- Email: [email protected]
Tolvaptan and Neurocognitive Function in Mild to Moderate Chronic Hyponatremia: A Randomized Trial (INSIGHT)
Original Investigation Tolvaptan and Neurocognitive Function in Mild to Moderate Chronic Hyponatremia: A Randomized Trial (INSIGHT) Joseph G. Verbalis, MD,1 Howard Ellison, MD,2 Mary Hobart, PhD,3 Holly Krasa, MS,3 John Ouyang, PhD,3 and Frank S. Czerwiec, MD, PhD,3 on behalf of the Investigation of the Neurocognitive Impact of Sodium Improvement in Geriatric Hyponatremia: Efficacy and Safety of Tolvaptan (INSIGHT) Investigators* Background: This trial assessed the effect of tolvaptan on cognition, gait, and postural stability in adult patients with mild to moderate asymptomatic hyponatremia. Study Design: Phase 3b, multicenter, randomized, double-blind, placebo-controlled, parallel-group pilot study. Setting & Participants: 57 men and women 50 years or older with chronic asymptomatic euvolemic or hypervolemic hyponatremia (serum sodium concentration .120-,135 mEq/L) at 16 sites. Intervention: Patients were randomly assigned 1:1 to receive tolvaptan or matching placebo beginning at a dose of 15 mg/d, with titration to 30 or 60 mg/d based on change in serum sodium concentration and tolerance. Outcomes: Primary: change from baseline in the neurocognitive composite score of speed domains. Secondary: changes from baseline in individual neurocognitive domain scores, overall neurocognitive composite score, gait and postural stability test results, and serum sodium concentrations. Results: Mean serum sodium concentration increased from 129 to 136 mEq/L in the tolvaptan group and from 130 to 132 mEq/L in the placebo group (P , 0.001). There was no difference in overall neurocognitive composite scores of speed domains between groups, except for the psychomotor speed domain, which was statistically improved following hyponatremia correction with tolvaptan (treatment effect, 0.27; 95% CI, 0.04-0.51; P 5 0.03). Limitations: There were some imbalances between treatment groups in baseline neurocognitive function scores and some baseline test results were near normal, leaving little opportunity for improvement. Formal sample size calculations were not performed because this was a pilot study. The study population was small (n 5 57) and treatment was of short duration (3 weeks). The primary end point of the study was not significant; thus, subgroup analyses are subject to errors of multiplicity and should be regarded as hypothesis generating. Conclusions: Tolvaptan was effective in reversing chronic hyponatremia, and this correlated with improvements in results of a variety of neurocognition tests, particularly rapid motor movements, which tended to reverse following return to a low baseline serum sodium concentration after treatment withdrawal. Am J Kidney Dis. 67(6):893-901. ª 2016 by the National Kidney Foundation, Inc. INDEX WORDS: Chronic hyponatremia; serum sodium; tolvaptan; vasopressin receptor antagonist; aquaresis; gait; postural stability; cognition; neurocognitive function; bone resorption; bone metabolism; osteoporosis; randomized controlled trial (RCT).
H
yponatremia is associated with a variety of underlying diseases and conditions, and the incidence of chronic mild hyponatremia increases with age.1,2 Chronic hyponatremia is often labeled as “asymptomatic,” and more subtle neurologic signs and symptoms of chronic hyponatremia associated with impaired mental functioning may be unrecognized or ascribed to “normal aging.” Subclinical impairment of mental functioning can be detected in patients with grossly normal results from neurologic tests and MiniMental State Examinations, leading to attention deficits similar to impairments seen after the ingestion of
moderate amounts of alcohol.3 Patients with mild chronic hyponatremia can manifest gait instability, and chronic hyponatremia has been associated with increased falls and fractures.3-5 Studies in experimental animals have indicated that hyponatremia also causes bone resorption, and analysis of epidemiologic data has shown a significantly increased odds ratio for osteoporosis in patients with hyponatremia,6 which was confirmed in 3 subsequent studies.7-9 Current options for treating hyponatremia are not optimal. This is particularly the case in the outpatient setting, in which fluid restriction is poorly effective and
From the 1Division of Endocrinology and Metabolism, Georgetown University, Washington, DC; 2Rockdale Medical Research Associates, Conyers, GA; and 3Otsuka Pharmaceutical Development and Commercialization, Inc, Rockville, MD. * A list of the INSIGHT Investigators appears in the Acknowledgements. Received July 1, 2015. Accepted in revised form December 28, 2015. Originally published online February 10, 2016.
Trial registration: www.ClinicalTrials.gov; study number: NCT00550459. Address correspondence to Joseph G. Verbalis, MD, Division of Endocrinology and Metabolism, 232 Bldg D, 4000 Reservoir Rd NW, Washington, DC 20007. E-mail: [email protected] 2016 by the National Kidney Foundation, Inc. 0272-6386 http://dx.doi.org/10.1053/j.ajkd.2015.12.024
Am J Kidney Dis. 2016;67(6):893-901
893
Verbalis et al
patients frequently do not adhere to fluid intake recommendations.10 Tolvaptan is a vasopressin receptor antagonist that induces water diuresis (aquaresis) without depletion of electrolytes. In 2 large randomized trials, tolvaptan was shown to increase serum sodium concentrations, in many cases to normalization.11 An open-label extension study showed continued efficacy without significant toxicity for treatment durations as long as 4 years.12 The major objective of the present trial was to determine the effect of oral tolvaptan on measures of cognition, gait, and postural stability in adult patients with mild to moderate hyponatremia and absence of overt neurologic symptoms. Exploratory objectives included evaluation of the effect of oral tolvaptan on self-reported cognitive variables and markers of bone metabolism.
METHODS Study Design The Investigation of the Neurocognitive Impact of Sodium Improvement in Geriatric Hyponatremia: Efficacy and Safety of Tolvaptan (INSIGHT) was a multicenter, randomized, doubleblind, placebo-controlled, parallel-group, titration-to-effect trial (Table S1, available as online supplementary material). The trial consisted of a 30-day screening period, 21-day treatment period, and 7-day posttreatment period (Fig 1A). During the screening period, attempts were made to identify and eliminate the underlying cause of hyponatremia, and patients were then treated using standard hyponatremia therapies. Patients who remained hyponatremic were randomly assigned to tolvaptan or placebo. The primary objective of this trial was to determine the effect of tolvaptan on change from baseline in a single composite score
comprising the sum of the z scores of individual cognitive variables under the domains of reaction time, psychomotor speed, and processing speed (hereafter referred to as the neurocognitive composite score of speed domains; Fig 1B) in patients with chronic mild to moderate hyponatremia after correction of serum sodium concentrations to normal ranges. Secondary and exploratory measures are described under Outcome Measures. The trial was conducted under the investigational new drug exemption of the US Food and Drug Administration in compliance with Good Clinical Practice guidelines, the sponsor’s standard operating procedures, and ethics principles for the protection of human research subjects in the Declaration of Helsinki. The protocol, amendments, and informed consent form were reviewed and approved by the governing institutional review board (IRB) or ethics committee of each investigational center prior to trial start at that center (central IRB: Chesapeake Research Review, CRRI 0607006; local IRB: Georgetown IRB, 2007-538; local IRB: University of Virginia IRB, 13606). Patients (or their guardians or legal representatives) provided written informed consent before any trial-related procedures were performed.
Patients Men and women 50 years or older with chronic asymptomatic hyponatremia (serum sodium concentration .122-,135 mEq/L) in the euvolemic or hypervolemic state were eligible for the trial. Two patients with serum sodium concentrations of 121 and 122 mEq/L were included following review. At least 50% of patients were targeted to have serum sodium concentrations # 130 mEq/L at baseline. Detailed inclusion and exclusion criteria are provided in Table S1.
Study Treatment Patients were randomly assigned 1:1 with stratification by baseline serum sodium concentration (,130 or $130 mEq/L) to receive oral tolvaptan or matching placebo tablets beginning at a dose of 15 mg/d, with titration up to 30 or 60 mg/d. A forced
Figure 1. (A) Study schematic and (B) neurocognitive composite score of speed domains (NCSSD). *Daily titration based on serum sodium levels. yThere were 26 completers per group analyzed for efficacy. 894
Am J Kidney Dis. 2016;67(6):893-901
Impact of Subclinical Hyponatremia titration in both treatment arms to 60 mg/d by days 3 to 7 was based on the patient’s change in serum sodium concentration and clinical tolerance. Titration occurred daily until the patient’s serum sodium concentration was increased and maintained at $138 mEq/L or the patient could not tolerate further titration. During this period, fluid restrictions were loosened or suspended until the patient’s response to therapy could be evaluated, typically during the first few days of therapy. Fluid restriction was to be reinstituted at any time during the trial in patients whose serum sodium concentrations did not improve or worsened. Patients remained on tolvaptan treatment for up to 21 days. Individual patient participation was approximately 60 days (30-day screening period, 21-day treatment period, and 7-day follow-up).
