing the mitochondrial membrane, or modulating the expression of specific upstream targets of apoptosis (6-11). TUDCA improved the survival and function of nigral transplants in a rat model of Parkinson's disease (12), and the neuroprotective effect was also reported in a transgenic mouse model of Huntington's disease (13), and rat models of ischemic and hemorrhagic stroke (14, 15).

ALS must be such a complex disease with numerous factors contributing to the degeneration of motor neurons, that an approach targeting a single molecule or process may not be effective (2). In this respect, UDCA may be a promising therapeutic candidate because of its multiple mechanisms of cytoprotective actions which may include antiapoptotic, immunomodulatory and antioxidant effects (6, 16, 17). Despite these therapeutic potentials, however, clinical application of UDCA has been limited to the hepatobiliary disorders exclusively. This is because of its unique pharmacokinetic properties which prevent the molecule from being delivered through systemic circulation, for example, to the central nervous system at potentially therapeutic concentrations (18). UDCA capsules or tablets, for instance, contain acid crystals with a very low solubility at pH < 7, and after administration of the pharmacological dose (10-15 mg/kg/day) UDCA is absorbed through passive non-ionic diffusion, which is largely limited by dilution mostly in the small intestine (18). Moreover, a large proportion of UDCA taken up from the portal blood is conjugated with glycine or taurine during its first hepatic passage, and actively secreted into the bile (18).

A new UDCA formula, oral solubilized UDCA (Yoo's solution; US patent 6251428, 7166299, 7303768 with PTC), was developed to broaden its therapeutic targets beyond the liver and biliary tracts eluding those pharmacokinetic limitations such as the enterohepatic circulation and high first-pass metabolism. The new formula contains intact UDCA molecules and soluble starch conversion product that enables UDCA to remain in solution. The product is highly soluble in water with a solubility of 80 mg/mL, and stable from pH 1 to 14 without producing precipitate (19, 20). It can deliver therapeutically effective amounts of solubilized intact UDCA to the systemic circulation with a Cmax of 20.4 µg/mL at a T_max of 15 min in human blood after oral administration of single dose containing 500 mg of UDCA (20). Moreover, a preliminary study showed that UDCA can be effectively delivered to the central nervous system (cerebrospinal fluid, CSF) with oral administration of solubilized UDCA (described in the methods). The present study was performed to evaluate the efficacy and safety of oral solubilized UDCA treatment in patients with ALS.

The trial profile is illustrated in Fig. 1. A total of 80 patients were assessed for eligibility, and 64 patients were enrolled in this study (one in error). Reasons for exclusion were the AALSRS total score greater than 120 or disease duration longer than 60 months (n = 10), severe depression (n = 3), and clinically possible ALS (n = 2) and cognitive impairment (n = 1). Forty of 63 patients were assigned to receive oral solubilized UDCA in the first period of study, and the other 23 patients to receive placebo. During the first period, 19 of 40 (47.5%) patients treated with oral solubilized UDCA and 15 of 23 (65.2%) patients treated with placebo completed the study. All patients who completed the first period switched to receive the other treatment in the second period. The patients who dropped out during the first period of study were also scheduled to enter the second period of study after originally planned three months had passed with the other treatment given for the second period. Thus, 31 of 40 (77.5%) randomized to oral solubilized UDCA and 19 of 23 (82.6%) randomized to placebo for the first period switched to receive placebo and oral solubilized UDCA, respectively, for the second period of study. During the second period, 6 of 19 (31.6%) treated with oral solubilized UDCA and 13 of 31 (41.9%) treated with placebo completed the study. Drop-out rates were higher during the second period (62%) than the first (46%), and also higher whilst on oral solubilized UDCA (57.6%) than placebo (48.1%). Only 16 out of 63 enrolled patients completed the entire cross-over trial of 8-month, receving both treatments. Missed visits was the most common reason for drop-out (62%), followed by lost to follow-up (18%), discontinuation due to adverse events (12%) and a rapid progression (8%).