Outcome Measures The primary efficacy end point was change from baseline to day 22 in neurocognitive composite score of speed domains. Secondary efficacy end points were the effect of tolvaptan treatment on: (1) changes from baseline in individual neurocognitive domain scores included in the neurocognitive composite score of speed domains (reaction time, psychomotor speed, and processing speed); (2) change from baseline in overall neurocognitive composite score (composite score of the following domains: reaction time, psychomotor speed, processing speed, continuity of attention, working memory/executive functions, quality of episodic verbal memory, and postural stability); (3) changes from baseline in postural stability, Timed Up and Go test, and Romberg test; (4) change from baseline in serum sodium concentration; and (5) correlations of changes from baseline in neurocognitive domain scores with changes in serum sodium concentrations. Exploratory efficacy end points included changes from baseline in cognitive
variables, Hyponatremia Disease-Specific Survey, and 12-Item Short Form Health Survey (SF-12) component scores and changes from baseline in serum osteocalcin concentration and urine N-telopeptide (NTx)-creatinine ratio. A bone resorption index was calculated for each patient as change from baseline in NTx-creatinine ratio divided by serum osteocalcin concentration. Secondary safety analyses were conducted to determine the effect of tolvaptan treatment on safety as determined by adverse events (AEs), vital signs, electrocardiograms (ECGs), and clinical laboratory tests.
Efficacy Measurements Efficacy assessments were conducted at baseline/day 1, testing day (day 22)/early termination, and the follow-up visit (day 28). Specimens for serum chemistry and urine pregnancy tests (if applicable) were collected and processed according to each trial center’s routine procedures. All specimens were sent to a central laboratory for analysis. Neurocognitive tests were administered using the Cognitive Drug Research system.13 Training of research personnel on the Cognitive Drug Research computerized assessment system was done prior to the first day of dosing of the trial to ensure an optimal level of performance for the baseline assessment. The purpose of training was to help overcome initial test anxiety, familiarize patients with the procedures, enable the development of strategies for task performance, and overcome any initial practice effects.14 Each patient completed 2 training sessions prior to the start of the trial. Training sessions were to be no more than 14 days apart to reduce practice effects. (See Neurocognitive Battery Methods in Item S1 for detailed methods relating to neurocognitive and postural stability testing.)
ENROLLMENT
Assessed for eligibility (n=117) Excluded (n=61) Not meeting inclusion criteria (n=40) Declined to participate (n=21)
Am J Kidney Dis. 2016;67(6):893-901
Placebo
ALLOCATED
Allocated to intervention (n=29) Received allocated intervention (n=29) Did not receive allocated intervention (n=0)
Allocated to intervention (n=27) Received allocated intervention (n=27) Did not receive allocated intervention (n=0)
FOLLOW-UP
Figure 2. Patient enrollment and outcomes.
Tolvaptan
Lost to follow-up (n=0) Discontinued intervention (n=3) Reason(s): – Adverse event (n=2) – Withdrew consent (n=1)
Lost to follow-up (n=0) Discontinued intervention (n=1) Reason(s): – Withdrew consent (n=1)
ANALYSIS
Randomized (n=56)
Analyzed (n=26) Excluded from analysis (n=3) Reason(s): – Missing baseline or postbaseline observation (n=3)
Analyzed (n=26) Excluded from analysis (n=1) Reason(s): – Missing baseline or postbaseline observation (n=1)
895
Verbalis et al Table 1. Patient Characteristics and Demographics
Statistical Methods Based on available data for serum sodium concentration improvements with tolvaptan, it was determined that about 30 patients per arm would be needed to see significant treatment effect differences in serum sodium concentrations. The ability to detect neurocognitive end point differences with similar patient numbers has been reported in neurocognitive trials involving other agents and disorders.15-19 Analysis of covariance with factors of treatment, disease severity (,130 or $130 mEq/L at baseline), age, severity-age interaction, and covariate baseline for individual measurements was used to fit the primary end point using the intention-to-treat data set. The estimated treatment effect and its 95% confidence interval (CI) were provided under the model, along with the P value. A 2-sided a 5 0.05 was applied to the primary analysis. The primary analysis was based on observed cases. (See Item S1 for detailed methods of analysis of the secondary and exploratory efficacy end points.) Safety analysis was conducted based on the safety data set, which was defined as all patients who received 1 or more dose of trial medication. Safety variables analyzed included AEs, clinical laboratory test results, vital signs, and ECGs. In general, summary statistics of changes from baseline were provided for safety variables (when applicable) based on all available data. Descriptive statistics were used to describe differences in AEs, vital signs, ECGs, and clinical laboratory findings between the tolvaptan and placebo groups.
RESULTS Patients In all, 57 men and women were enrolled at 16 US sites in September 2007 to March 2009; a total of 53 patients were analyzed for efficacy (26 tolvaptan and 27 placebo), and 57, for safety (29 tolvaptan and 28 placebo; Fig 2). The 2 patient groups were relatively well matched for relevant baseline characteristics (Table 1). Compared with normative values from agematched individuals, patients in this trial at baseline had significantly impaired cognitive performance on measures of correct choice and speed of choice, modestly impaired gait and postural stability, and elevated bone turnover marker levels, but preserved cognitive performance on measures relying on word and number recognition (Fig 3). However, mean SF12 scores were generally higher than those of patients enrolled in previous trials of tolvaptan (SALT-1 and SALT-2 [Study of Ascending Levels of Tolvaptan in Hyponatremia 1 and 2]) for both the physical and mental component summaries (Table S2).11 Serum Sodium Concentrations Mean serum sodium concentrations increased from 129 to 136 mEq/L in the tolvaptan group and from 130 to 132 mEq/L in the placebo group (P , 0.01 in treatment comparison); levels in the tolvaptan group returned to baseline after withdrawal (Fig S1). Dose titration to the 30- and 60-mg dose levels (by days 3-7) was based on the patient’s change in serum sodium concentration (to $138 mEq/L [$138 mmol/L]) 896
Male sex Race White Non-Hispanic Age, y Baseline serum sodium, mEq/L Cause SIADH/other Cirrhosis CHF Baseline medication Diabetes Psycholeptic Psychoanaleptic MMSE score
Tolvaptan 15-60 mg (n 5 29)
Placebo (n 5 28)
14 (48)
9 (32)
29 (100) 29 (100)
26 (93) 25 (89)
71.1 6 10.2 (50-86) 129.38 6 3.31
71.3 6 9.7 (50-87) 129.81 6 3.04
29 (100) 0 (0) 0 (0)
26 (93) 2 (7) 0 (0)
1 (3.4) 8 (27.6) 10 (34.5)
5 (17.9) 10 (35.7) 9 (32.1)
28.5 6 1.2
28.4 6 1.7
Note: Values for categorical variables are given as number (percentage); values for continuous variables, as mean 6 standard deviation or mean 6 standard deviation (range). Abbreviations: CHF, congestive heart failure; MMSE, MiniMental State Examination; SIADH, syndrome of inappropriate antidiuretic hormone secretion.
and clinical tolerance of the trial drug. Approximately 55% of tolvaptan-treated patients had doses titrated to the highest permitted dose (60 mg), whereas 89% of placebo patients had doses titrated to the highest sham dose (60 mg). Pharmacodynamic results (serum tolvaptan and metabolite levels) indicated that all patients were dosed appropriately and appeared to adhere to dosing. Efficacy Analysis Mean baseline z scores were negative compared with normative values from age-matched individuals, representing a more than 1–standard deviation (SD) deficit in cognitive speed, and were comparable between the tolvaptan and placebo groups (21.37 6 1.20 vs 21.55 6 1.26, respectively). Although mean changes from baseline in neurocognitive composite scores of speed domains at day 22 indicated a trend for greater improvement with tolvaptan; that is, a modest effect size for tolvaptan (0.39 6 0.49 [SD]) versus a smaller effect size for placebo (0.20 6 0.61), the difference was not statistically significant (estimated treatment effect, 0.23; 95% CI, 20.03 to 0.50; P 5 0.08; Fig 4). At follow-up on day 28, the neurocognitive composite score of speed domains decreased to near baseline values in the tolvaptan group. Effects were similar between the overall patient population and patients with more severe hyponatremia (serum sodium concentration , 130 mEq/L). The effect had contributions from each end point component but was Am J Kidney Dis. 2016;67(6):893-901
Impact of Subclinical Hyponatremia
Figure 3. Primary end point cognition score domains (mean 6 standard deviation) at baseline.
largely driven by psychomotor speed (tapping), which was the only neurocognitive measure among the speed domains to show a statistically significance difference between tolvaptan and placebo treatment (P 5 0.02; Fig S2A). Overall neurocognitive composite scores (Fig S3A) tended to favor tolvaptan, but this was not statistically significant (P 5 0.2); results of individual neurocognitive tests making up the neurocognitive composite score domains are shown in Fig S3B. Results for measures of gait and stability were mixed. Improvements were seen on the Romberg test (P 5 0.2). Postural stability and Timed Up and Go Test scores were divergent, with the former favoring placebo and the latter favoring tolvaptan (Fig S2B). Exploratory analyses showed that correlations between changes in serum sodium concentrations and
neurocognitive test results were positive in nearly all measures, demonstrating a trend for improvement with improved serum sodium concentration. Correlations were most pronounced (Pearson coefficient, 0.3) and statistically significant (P 5 0.03) for the psychomotor speed domain (Table 2). In the tolvaptan group, nonsignificant greater increases in serum osteocalcin levels (Fig S4A) and decreases in urine NTx-creatinine ratios (Fig S4B) were seen, as well as significant decreases from baseline in calculated bone resorption indexes at day 22 or end of treatment (Fig S4C). Safety Evaluation The most common AEs ($10%) are listed in Table S3. Few patients had serious AEs. The titration
Figure 4. Treatment effects (95% confidence interval [CI]) on the primary cognitive end point and the 3 speed domain components in all patients (serum sodium concentration , 135 mEq/L) and in those with a baseline serum sodium concentration , 130 mEq/L. Am J Kidney Dis. 2016;67(6):893-901
897
898
0.3 0.03 0.5 0.14 (20.14 to 0.40) 0.31 (0.04 to 0.54) 0.11 (20.17 to 0.37)
algorithm in this trial was aimed at achieving a normal serum sodium concentration without overly rapid correction, and no overly rapid correction was seen in either treatment group. Although no AEs related to liver abnormalities and no potentially clinically significant abnormalities in aminotransaminase levels were reported in this relatively short-term treatment study, monthly transaminase monitoring is now recommended based on an increased incidence of liver injury biomarker levels during 18 months of highdose tolvaptan treatment in a patient population treated for polycystic kidney disease.20
DISCUSSION
Note: N 5 51 for all rows. Unless otherwise indicated, values given as mean 6 standard deviation. Abbreviation: CI, confidence interval. a Probability of correlation coefficient.
21.21 6 1.63 22.86 6 0.77 21.44 6 1.33 20.90 6 1.08 22.73 6 0.85 21.37 6 1.75
20.93 6 0.92 22.57 6 0.99 20.86 6 1.34
21.27 6 1.67 22.86 6 0.83 21.04 6 1.42
20.04 6 0.59 0.16 6 0.46 0.51 6 0.86
20.06 6 0.82 20.01 6 0.32 0.40 6 0.81
0.07 (20.31 to 0.45) 0.17 (20.06 to 0.40) 0.13 (20.26 to 0.53)
0.2 0.2 0.18 (20.11 to 0.43) 0.19 (20.09 to 0.44) 0.11 (20.18 to 0.40) 0.04 (20.15 to 0.22) 21.55 6 1.26 20.84 6 0.85
Primary cognitive end point Overall composite cognitive end point Reaction time domain Psychomotor speed domain Processing speed domain
21.37 6 1.20 20.81 6 0.71
21.14 6 0.96 20.64 6 0.64
21.40 6 1.30 20.70 6 0.79
0.22 6 0.57 0.17 6 0.36
0.15 6 0.58 0.14 6 0.41
Pearson Correlation Coefficient (95% CI) Estimated Treatment Effect (95% CI) Placebo Tolvaptan Placebo
Tolvaptan
Placebo
Change From Baseline Follow-up Baseline
Tolvaptan Neurocognitive Composite z Score
Table 2. Correlations Between Changes From Baseline in Serum Sodium Concentration and Neurocognitive Composite z Score End Points
Pa
Verbalis et al
This trial demonstrated the effects of oral tolvaptan on correction of serum sodium concentrations to the normal range in adult patients with mild hyponatremia and the effects of this correction on measures of cognition, gait, and bone metabolism. Compared with normative values from age-matched individuals, study participants had impaired cognitive performance at baseline on measures of correct choice and speed of choice, impaired postural and gait stability, and reduced SF-12 physical component scores, especially for the Physical Functioning, Role– Physical, and Bodily Pain domains. Participants showed preserved cognitive performance on measures relying on word and number recognition and preserved SF-12 mental component scores, particularly the Mental Health and Social Functioning domains. It is important to note that some imbalances between treatment groups in baseline neurocognitive function scores were noted in this population. Further, some baseline test results were close to normal. For example, a ceiling effect was observed for tolvaptantreated patients with normal SF-12 Mental Component Summary scores at baseline, who therefore had no opportunity for improvement. Major limitations of this study include the small sample size and short treatment duration. Because this was a pilot study, a formal sample size power calculation for the primary end point was not performed. However, previous unblinded crossover trials have shown significant neurocognitive improvements with as few as 15 patients.15-19 The present trial studied a different population using related, but distinct, instruments. Tests that measure reaction time, speed, and processing speed were chosen for the primary end point based on previous published findings3; however, the trial lacked sufficient power to detect changes in neurocognitive test and SF-12 scores. An additional problem related to the small sample size was the inability to control for underlying comorbid conditions or medications. However, because the study was designed to evaluate hyponatremic patients with Am J Kidney Dis. 2016;67(6):893-901
Impact of Subclinical Hyponatremia
chronic asymptomatic hyponatremia in good general health, significant changes in underlying comorbid conditions or medications were not likely during the short period of study. Although the 21-day study duration is short, this period was sufficient to observe both neurocognitive and gait stability improvements in previous studies of hyponatremic patients3 and served to reduce the likelihood of interval changes in underlying comorbid conditions or medications to affect outcome measures. Tolvaptan was effective in reversing hyponatremia in this small population of patients; however, a goal of serum sodium concentration $ 138 mEq/L was reached in only 50% of patients treated with tolvaptan. This was likely secondary to increased thirst at higher serum sodium concentrations, resulting in equilibration at a new set point at which fluid intake balanced the tolvaptan-induced aquaresis. Although the difference between groups did not reach statistical significance, a trend for improvement in the neurocognitive composite score of speed domains, the primary study end point, was observed in the tolvaptan group. Changes from baseline in cognition test results pertaining to speed domains were uniformly numerically in favor of tolvaptan (small effect sizes . 0.2), with most missing statistical significance but showing positive correlations with changes in serum sodium concentrations. However, the psychomotor speed domain both reached statistical significance for tolvaptan versus placebo (small and moderate effect sizes of 0.27 for sodium , 135 mEq/L and 0.53 for sodium , 130 mEq/L, respectively) and correlated well with increased serum sodium concentrations. Compared to the placebo group, a tendency to regress to baseline was observed for each of the end points on withdrawal of tolvaptan treatment (ie, at the end of 7-day follow-up), which again suggests that the effect was related to the correction of serum sodium concentration. Effect sizes of 0.27 and 0.53 for improvement in the psychomotor speed domain in tolvaptan-treated patients are clinically relevant, as defined by Cohen D criteria (small effect size 5 0.2 SD, moderate effect size 5 0.5 SD, and large effect size 5 0.8 SD).21 The correlation of effect with serum sodium concentration and the tendency to regress on withdrawal of tolvaptan treatment support the hypothesis that the observed small effect of tolvaptan (and subsequent correction of serum sodium concentration) on each neurocognition test or domain is nonrandom.22 An exception was noted with the negative change in neurocognitive postural stability in patients treated with tolvaptan, which further declined through the 7day follow-up, suggesting a possible random effect for this measure. Baseline imbalances between the 2 treatment groups may have contributed to this result. Am J Kidney Dis. 2016;67(6):893-901
Similar effect sizes for improvements in the range of 0.3 to 0.6 using the Cognitive Drug Research system have been previously documented to demonstrate clinically meaningful improvements in various parameters of attention (eg, rivastigmine for Parkinson disease15 and Lewy body dementia,16 NS2359 for attention-deficit/hyperactivity disorder,17 and modafinil for narcolepsy19) and improved processing speed (eg, reboxetine for major depressive disorder18 and modafinil for narcolepsy19). The bone markers osteocalcin concentration and NTx-creatinine ratio were exploratory end points based on the reported impact of hyponatremia on fractures and osteoporosis.4,6-8,23-25 Results were favorable for tolvaptan treatment because osteocalcin concentration (a marker of bone formation) increased and NTx-creatinine ratio (a bone marker of resorption) decreased to a greater extent than with placebo. Although changes in individual markers of bone formation and resorption were not statistically significant, calculation of a bone resorption index, defined as change from baseline in urine NTx-creatinine ratio divided by serum osteocalcin concentration, showed a highly significant decrease in the tolvaptan group compared to the placebo group despite a small sample size of patients with both bone markers measured (tolvaptan, n 5 15; placebo, n 5 18). Markers of bone metabolism are used clinically to detect changes in bone resorption and formation in response to therapeutic interventions in patients with osteopenia and osteoporosis,26,27 and the concept of a bone resorptive index (or “uncoupling factor”) that arithmetically combines both resorption and formation markers has been proposed as a potentially more sensitive measure of bone metabolism.28 The striking improvement in resorptive index after just 22 days of hyponatremia correction with tolvaptan suggests that long-term treatment of hyponatremia may result in an improvement in bone mineral density, which was recently reported in an individual patient.29 The improvement in the psychomotor speed domain also has important potential clinical relevance. Patients with mild chronic hyponatremia have been found to have gait instability and an increased rate of falls.3 This suggests impairment of the fine motor control that is necessary to maintain upright stability during walking, which is supported by the significant decrease in tapping speed measured by the psychomotor speed domain in the present trial. Although the cause of this deficit is speculative, it is notable that the chronically hyponatremic brain undergoes a volume regulatory decrease in cell volume through solute extrusion from brain cells.30 Among the solutes lost are low-molecular-weight organic osmolytes, including some amino acids that serve as excitatory neurotransmitters.31 These include glutamate, the 899
Verbalis et al
major excitatory neurotransmitter in motor circuits, which can be decreased by as much as 40% in rats with chronic hyponatremia.32 It is reasonable to speculate that such profound depletion of transmitters that are vitally important for control of motor function could slow nerve conduction and thereby impair the rapid motor corrections necessary to maintain a stable gait or enable rapid movements, such as finger tapping. The combined effects of gait instability, increased falls, and osteoporosis could lead to the increased fracture rates observed in multiple independent studies of patients with hyponatremia.4,23-25 Because the gait instability appears to be reversible with correction of hyponatremia,3 and the bone resorption index observed in the present trial suggests that the bone loss associated with hyponatremia may be reversible as well, there is a strong rationale to develop clinical trials to assess the effect of hyponatremia correction on falls, fractures, and osteoporosis. Tolvaptan was generally safe and well tolerated in this trial, and no new safety concerns were raised. Commonly reported ($10%) treatment-emergent AEs were consistent with the aquaretic action of tolvaptan, and few serious AEs were reported. No apparent trends were observed in response to treatment with tolvaptan regarding vital signs, ECGs, or clinical laboratory parameters. No incidence of overly rapid correction of serum sodium concentration (.12 mEq/L in a 24-hour period) was observed. In conclusion, this trial demonstrated that in a small adult population with hyponatremia without overt neurologic or mental status changes at baseline, tolvaptan was effective in reversing hyponatremia, which correlated with improvements in tests of neurocognition, particularly rapid motor movements. The results also suggest a potential reversal of hyponatremia-induced resorptive bone loss, as reflected by the significant improvement in bone resorption index following short-term correction of hyponatremia. Although tolvaptan was used to correct hyponatremia in this study, improvements in neurocognitive deficits correlated with improvements in serum sodium concentrations and would be expected regardless of the methods used to improve serum sodium concentrations. Future studies in larger groups of hyponatremic patients are needed to determine what type, degree, and length of therapy are necessary to produce and maintain neurocognitive and metabolic benefits.