The demographic and clinical characteristics of patients treated with either oral solubilized UDCA or placebo are outlined in Table 1. The patients treated with oral solubilized UDCA and placebo were comparable to each other at baseline with respect to sex, age at onset, disease duration from symptom onset to diagnosis, onset region, severity, and level of diagnostic certainty. We constructed the dataset for efficacy analysis separately for each outcome, because the AALSRS and FVC measures were missing more frequently than the ALSFRS-R. The results of efficacy analysis for each outcome were summarized in Table 2. As for the PP datasets, the slope of AALSRS was 1.17 points/month lower while the patients were treated with oral solubilized UDCA than with placebo (2.3 vs 3.47 points/month, 95% CI for difference 0.08-2.26, P = 0.037). However, the slope of neither ALSFRS-R nor FVC (%) did not show any significant differences between treatments. Efficacy analysis from the PP' datasets revealed the same results as from the PP datasets. The slope of AALSRS was 1.63 points/month slower while the patients were treated with oral solubilized UDCA than with placebo (2.24 vs 3.88 points/month, 95% CI for difference 0.60-2.68, P = 0.004), whereas the ALSFRS-R and FVC (%) deteriorated without significant differences between the two treatments. To infer the clinical meaning from the slope data, Kaplan-Meier estimation was performed using a 20-points progression in the AALSRS total score as the end point which was reported to represent a clinically meaningful lifestyle change. However, because of the short study period of this trial, the time to reach this endpoint had to be estimated in all patients, and was calculated by assuming that the slope of AALSRS would be linear from the start of treatment through the endpoint. As for the PP' datasets, the time to a 20-points progression in the AALSRS total score was estimated to be delayed by 14.9 months in oral solubilized UDCA-treated group compared to placebo-treated group (22.5 vs 7.6 months, 95% CI for difference 2.97-26.8 months, P = 0.018) (Fig. 2). Additionally, we compared clinical features between 60 patients who dropped out and 53 patients who completed at least one part of the study. At baseline, sex, age at onset, age at enrollment, disease duration from symptom onset to diagnosis, onset region, severity, and level of diagnostic certainty were not significantly different between two groups (all, P > 0.05). Safety data from each period of study were pooled to provide a whole picture of the adverse event profile. The population for safety analysis consisted of 74 patients who took oral solubilized UDCA, and 70 patients who took placebo (Table 3). Not included for safety analysis were the patien ts who did not take the study drug or placebo at all, 6 patients assigned to oral solubilized UDCA and 10 patients to placebo.

A total of 25 occasions of adverse events were reported in 17 patients (23.0%) treated with oral solubilized UDCA, and 12 occasions in 10 patients (14.3%) treated with placebo. Except for the expected complications of ALS such as dyspnea and dysphagia, adverse events that could be possibly attributed to the study drug or placebo were reported in 12 patients (16.2%, 16 occasions) treated with oral solubilized UDCA, and in 6 patients (8.6%, 7 occasions) treated with placebo. Gastrointestinal adverse events were significantly more common in patients treated with oral solubilized UDCA (P < 0.05), but serious events were reported in only one patient who discontinued the drug due to abdominal pain, anorexia and vomiting. None of the 80 patients who were randomized died during the 8-month period of study, with tracheostomy performed in 2 patients.

The results of this study suggest that oral solubilized UDCA therapy may have a beneficial effect on the rate of functional decline in patients with ALS. As for the slope of AALSRS, the difference between oral solubilized UDCA and placebo was 1.17 and 1.63 points per month on the average for the PP and PP' datasets, respectively, which correspond to 34% and 42% relative slowing of the rate of deterioration in this scale. Although estimated, the time to a 20-points progression in the AALSRS also showed that oral solubilized UDCA treatment could delay this endpoint by more than 1 yr compared to placebo. Although it was not statistically significant, the efficacy data in the other outcomes also showed a trend favoring oral solubilized UDCA. Although gastrointestinal adverse events developed more frequently in patients treated with oral solubilized UDCA, these were mostly tolerable and oral high dose solubilized UDCA was well tolerated in vast majority of patients. Taken together, this study suggests a therapeutic potential and acceptable safety profile of oral solubilized UDCA in ALS patients.