Research Center; Ali Khojasteh, Columbia Comprehensive Care Clinic; John LaFata, Progressive Clinical Research; Michael J. Lillestol, Lillestol Research LLC; Sarah Olelewe, Hawthorne, CA; David Rosenbaum, Pikes Peak Cardiology; Mitchell H. Rosner, University of Virginia Health System West Hospital Complex Nephrology Clinical Research Center; Miguel Trevino, Innovative Research of West FL; Joseph Verbalis, Georgetown University; and Wayne Wells, Lebanon, TN. Support: Otsuka Pharmaceutical Development & Commercialization Inc provided the financial support for this study. Otsuka designed the study, collected and analyzed the data, and played a role in its interpretation. Otsuka assisted in reviewing the report and was involved in the decision to submit the results of this study for publication. BioScience Communications (New York, NY) provided assistance with formatting the manuscript for submission that was funded by Otsuka. Dr Verbalis has received research support and non2continuing medical education2related fees from Otsuka; Dr Ellison has received research support from Otsuka; Drs Hobart, Ouyang, and Czerwiec and Ms Krasa are employees of Otsuka. Financial Disclosure: Dr Verbalis has received non2 continuing medical education2related fees from Cornerstone Therapeutics and Ferring Pharmaceuticals. Contributions: Study design: JGV, MH, HK, FSC; data acquisition: JGV, HE; data analysis/interpretation: JGV, MH, HK, JO, FSC; statistical analysis: JGV, MH, HK, JO, FSC. All prespecified primary and secondary outcomes of the study were reported in this study. Each author contributed important intellectual content during manuscript drafting or revision and accepts accountability for the overall work by ensuring that questions pertaining to the accuracy or integrity of any portion of the work are appropriately investigated and resolved. JGV takes responsibility that this study has been reported honestly, accurately, and transparently; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and registered) have been explained. Peer Review: Evaluated by 3 external peer reviewers, a Statistical Editor, a Co-Editor, and the Editor-in-Chief.
ACKNOWLEDGEMENTS
1. Upadhyay A, Jaber BL, Madias NE. Epidemiology of hyponatremia. Semin Nephrol. 2009;29(3):227-238. 2. Hawkins RC. Age and gender as risk factors for hyponatremia and hypernatremia. Clin Chim Acta. 2003;337(1-2):169-172. 3. Renneboog B, Musch W, Vandemergel X, Manto MU, Decaux G. Mild chronic hyponatremia is associated with falls, unsteadiness, and attention deficits. Am J Med. 2006;119(1): 71.e1-71.e8.
The INSIGHT Investigators are Otis Barnum, Natchitoches Hospital; Karthikeya Devireddy, KD Medical Group Inc; Howard Ellison, Rockdale Medical Research Associates; Miguel Franco, Memorial Clinical Associates; Francis Goldstein, Carolina Research Associates; Terrence C. Hack, Primary Care Cardiology Research Inc; Kianoosh Kaveh, Coastal Nephrology Associates 900
SUPPLEMENTARY MATERIAL Table S1: Entry criteria. Table S2: Baseline SF-12, mental and physical component scores. Table S3: Adverse events. Figure S1: Mean serum sodium concentrations at study visits. Figure S2: Treatment effects on overall NCS and all 7 components, and on gait and stability measures. Figure S3: Diagram of neurocognitive tests for each NCS domain and treatment effects. Figure S4: Effects on markers of bone turnover. Item S1: Neurocognitive battery methods, gait and stability, and statistical methods for secondary and exploratory end points. Note: The supplementary material accompanying this article (http://dx.doi.org/10.1053/j.ajkd.2015.12.024) is available at www.ajkd.org
REFERENCES
Am J Kidney Dis. 2016;67(6):893-901
Impact of Subclinical Hyponatremia 4. Kinsella S, Moran S, Sullivan MO, Molloy MG, Eustace JA. Hyponatremia independent of osteoporosis is associated with fracture occurrence. Clin J Am Soc Nephrol. 2010;5(2):275-280. 5. Hoorn EJ, Rivadeneira F, van Meurs JB, et al. Mild hyponatremia as a risk factor for fractures: the Rotterdam Study. J Bone Miner Res. 2011;26(8):1822-1828. 6. Verbalis JG, Barsony J, Sugimura Y, et al. Hyponatremiainduced osteoporosis. J Bone Miner Res. 2010;25(3):554-563. 7. Jamal SA, Arampatzis S, Harrison SL, et al. Hyponatremia and fractures: findings from the Osteoporotic Fractures in Men Study. J Bone Miner Res. 2015;30(6):970-975. 8. Kruse C, Eiken P, Vestergaard P. Hyponatremia and osteoporosis: insights from the Danish National Patient Registry. Osteoporos Int. 2015;26(3):1005-1016. 9. Usala RL, Fernandez SJ, Mete M, et al. Hyponatremia is associated with increased osteoporosis and bone fractures in a large US health system population. J Clin Endocrinol Metab. 2015;100(8):3021-3031. 10. Greenberg A, Verbalis JG, Amin AN, et al. Current treatment practice and outcomes. Report of the hyponatremia registry. Kidney Int. 2015;88(1):167-177. 11. Schrier RW, Gross P, Gheorghiade M, et al. Tolvaptan, a selective oral vasopressin V2-receptor antagonist, for hyponatremia. N Engl J Med. 2006;355(20):2099-2112. 12. Berl T, Quittnat-Pelletier F, Verbalis JG, et al. Oral tolvaptan is safe and effective in chronic hyponatremia. J Am Soc Nephrol. 2010;21(4):705-712. 13. Simpson PM, Surmon DJ, Wesnes KA, Wilcock GK. The Cognitive Drug Research computerised assessment system for demented patients: a validation study. Int J Geriatr Psychiatry. 1991;6:95-102. 14. Wesnes K, Pincock C. Practice effects on cognitive tasks: a major problem? Lancet Neurol. 2002;1(8):473. 15. Wesnes KA, McKeith IG, Edgar C, et al. Benefits of rivastigmine on attention in dementia associated with Parkinson disease. Neurology. 2005;65(10):1654-1656. 16. Wesnes KA, McKeith IG, Ferrara R, et al. Effects of rivastigmine on cognitive function in dementia with Lewy bodies: a randomised placebo-controlled international study using the Cognitive Drug Research computerised assessment system. Dement Geriatr Cogn Disord. 2002;13(3):183-192. 17. Wilens TE, Klint T, Adler L, et al. A randomized controlled trial of a novel mixed monoamine reuptake inhibitor in adults with ADHD. Behav Brain Funct. 2008;4:24. 18. Ferguson JM, Wesnes KA, Schwartz GE. Reboxetine versus paroxetine versus placebo: effects on cognitive functioning
Am J Kidney Dis. 2016;67(6):893-901
in depressed patients. Int Clin Psychopharmacol. 2003;18(1): 9-14. 19. Harsh JR, Hayduk R, Rosenberg R, et al. The efficacy and safety of armodafinil as treatment for adults with excessive sleepiness associated with narcolepsy. Curr Med Res Opin. 2006;22(4): 761-774. 20. Torres VE, Chapman AB, Devuyst O, Gansevoort RT, Grantham JJ, Higashihara E, Perrone RD, Krasa HB, Ouyang J, Czerwiec FS; TEMPO 3:4 Trial Investigators. Tolvaptan in patients with autosomal dominant polycystic kidney disease. N Engl J Med. 2012;367(25):2407-2418. 21. Cohen J. Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Hillsdale, NJ: Lawrence Erlbaum; 2015. 22. Verbalis JG, Ouyang J, Orlandi C, Czerwiec FS. Serum sodium correction with tolvaptan improves self-reported mental health status in hyponatremic patients: a subanalysis from the SALT trials [abstract]. Presented at: The Endocrine Society’s 89th Annual Meeting; June 2-5, 2007; Toronto, Canada. 23. Sandhu HS, Gilles E, DeVita MV, Panagopoulos G, Michelis MF. Hyponatremia associated with large-bone fracture in elderly patients. Int Urol Nephrol. 2009;41(3):733-737. 24. Gankam KF, Andres C, Sattar L, Melot C, Decaux G. Mild hyponatremia and risk of fracture in the ambulatory elderly. QJM. 2008;101(7):583-588. 25. Hoorn EJ, Liamis G, Zietse R, Zillikens MC. Hyponatremia and bone: an emerging relationship. Nat Rev Endocrinol. 2011;8(1):33-39. 26. Naylor K, Eastell R. Bone turnover markers: use in osteoporosis. Nat Rev Rheumatol. 2012;8(7):379-389. 27. Eastell R, Hannon RA. Biomarkers of bone health and osteoporosis risk. Proc Nutr Soc. 2008;67(2):157-162. 28. Eastell R, Robins SP, Colwell T, Assiri AM, Riggs BL, Russell RG. Evaluation of bone turnover in type I osteoporosis using biochemical markers specific for both bone formation and bone resorption. Osteoporos Int. 1993;3(5):255-260. 29. Sejling AS, Thorsteinsson AL, Pedersen-Bjergaard U, Eiken P. Recovery from SIADH-associated osteoporosis: a case report. J Clin Endocrinol Metab. 2014;99(10):3527-3530. 30. Verbalis JG. Brain volume regulation in response to changes in osmolality. Neuroscience. 2010;168(4):862-870. 31. Gullans SR, Verbalis JG. Control of brain volume during hyperosmolar and hypoosmolar conditions. Annu Rev Med. 1993;44:289-301. 32. Verbalis JG, Gullans SR. Hyponatremia causes large sustained reductions in brain content of multiple organic osmolytes in rats. Brain Res. 1991;567:274-282.
901
H
yponatremia is associated with a variety of underlying diseases and conditions, and the incidence of chronic mild hyponatremia increases with age.1,2 Chronic hyponatremia is often labeled as “asymptomatic,” and more subtle neurologic signs and symptoms of chronic hyponatremia associated with impaired mental functioning may be unrecognized or ascribed to “normal aging.” Subclinical impairment of mental functioning can be detected in patients with grossly normal results from neurologic tests and MiniMental State Examinations, leading to attention deficits similar to impairments seen after the ingestion of
moderate amounts of alcohol.3 Patients with mild chronic hyponatremia can manifest gait instability, and chronic hyponatremia has been associated with increased falls and fractures.3-5 Studies in experimental animals have indicated that hyponatremia also causes bone resorption, and analysis of epidemiologic data has shown a significantly increased odds ratio for osteoporosis in patients with hyponatremia,6 which was confirmed in 3 subsequent studies.7-9 Current options for treating hyponatremia are not optimal. This is particularly the case in the outpatient setting, in which fluid restriction is poorly effective and
From the 1Division of Endocrinology and Metabolism, Georgetown University, Washington, DC; 2Rockdale Medical Research Associates, Conyers, GA; and 3Otsuka Pharmaceutical Development and Commercialization, Inc, Rockville, MD. * A list of the INSIGHT Investigators appears in the Acknowledgements. Received July 1, 2015. Accepted in revised form December 28, 2015. Originally published online February 10, 2016.
Trial registration: www.ClinicalTrials.gov; study number: NCT00550459. Address correspondence to Joseph G. Verbalis, MD, Division of Endocrinology and Metabolism, 232 Bldg D, 4000 Reservoir Rd NW, Washington, DC 20007. E-mail: [email protected] 2016 by the National Kidney Foundation, Inc. 0272-6386 http://dx.doi.org/10.1053/j.ajkd.2015.12.024
Am J Kidney Dis. 2016;67(6):893-901
893
Verbalis et al
patients frequently do not adhere to fluid intake recommendations.10 Tolvaptan is a vasopressin receptor antagonist that induces water diuresis (aquaresis) without depletion of electrolytes. In 2 large randomized trials, tolvaptan was shown to increase serum sodium concentrations, in many cases to normalization.11 An open-label extension study showed continued efficacy without significant toxicity for treatment durations as long as 4 years.12 The major objective of the present trial was to determine the effect of oral tolvaptan on measures of cognition, gait, and postural stability in adult patients with mild to moderate hyponatremia and absence of overt neurologic symptoms. Exploratory objectives included evaluation of the effect of oral tolvaptan on self-reported cognitive variables and markers of bone metabolism.
METHODS Study Design The Investigation of the Neurocognitive Impact of Sodium Improvement in Geriatric Hyponatremia: Efficacy and Safety of Tolvaptan (INSIGHT) was a multicenter, randomized, doubleblind, placebo-controlled, parallel-group, titration-to-effect trial (Table S1, available as online supplementary material). The trial consisted of a 30-day screening period, 21-day treatment period, and 7-day posttreatment period (Fig 1A). During the screening period, attempts were made to identify and eliminate the underlying cause of hyponatremia, and patients were then treated using standard hyponatremia therapies. Patients who remained hyponatremic were randomly assigned to tolvaptan or placebo. The primary objective of this trial was to determine the effect of tolvaptan on change from baseline in a single composite score
comprising the sum of the z scores of individual cognitive variables under the domains of reaction time, psychomotor speed, and processing speed (hereafter referred to as the neurocognitive composite score of speed domains; Fig 1B) in patients with chronic mild to moderate hyponatremia after correction of serum sodium concentrations to normal ranges. Secondary and exploratory measures are described under Outcome Measures. The trial was conducted under the investigational new drug exemption of the US Food and Drug Administration in compliance with Good Clinical Practice guidelines, the sponsor’s standard operating procedures, and ethics principles for the protection of human research subjects in the Declaration of Helsinki. The protocol, amendments, and informed consent form were reviewed and approved by the governing institutional review board (IRB) or ethics committee of each investigational center prior to trial start at that center (central IRB: Chesapeake Research Review, CRRI 0607006; local IRB: Georgetown IRB, 2007-538; local IRB: University of Virginia IRB, 13606). Patients (or their guardians or legal representatives) provided written informed consent before any trial-related procedures were performed.
Patients Men and women 50 years or older with chronic asymptomatic hyponatremia (serum sodium concentration .122-,135 mEq/L) in the euvolemic or hypervolemic state were eligible for the trial. Two patients with serum sodium concentrations of 121 and 122 mEq/L were included following review. At least 50% of patients were targeted to have serum sodium concentrations # 130 mEq/L at baseline. Detailed inclusion and exclusion criteria are provided in Table S1.
Study Treatment Patients were randomly assigned 1:1 with stratification by baseline serum sodium concentration (,130 or $130 mEq/L) to receive oral tolvaptan or matching placebo tablets beginning at a dose of 15 mg/d, with titration up to 30 or 60 mg/d. A forced
Figure 1. (A) Study schematic and (B) neurocognitive composite score of speed domains (NCSSD). *Daily titration based on serum sodium levels. yThere were 26 completers per group analyzed for efficacy. 894
Am J Kidney Dis. 2016;67(6):893-901
Impact of Subclinical Hyponatremia titration in both treatment arms to 60 mg/d by days 3 to 7 was based on the patient’s change in serum sodium concentration and clinical tolerance. Titration occurred daily until the patient’s serum sodium concentration was increased and maintained at $138 mEq/L or the patient could not tolerate further titration. During this period, fluid restrictions were loosened or suspended until the patient’s response to therapy could be evaluated, typically during the first few days of therapy. Fluid restriction was to be reinstituted at any time during the trial in patients whose serum sodium concentrations did not improve or worsened. Patients remained on tolvaptan treatment for up to 21 days. Individual patient participation was approximately 60 days (30-day screening period, 21-day treatment period, and 7-day follow-up).
Outcome Measures The primary efficacy end point was change from baseline to day 22 in neurocognitive composite score of speed domains. Secondary efficacy end points were the effect of tolvaptan treatment on: (1) changes from baseline in individual neurocognitive domain scores included in the neurocognitive composite score of speed domains (reaction time, psychomotor speed, and processing speed); (2) change from baseline in overall neurocognitive composite score (composite score of the following domains: reaction time, psychomotor speed, processing speed, continuity of attention, working memory/executive functions, quality of episodic verbal memory, and postural stability); (3) changes from baseline in postural stability, Timed Up and Go test, and Romberg test; (4) change from baseline in serum sodium concentration; and (5) correlations of changes from baseline in neurocognitive domain scores with changes in serum sodium concentrations. Exploratory efficacy end points included changes from baseline in cognitive
variables, Hyponatremia Disease-Specific Survey, and 12-Item Short Form Health Survey (SF-12) component scores and changes from baseline in serum osteocalcin concentration and urine N-telopeptide (NTx)-creatinine ratio. A bone resorption index was calculated for each patient as change from baseline in NTx-creatinine ratio divided by serum osteocalcin concentration. Secondary safety analyses were conducted to determine the effect of tolvaptan treatment on safety as determined by adverse events (AEs), vital signs, electrocardiograms (ECGs), and clinical laboratory tests.
Efficacy Measurements Efficacy assessments were conducted at baseline/day 1, testing day (day 22)/early termination, and the follow-up visit (day 28). Specimens for serum chemistry and urine pregnancy tests (if applicable) were collected and processed according to each trial center’s routine procedures. All specimens were sent to a central laboratory for analysis. Neurocognitive tests were administered using the Cognitive Drug Research system.13 Training of research personnel on the Cognitive Drug Research computerized assessment system was done prior to the first day of dosing of the trial to ensure an optimal level of performance for the baseline assessment. The purpose of training was to help overcome initial test anxiety, familiarize patients with the procedures, enable the development of strategies for task performance, and overcome any initial practice effects.14 Each patient completed 2 training sessions prior to the start of the trial. Training sessions were to be no more than 14 days apart to reduce practice effects. (See Neurocognitive Battery Methods in Item S1 for detailed methods relating to neurocognitive and postural stability testing.)