However, no firm conclusion cannot be made on the efficacy of oral solubilized UDCA because of several critical problems and limitations of this study which should be acknowledged herein. First of all, the major problem was the large drop-out rate in this trial. In fact, only about a fourth of the initially randomized patients completed both parts of the study with missed visits, followed by lost to follow-up, being the most common drop-out reason. The patients, who had already severe disease at enrollment or became progressively more disabled during the study, tended to refuse to visit the clinic for outcome measures. These patients were also hard to keep contact with for consistent follow-up. Thus, it is likely that the broad inclusion criteria regarding disease severity, and the lack of adherence to visit schedule for outcome measures account for the large attrition rate. The second problem of this study is the way data were analyzed, i.e., 'completers-only' analysis. Since only the patients who completed the planned treatment course were included for efficacy analysis, this does not conform to the intention-to-principle. We cannot rule out faster declines for patients after their dropout, and excluded cases can cause a bias, although baseline clinical features of patients completing the study and patients dropping-out the study were not different. However, we could not apply the intention-to-treat principle, since the attrition rate was too large for us to perform any valid imputation procedure for missing data. Other limitations of this study are the short treatment period and limited number of datapoints to assess the rates of functional deterioration. Treatment period of 3 months might not be sufficient duration to capture a small effect. In addition, the outcome measures may not be reproducible to give accurate statistics with only two datapoints. Finally, it also should be pointed out that all the patients who were excluded due to their rapid progression rates (> 8 points/month) were those taking oral solubilized UDCA.

An ethical concern about treatment with placebo was the main reason we adopted a cross-over design instead of conventional two parallel arm study (27). It was also expected that the crossover design would be more efficient in detecting the same effect with a smaller sample size, since within-patient variability may be smaller than between-patient variability when the deterioration rates of functional scales are used as endpoints (28, 29). However, there might be disadvantages to the cross-over trial as well, and some of these indeed caused difficulties for analysis and interpretation of our study. First, there may be a bias which is related to the unequal randomization for the first treatment period. It is likely that merely the sequence of treatment may explain the difference between treatments, although we adjusted the effects of sequence and period using ANCOVA to evaluate the effect of treatment itself. The other problem is a treatmentperiod interaction, so-called a carry-over effect. The wash-out period of 1 month may be insufficient to completely eliminate the effect of treatment given before. However, we could exclude the inflation of alpha-error resulting from the carry-over effect, because any persistence of the drug effect would have reduced the magnitude of difference during the second treatment period.

It is worth emphasizing that UDCA can be effectively delivered to the central nervous system (CSF) with oral administration of solubilized UDCA in our study. A recent pilot study by Parry et al. also showed that UDCA was well absorbed after oral administration and crossed the blood-brain barrier in a dose-dependent manner in ALS patients (30). Eighteen patients were received UDCA in tablet formula at doses of 15, 30, and 50 mg/kg of body weight per day and the concentration of UDCA in CSF ranged 43.1 to 281.7 nM/L after 4 weeks of treatment. These results suggest that oral UDCA is well delivered to CSF of ALS patients and the therapeutic effects deserve to be expected although it remains to be elucidated which formula or dose of UDCA is more tolerated, absorbed, delivered and effective in ALS patients.

Although the mechanisms of cytoprotection of UDCA remain to be elusive, it does not seem to be limited to hepatic cells. UDCA modulates the apoptotic threshold in both hepatic and non-hepatic cells through the downregulation of p53 pathway (10, 11). Recently, a study using oral solubilized UDCA demonstrated that UDCA treatment can suppress cisplatin-induced p53 accumulation and apoptosis in mouse sensory neurons (19). It was also reported that this highly soluble and acid stable UDCA formula can scavenge the reactive oxygen species and prevent apoptosis in Helicobacter pylori-induced gastritis (20). In line with these previous reports, we also found that the new UDCA formula improved the motor performance and survival of G93A SOD1 mutant transgenic mice (in preparation). Therefore, the therapeutic effect of UDCA may be ascribed to the successful delivery of UDCA to the central nervous system and its multiple cytoprotective actions.

Oral solubilized UDCA treatment seems to be tolerable in ALS patients, but regarding the efficacy we could not make any conclusions due to several problems and limitations of this pilot study as discussed above. Still, the effect of UDCA deserves to be expected in ALS, due to its multiple neuroprotective mechanism. As we know, a phase II trial for the efficacy and tolerability of tauroursodeoxycholic acid, taurine conjugate of UDCA in ALS patients is also ongoing (NCT00877604) by Italian researchers. Further large scale study with minimal attrition rate is warranted to confirm the efficacy and safety of oral solubilized UDCA for patients with ALS.