ENROLLMENT
Assessed for eligibility (n=117) Excluded (n=61) Not meeting inclusion criteria (n=40) Declined to participate (n=21)
Am J Kidney Dis. 2016;67(6):893-901
Placebo
ALLOCATED
Allocated to intervention (n=29) Received allocated intervention (n=29) Did not receive allocated intervention (n=0)
Allocated to intervention (n=27) Received allocated intervention (n=27) Did not receive allocated intervention (n=0)
FOLLOW-UP
Figure 2. Patient enrollment and outcomes.
Tolvaptan
Lost to follow-up (n=0) Discontinued intervention (n=3) Reason(s): – Adverse event (n=2) – Withdrew consent (n=1)
Lost to follow-up (n=0) Discontinued intervention (n=1) Reason(s): – Withdrew consent (n=1)
ANALYSIS
Randomized (n=56)
Analyzed (n=26) Excluded from analysis (n=3) Reason(s): – Missing baseline or postbaseline observation (n=3)
Analyzed (n=26) Excluded from analysis (n=1) Reason(s): – Missing baseline or postbaseline observation (n=1)
895
Verbalis et al Table 1. Patient Characteristics and Demographics
Statistical Methods Based on available data for serum sodium concentration improvements with tolvaptan, it was determined that about 30 patients per arm would be needed to see significant treatment effect differences in serum sodium concentrations. The ability to detect neurocognitive end point differences with similar patient numbers has been reported in neurocognitive trials involving other agents and disorders.15-19 Analysis of covariance with factors of treatment, disease severity (,130 or $130 mEq/L at baseline), age, severity-age interaction, and covariate baseline for individual measurements was used to fit the primary end point using the intention-to-treat data set. The estimated treatment effect and its 95% confidence interval (CI) were provided under the model, along with the P value. A 2-sided a 5 0.05 was applied to the primary analysis. The primary analysis was based on observed cases. (See Item S1 for detailed methods of analysis of the secondary and exploratory efficacy end points.) Safety analysis was conducted based on the safety data set, which was defined as all patients who received 1 or more dose of trial medication. Safety variables analyzed included AEs, clinical laboratory test results, vital signs, and ECGs. In general, summary statistics of changes from baseline were provided for safety variables (when applicable) based on all available data. Descriptive statistics were used to describe differences in AEs, vital signs, ECGs, and clinical laboratory findings between the tolvaptan and placebo groups.
RESULTS Patients In all, 57 men and women were enrolled at 16 US sites in September 2007 to March 2009; a total of 53 patients were analyzed for efficacy (26 tolvaptan and 27 placebo), and 57, for safety (29 tolvaptan and 28 placebo; Fig 2). The 2 patient groups were relatively well matched for relevant baseline characteristics (Table 1). Compared with normative values from agematched individuals, patients in this trial at baseline had significantly impaired cognitive performance on measures of correct choice and speed of choice, modestly impaired gait and postural stability, and elevated bone turnover marker levels, but preserved cognitive performance on measures relying on word and number recognition (Fig 3). However, mean SF12 scores were generally higher than those of patients enrolled in previous trials of tolvaptan (SALT-1 and SALT-2 [Study of Ascending Levels of Tolvaptan in Hyponatremia 1 and 2]) for both the physical and mental component summaries (Table S2).11 Serum Sodium Concentrations Mean serum sodium concentrations increased from 129 to 136 mEq/L in the tolvaptan group and from 130 to 132 mEq/L in the placebo group (P , 0.01 in treatment comparison); levels in the tolvaptan group returned to baseline after withdrawal (Fig S1). Dose titration to the 30- and 60-mg dose levels (by days 3-7) was based on the patient’s change in serum sodium concentration (to $138 mEq/L [$138 mmol/L]) 896
Male sex Race White Non-Hispanic Age, y Baseline serum sodium, mEq/L Cause SIADH/other Cirrhosis CHF Baseline medication Diabetes Psycholeptic Psychoanaleptic MMSE score
Tolvaptan 15-60 mg (n 5 29)
Placebo (n 5 28)
14 (48)
9 (32)
29 (100) 29 (100)
26 (93) 25 (89)
71.1 6 10.2 (50-86) 129.38 6 3.31
71.3 6 9.7 (50-87) 129.81 6 3.04
29 (100) 0 (0) 0 (0)
26 (93) 2 (7) 0 (0)
1 (3.4) 8 (27.6) 10 (34.5)
5 (17.9) 10 (35.7) 9 (32.1)
28.5 6 1.2
28.4 6 1.7
Note: Values for categorical variables are given as number (percentage); values for continuous variables, as mean 6 standard deviation or mean 6 standard deviation (range). Abbreviations: CHF, congestive heart failure; MMSE, MiniMental State Examination; SIADH, syndrome of inappropriate antidiuretic hormone secretion.
and clinical tolerance of the trial drug. Approximately 55% of tolvaptan-treated patients had doses titrated to the highest permitted dose (60 mg), whereas 89% of placebo patients had doses titrated to the highest sham dose (60 mg). Pharmacodynamic results (serum tolvaptan and metabolite levels) indicated that all patients were dosed appropriately and appeared to adhere to dosing. Efficacy Analysis Mean baseline z scores were negative compared with normative values from age-matched individuals, representing a more than 1–standard deviation (SD) deficit in cognitive speed, and were comparable between the tolvaptan and placebo groups (21.37 6 1.20 vs 21.55 6 1.26, respectively). Although mean changes from baseline in neurocognitive composite scores of speed domains at day 22 indicated a trend for greater improvement with tolvaptan; that is, a modest effect size for tolvaptan (0.39 6 0.49 [SD]) versus a smaller effect size for placebo (0.20 6 0.61), the difference was not statistically significant (estimated treatment effect, 0.23; 95% CI, 20.03 to 0.50; P 5 0.08; Fig 4). At follow-up on day 28, the neurocognitive composite score of speed domains decreased to near baseline values in the tolvaptan group. Effects were similar between the overall patient population and patients with more severe hyponatremia (serum sodium concentration , 130 mEq/L). The effect had contributions from each end point component but was Am J Kidney Dis. 2016;67(6):893-901
Impact of Subclinical Hyponatremia
Figure 3. Primary end point cognition score domains (mean 6 standard deviation) at baseline.
largely driven by psychomotor speed (tapping), which was the only neurocognitive measure among the speed domains to show a statistically significance difference between tolvaptan and placebo treatment (P 5 0.02; Fig S2A). Overall neurocognitive composite scores (Fig S3A) tended to favor tolvaptan, but this was not statistically significant (P 5 0.2); results of individual neurocognitive tests making up the neurocognitive composite score domains are shown in Fig S3B. Results for measures of gait and stability were mixed. Improvements were seen on the Romberg test (P 5 0.2). Postural stability and Timed Up and Go Test scores were divergent, with the former favoring placebo and the latter favoring tolvaptan (Fig S2B). Exploratory analyses showed that correlations between changes in serum sodium concentrations and
neurocognitive test results were positive in nearly all measures, demonstrating a trend for improvement with improved serum sodium concentration. Correlations were most pronounced (Pearson coefficient, 0.3) and statistically significant (P 5 0.03) for the psychomotor speed domain (Table 2). In the tolvaptan group, nonsignificant greater increases in serum osteocalcin levels (Fig S4A) and decreases in urine NTx-creatinine ratios (Fig S4B) were seen, as well as significant decreases from baseline in calculated bone resorption indexes at day 22 or end of treatment (Fig S4C). Safety Evaluation The most common AEs ($10%) are listed in Table S3. Few patients had serious AEs. The titration
Figure 4. Treatment effects (95% confidence interval [CI]) on the primary cognitive end point and the 3 speed domain components in all patients (serum sodium concentration , 135 mEq/L) and in those with a baseline serum sodium concentration , 130 mEq/L. Am J Kidney Dis. 2016;67(6):893-901
897
898
0.3 0.03 0.5 0.14 (20.14 to 0.40) 0.31 (0.04 to 0.54) 0.11 (20.17 to 0.37)
algorithm in this trial was aimed at achieving a normal serum sodium concentration without overly rapid correction, and no overly rapid correction was seen in either treatment group. Although no AEs related to liver abnormalities and no potentially clinically significant abnormalities in aminotransaminase levels were reported in this relatively short-term treatment study, monthly transaminase monitoring is now recommended based on an increased incidence of liver injury biomarker levels during 18 months of highdose tolvaptan treatment in a patient population treated for polycystic kidney disease.20
DISCUSSION
Note: N 5 51 for all rows. Unless otherwise indicated, values given as mean 6 standard deviation. Abbreviation: CI, confidence interval. a Probability of correlation coefficient.
21.21 6 1.63 22.86 6 0.77 21.44 6 1.33 20.90 6 1.08 22.73 6 0.85 21.37 6 1.75
20.93 6 0.92 22.57 6 0.99 20.86 6 1.34
21.27 6 1.67 22.86 6 0.83 21.04 6 1.42
20.04 6 0.59 0.16 6 0.46 0.51 6 0.86
20.06 6 0.82 20.01 6 0.32 0.40 6 0.81
0.07 (20.31 to 0.45) 0.17 (20.06 to 0.40) 0.13 (20.26 to 0.53)
0.2 0.2 0.18 (20.11 to 0.43) 0.19 (20.09 to 0.44) 0.11 (20.18 to 0.40) 0.04 (20.15 to 0.22) 21.55 6 1.26 20.84 6 0.85
Primary cognitive end point Overall composite cognitive end point Reaction time domain Psychomotor speed domain Processing speed domain
21.37 6 1.20 20.81 6 0.71
21.14 6 0.96 20.64 6 0.64
21.40 6 1.30 20.70 6 0.79
0.22 6 0.57 0.17 6 0.36
0.15 6 0.58 0.14 6 0.41
Pearson Correlation Coefficient (95% CI) Estimated Treatment Effect (95% CI) Placebo Tolvaptan Placebo
Tolvaptan
Placebo
Change From Baseline Follow-up Baseline
Tolvaptan Neurocognitive Composite z Score
Table 2. Correlations Between Changes From Baseline in Serum Sodium Concentration and Neurocognitive Composite z Score End Points
Pa
Verbalis et al
This trial demonstrated the effects of oral tolvaptan on correction of serum sodium concentrations to the normal range in adult patients with mild hyponatremia and the effects of this correction on measures of cognition, gait, and bone metabolism. Compared with normative values from age-matched individuals, study participants had impaired cognitive performance at baseline on measures of correct choice and speed of choice, impaired postural and gait stability, and reduced SF-12 physical component scores, especially for the Physical Functioning, Role– Physical, and Bodily Pain domains. Participants showed preserved cognitive performance on measures relying on word and number recognition and preserved SF-12 mental component scores, particularly the Mental Health and Social Functioning domains. It is important to note that some imbalances between treatment groups in baseline neurocognitive function scores were noted in this population. Further, some baseline test results were close to normal. For example, a ceiling effect was observed for tolvaptantreated patients with normal SF-12 Mental Component Summary scores at baseline, who therefore had no opportunity for improvement. Major limitations of this study include the small sample size and short treatment duration. Because this was a pilot study, a formal sample size power calculation for the primary end point was not performed. However, previous unblinded crossover trials have shown significant neurocognitive improvements with as few as 15 patients.15-19 The present trial studied a different population using related, but distinct, instruments. Tests that measure reaction time, speed, and processing speed were chosen for the primary end point based on previous published findings3; however, the trial lacked sufficient power to detect changes in neurocognitive test and SF-12 scores. An additional problem related to the small sample size was the inability to control for underlying comorbid conditions or medications. However, because the study was designed to evaluate hyponatremic patients with Am J Kidney Dis. 2016;67(6):893-901
Impact of Subclinical Hyponatremia
chronic asymptomatic hyponatremia in good general health, significant changes in underlying comorbid conditions or medications were not likely during the short period of study. Although the 21-day study duration is short, this period was sufficient to observe both neurocognitive and gait stability improvements in previous studies of hyponatremic patients3 and served to reduce the likelihood of interval changes in underlying comorbid conditions or medications to affect outcome measures. Tolvaptan was effective in reversing hyponatremia in this small population of patients; however, a goal of serum sodium concentration $ 138 mEq/L was reached in only 50% of patients treated with tolvaptan. This was likely secondary to increased thirst at higher serum sodium concentrations, resulting in equilibration at a new set point at which fluid intake balanced the tolvaptan-induced aquaresis. Although the difference between groups did not reach statistical significance, a trend for improvement in the neurocognitive composite score of speed domains, the primary study end point, was observed in the tolvaptan group. Changes from baseline in cognition test results pertaining to speed domains were uniformly numerically in favor of tolvaptan (small effect sizes . 0.2), with most missing statistical significance but showing positive correlations with changes in serum sodium concentrations. However, the psychomotor speed domain both reached statistical significance for tolvaptan versus placebo (small and moderate effect sizes of 0.27 for sodium , 135 mEq/L and 0.53 for sodium , 130 mEq/L, respectively) and correlated well with increased serum sodium concentrations. Compared to the placebo group, a tendency to regress to baseline was observed for each of the end points on withdrawal of tolvaptan treatment (ie, at the end of 7-day follow-up), which again suggests that the effect was related to the correction of serum sodium concentration. Effect sizes of 0.27 and 0.53 for improvement in the psychomotor speed domain in tolvaptan-treated patients are clinically relevant, as defined by Cohen D criteria (small effect size 5 0.2 SD, moderate effect size 5 0.5 SD, and large effect size 5 0.8 SD).21 The correlation of effect with serum sodium concentration and the tendency to regress on withdrawal of tolvaptan treatment support the hypothesis that the observed small effect of tolvaptan (and subsequent correction of serum sodium concentration) on each neurocognition test or domain is nonrandom.22 An exception was noted with the negative change in neurocognitive postural stability in patients treated with tolvaptan, which further declined through the 7day follow-up, suggesting a possible random effect for this measure. Baseline imbalances between the 2 treatment groups may have contributed to this result. Am J Kidney Dis. 2016;67(6):893-901
Similar effect sizes for improvements in the range of 0.3 to 0.6 using the Cognitive Drug Research system have been previously documented to demonstrate clinically meaningful improvements in various parameters of attention (eg, rivastigmine for Parkinson disease15 and Lewy body dementia,16 NS2359 for attention-deficit/hyperactivity disorder,17 and modafinil for narcolepsy19) and improved processing speed (eg, reboxetine for major depressive disorder18 and modafinil for narcolepsy19). The bone markers osteocalcin concentration and NTx-creatinine ratio were exploratory end points based on the reported impact of hyponatremia on fractures and osteoporosis.4,6-8,23-25 Results were favorable for tolvaptan treatment because osteocalcin concentration (a marker of bone formation) increased and NTx-creatinine ratio (a bone marker of resorption) decreased to a greater extent than with placebo. Although changes in individual markers of bone formation and resorption were not statistically significant, calculation of a bone resorption index, defined as change from baseline in urine NTx-creatinine ratio divided by serum osteocalcin concentration, showed a highly significant decrease in the tolvaptan group compared to the placebo group despite a small sample size of patients with both bone markers measured (tolvaptan, n 5 15; placebo, n 5 18). Markers of bone metabolism are used clinically to detect changes in bone resorption and formation in response to therapeutic interventions in patients with osteopenia and osteoporosis,26,27 and the concept of a bone resorptive index (or “uncoupling factor”) that arithmetically combines both resorption and formation markers has been proposed as a potentially more sensitive measure of bone metabolism.28 The striking improvement in resorptive index after just 22 days of hyponatremia correction with tolvaptan suggests that long-term treatment of hyponatremia may result in an improvement in bone mineral density, which was recently reported in an individual patient.29 The improvement in the psychomotor speed domain also has important potential clinical relevance. Patients with mild chronic hyponatremia have been found to have gait instability and an increased rate of falls.3 This suggests impairment of the fine motor control that is necessary to maintain upright stability during walking, which is supported by the significant decrease in tapping speed measured by the psychomotor speed domain in the present trial. Although the cause of this deficit is speculative, it is notable that the chronically hyponatremic brain undergoes a volume regulatory decrease in cell volume through solute extrusion from brain cells.30 Among the solutes lost are low-molecular-weight organic osmolytes, including some amino acids that serve as excitatory neurotransmitters.31 These include glutamate, the 899
Verbalis et al
major excitatory neurotransmitter in motor circuits, which can be decreased by as much as 40% in rats with chronic hyponatremia.32 It is reasonable to speculate that such profound depletion of transmitters that are vitally important for control of motor function could slow nerve conduction and thereby impair the rapid motor corrections necessary to maintain a stable gait or enable rapid movements, such as finger tapping. The combined effects of gait instability, increased falls, and osteoporosis could lead to the increased fracture rates observed in multiple independent studies of patients with hyponatremia.4,23-25 Because the gait instability appears to be reversible with correction of hyponatremia,3 and the bone resorption index observed in the present trial suggests that the bone loss associated with hyponatremia may be reversible as well, there is a strong rationale to develop clinical trials to assess the effect of hyponatremia correction on falls, fractures, and osteoporosis. Tolvaptan was generally safe and well tolerated in this trial, and no new safety concerns were raised. Commonly reported ($10%) treatment-emergent AEs were consistent with the aquaretic action of tolvaptan, and few serious AEs were reported. No apparent trends were observed in response to treatment with tolvaptan regarding vital signs, ECGs, or clinical laboratory parameters. No incidence of overly rapid correction of serum sodium concentration (.12 mEq/L in a 24-hour period) was observed. In conclusion, this trial demonstrated that in a small adult population with hyponatremia without overt neurologic or mental status changes at baseline, tolvaptan was effective in reversing hyponatremia, which correlated with improvements in tests of neurocognition, particularly rapid motor movements. The results also suggest a potential reversal of hyponatremia-induced resorptive bone loss, as reflected by the significant improvement in bone resorption index following short-term correction of hyponatremia. Although tolvaptan was used to correct hyponatremia in this study, improvements in neurocognitive deficits correlated with improvements in serum sodium concentrations and would be expected regardless of the methods used to improve serum sodium concentrations. Future studies in larger groups of hyponatremic patients are needed to determine what type, degree, and length of therapy are necessary to produce and maintain neurocognitive and metabolic benefits.
Research Center; Ali Khojasteh, Columbia Comprehensive Care Clinic; John LaFata, Progressive Clinical Research; Michael J. Lillestol, Lillestol Research LLC; Sarah Olelewe, Hawthorne, CA; David Rosenbaum, Pikes Peak Cardiology; Mitchell H. Rosner, University of Virginia Health System West Hospital Complex Nephrology Clinical Research Center; Miguel Trevino, Innovative Research of West FL; Joseph Verbalis, Georgetown University; and Wayne Wells, Lebanon, TN. Support: Otsuka Pharmaceutical Development & Commercialization Inc provided the financial support for this study. Otsuka designed the study, collected and analyzed the data, and played a role in its interpretation. Otsuka assisted in reviewing the report and was involved in the decision to submit the results of this study for publication. BioScience Communications (New York, NY) provided assistance with formatting the manuscript for submission that was funded by Otsuka. Dr Verbalis has received research support and non2continuing medical education2related fees from Otsuka; Dr Ellison has received research support from Otsuka; Drs Hobart, Ouyang, and Czerwiec and Ms Krasa are employees of Otsuka. Financial Disclosure: Dr Verbalis has received non2 continuing medical education2related fees from Cornerstone Therapeutics and Ferring Pharmaceuticals. Contributions: Study design: JGV, MH, HK, FSC; data acquisition: JGV, HE; data analysis/interpretation: JGV, MH, HK, JO, FSC; statistical analysis: JGV, MH, HK, JO, FSC. All prespecified primary and secondary outcomes of the study were reported in this study. Each author contributed important intellectual content during manuscript drafting or revision and accepts accountability for the overall work by ensuring that questions pertaining to the accuracy or integrity of any portion of the work are appropriately investigated and resolved. JGV takes responsibility that this study has been reported honestly, accurately, and transparently; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and registered) have been explained. Peer Review: Evaluated by 3 external peer reviewers, a Statistical Editor, a Co-Editor, and the Editor-in-Chief.
ACKNOWLEDGEMENTS
1. Upadhyay A, Jaber BL, Madias NE. Epidemiology of hyponatremia. Semin Nephrol. 2009;29(3):227-238. 2. Hawkins RC. Age and gender as risk factors for hyponatremia and hypernatremia. Clin Chim Acta. 2003;337(1-2):169-172. 3. Renneboog B, Musch W, Vandemergel X, Manto MU, Decaux G. Mild chronic hyponatremia is associated with falls, unsteadiness, and attention deficits. Am J Med. 2006;119(1): 71.e1-71.e8.
The INSIGHT Investigators are Otis Barnum, Natchitoches Hospital; Karthikeya Devireddy, KD Medical Group Inc; Howard Ellison, Rockdale Medical Research Associates; Miguel Franco, Memorial Clinical Associates; Francis Goldstein, Carolina Research Associates; Terrence C. Hack, Primary Care Cardiology Research Inc; Kianoosh Kaveh, Coastal Nephrology Associates 900
SUPPLEMENTARY MATERIAL Table S1: Entry criteria. Table S2: Baseline SF-12, mental and physical component scores. Table S3: Adverse events. Figure S1: Mean serum sodium concentrations at study visits. Figure S2: Treatment effects on overall NCS and all 7 components, and on gait and stability measures. Figure S3: Diagram of neurocognitive tests for each NCS domain and treatment effects. Figure S4: Effects on markers of bone turnover. Item S1: Neurocognitive battery methods, gait and stability, and statistical methods for secondary and exploratory end points. Note: The supplementary material accompanying this article (http://dx.doi.org/10.1053/j.ajkd.2015.12.024) is available at www.ajkd.org
REFERENCES
Am J Kidney Dis. 2016;67(6):893-901
Impact of Subclinical Hyponatremia 4. Kinsella S, Moran S, Sullivan MO, Molloy MG, Eustace JA. Hyponatremia independent of osteoporosis is associated with fracture occurrence. Clin J Am Soc Nephrol. 2010;5(2):275-280. 5. Hoorn EJ, Rivadeneira F, van Meurs JB, et al. Mild hyponatremia as a risk factor for fractures: the Rotterdam Study. J Bone Miner Res. 2011;26(8):1822-1828. 6. Verbalis JG, Barsony J, Sugimura Y, et al. Hyponatremiainduced osteoporosis. J Bone Miner Res. 2010;25(3):554-563. 7. Jamal SA, Arampatzis S, Harrison SL, et al. Hyponatremia and fractures: findings from the Osteoporotic Fractures in Men Study. J Bone Miner Res. 2015;30(6):970-975. 8. Kruse C, Eiken P, Vestergaard P. Hyponatremia and osteoporosis: insights from the Danish National Patient Registry. Osteoporos Int. 2015;26(3):1005-1016. 9. Usala RL, Fernandez SJ, Mete M, et al. Hyponatremia is associated with increased osteoporosis and bone fractures in a large US health system population. J Clin Endocrinol Metab. 2015;100(8):3021-3031. 10. Greenberg A, Verbalis JG, Amin AN, et al. Current treatment practice and outcomes. Report of the hyponatremia registry. Kidney Int. 2015;88(1):167-177. 11. Schrier RW, Gross P, Gheorghiade M, et al. Tolvaptan, a selective oral vasopressin V2-receptor antagonist, for hyponatremia. N Engl J Med. 2006;355(20):2099-2112. 12. Berl T, Quittnat-Pelletier F, Verbalis JG, et al. Oral tolvaptan is safe and effective in chronic hyponatremia. J Am Soc Nephrol. 2010;21(4):705-712. 13. Simpson PM, Surmon DJ, Wesnes KA, Wilcock GK. The Cognitive Drug Research computerised assessment system for demented patients: a validation study. Int J Geriatr Psychiatry. 1991;6:95-102. 14. Wesnes K, Pincock C. Practice effects on cognitive tasks: a major problem? Lancet Neurol. 2002;1(8):473. 15. Wesnes KA, McKeith IG, Edgar C, et al. Benefits of rivastigmine on attention in dementia associated with Parkinson disease. Neurology. 2005;65(10):1654-1656. 16. Wesnes KA, McKeith IG, Ferrara R, et al. Effects of rivastigmine on cognitive function in dementia with Lewy bodies: a randomised placebo-controlled international study using the Cognitive Drug Research computerised assessment system. Dement Geriatr Cogn Disord. 2002;13(3):183-192. 17. Wilens TE, Klint T, Adler L, et al. A randomized controlled trial of a novel mixed monoamine reuptake inhibitor in adults with ADHD. Behav Brain Funct. 2008;4:24. 18. Ferguson JM, Wesnes KA, Schwartz GE. Reboxetine versus paroxetine versus placebo: effects on cognitive functioning
Am J Kidney Dis. 2016;67(6):893-901
in depressed patients. Int Clin Psychopharmacol. 2003;18(1): 9-14. 19. Harsh JR, Hayduk R, Rosenberg R, et al. The efficacy and safety of armodafinil as treatment for adults with excessive sleepiness associated with narcolepsy. Curr Med Res Opin. 2006;22(4): 761-774. 20. Torres VE, Chapman AB, Devuyst O, Gansevoort RT, Grantham JJ, Higashihara E, Perrone RD, Krasa HB, Ouyang J, Czerwiec FS; TEMPO 3:4 Trial Investigators. Tolvaptan in patients with autosomal dominant polycystic kidney disease. N Engl J Med. 2012;367(25):2407-2418. 21. Cohen J. Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Hillsdale, NJ: Lawrence Erlbaum; 2015. 22. Verbalis JG, Ouyang J, Orlandi C, Czerwiec FS. Serum sodium correction with tolvaptan improves self-reported mental health status in hyponatremic patients: a subanalysis from the SALT trials [abstract]. Presented at: The Endocrine Society’s 89th Annual Meeting; June 2-5, 2007; Toronto, Canada. 23. Sandhu HS, Gilles E, DeVita MV, Panagopoulos G, Michelis MF. Hyponatremia associated with large-bone fracture in elderly patients. Int Urol Nephrol. 2009;41(3):733-737. 24. Gankam KF, Andres C, Sattar L, Melot C, Decaux G. Mild hyponatremia and risk of fracture in the ambulatory elderly. QJM. 2008;101(7):583-588. 25. Hoorn EJ, Liamis G, Zietse R, Zillikens MC. Hyponatremia and bone: an emerging relationship. Nat Rev Endocrinol. 2011;8(1):33-39. 26. Naylor K, Eastell R. Bone turnover markers: use in osteoporosis. Nat Rev Rheumatol. 2012;8(7):379-389. 27. Eastell R, Hannon RA. Biomarkers of bone health and osteoporosis risk. Proc Nutr Soc. 2008;67(2):157-162. 28. Eastell R, Robins SP, Colwell T, Assiri AM, Riggs BL, Russell RG. Evaluation of bone turnover in type I osteoporosis using biochemical markers specific for both bone formation and bone resorption. Osteoporos Int. 1993;3(5):255-260. 29. Sejling AS, Thorsteinsson AL, Pedersen-Bjergaard U, Eiken P. Recovery from SIADH-associated osteoporosis: a case report. J Clin Endocrinol Metab. 2014;99(10):3527-3530. 30. Verbalis JG. Brain volume regulation in response to changes in osmolality. Neuroscience. 2010;168(4):862-870. 31. Gullans SR, Verbalis JG. Control of brain volume during hyperosmolar and hypoosmolar conditions. Annu Rev Med. 1993;44:289-301. 32. Verbalis JG, Gullans SR. Hyponatremia causes large sustained reductions in brain content of multiple organic osmolytes in rats. Brain Res. 1991;567:274-282.
901