Cosmetic side effects of antiepileptic drugs in adults with epilepsy.

OBJECTIVE: Cosmetic side effects (CSEs) such as weight gain and alopecia are common, undesirable effects associated with several AEDs. The objective of the study was to compare the CSE profiles in a large specialty practice-based sample of patients taking both older and newer AEDs. METHODS: As part of the Columbia and Yale AED Database Project, we reviewed patient records including demographics, medical history, AED use, and side effects for 1903 adult patients (≥16years of age) newly started on an AED. Cosmetic side effects were determined by patient or physician report in the medical record and included acne, gingival hyperplasia, hair loss, hirsutism, and weight gain. We compared the overall rate of CSEs and intolerable CSEs (ICSEs-CSEs that led to dosage reduction or discontinuation) between different AEDs in both monotherapy and polytherapy. RESULTS: Overall, CSEs occurred in 110/1903 (5.8%) patients and led to intolerability in 70/1903 (3.7%) patients. Weight gain was the most commonly reported CSE (68/1903, 3.6%) and led to intolerability in 63 (3.3%) patients. Alopecia was the second most common patient-reported CSE (36/1903, 1.9%) and was intolerable in 33/1903 (1.7%) patients. Risk factors for CSEs included female sex (7.0% vs. 4.3% in males; p<0.05) and any prior CSE (37% vs. 2.9% in patients without prior CSE; p<0.001). Significantly more CSEs were attributed to valproic acid (59/270; 21.9%; p<0.001) and pregabalin (14/143; 9.8%; p<0.001) than to all other AEDs. Significantly less CSEs were attributed to levetiracetam (7/524; 1.3%; p=0.002). Weight gain was most frequently associated with valproic acid (35/270; 13.0%; p<0.001) and pregabalin (12/143; 8.4%; p<0.001). Hair loss was most commonly reported among patients taking valproic acid (24/270; 8.9%; p<0.001). Finally, gingival hyperplasia was most commonly reported in patients taking phenytoin (10/404; 2.5%; p<0.001). Cosmetic side effects leading to dosage change or discontinuation occurred most frequently with pregabalin and valproic acid compared with all other AEDs (13.3 and 5.6% vs. 2.3%; p<0.001). For patients who had been on an AED in monotherapy (n=677), CSEs and ICSEs were still more likely to be attributed to valproic acid (30.2% and 17.1%, respectively) than to any other AED (both p<0.001). SIGNIFICANCE: Weight gain and alopecia were the most common patient-reported CSEs in this study, and weight gain was the most likely cosmetic side effect to result in dosage adjustment or medication discontinuation. Particular attention should be paid to pregabalin, phenytoin, and valproic acid when considering cosmetic side effects. Female patients and patients who have had prior CSE(s) to AED(s) were more likely to report CSEs. Knowledge of specific CSE rates for each AED found in this study may be useful in clinical practice.

The N-terminal pleckstrin, coiled-coil, and IQ domains of the exchange factor Ras-GRF act cooperatively to facilitate activation by calcium.

We have recently shown that the neuronal exchange factor p140 Ras-GRF becomes activated in vivo in response to elevated calcium levels [C. L. Farnsworth, N. W. Freshney, L. B. Rosen, A. Ghosh, M. E. Greenberg, and L. A. Feig, Nature (London) 376:524-527, 1995]. Activation is mediated by calcium-induced calmodulin binding to an IQ domain near the N terminus of Ras-GRF. Here we show that the adjacent N-terminal pleckstrin homology (PH), coiled-coil, and IQ domains function cooperatively to allow Ras-GRF activation. Deletion of the N-terminal PH domain redistributes a large percentage of Ras-GRF from the particulate to the cytosolic fraction of cells and renders the protein insensitive to calcium stimulation. A similar cellular distribution and biological activity are observed when only the core catalytic domain is expressed. Although the PH domain is necessary for particulate association of Ras-GRF, it is not sufficient for targeting the core catalytic domain to this cellular location. This requires the PH domain and the adjacent coiled-coil and IQ sequences. Remarkably, this form of Ras-GRF is constitutively activated. The PH and coiled-coil domains must also perform an additional function, since targeting to the particulate fraction of cells is not sufficient to allow Ras-GRF activation by calcium. A Ras-GRF mutant containing the PH domain from Ras-GTPase-activating protein in place of its own N-terminal PH domain localizes to the particulate fraction of cells but does not respond to calcium. Similar phenotypes are seen with mutant Ras-GRFs containing point mutations in either the PH or coiled-coil domain. These findings argue that the N-terminal PH, coiled-coil, and IQ domains of Ras-GRF function together to connect Ras-GRF to multiple components in the particulate fractions of cells that are required for responsiveness of the protein to calcium signaling.

Psychiatric and behavioral side effects of the newer antiepileptic drugs in adults with epilepsy.

OBJECTIVE: Psychiatric/behavioral side effects (PSEs) are common in patients taking antiepileptic drugs (AEDs). The objective of the study described here was to compare the PSE profiles of the newer AEDs. METHODS: We examined the charts of 1394 adult outpatients seen at the Columbia Comprehensive Epilepsy Center who had taken one of the newer AEDs. We compared the rate of AED-related PSEs in patients newly started on the newer AEDs both before and after controlling for non-AED predictors of PSEs. RESULTS: Overall, 221 of 1394 (16%) patients experienced PSEs. The average rate of AED-related PSEs for a single AED was 8.4%, with 6.1% resulting in dosage change and 4.3% resulting in AED discontinuation. Significantly fewer PSEs were attributed to gabapentin (n=160, 0.6% incidence, P<0.001) and lamotrigine (n=547, 4.8% incidence, P<0.001), and significantly more PSEs were attributed to levetiracetam (n=521, 15.7% incidence, P<0.001; 8.8% discontinued LEV because of PSEs). Vigabatrin, felbamate, and oxcarbazepine were associated with similarly low rates of PSEs in many analyses but with fewer of patients. Tiagabine was associated with high PSE rates (similar to those for levetiracetam), but was used much less commonly at our center. Intermediate rates of PSEs were attributed to topiramate and zonisamide (both nonsignificant). Psychiatric history was the most significant nondrug predictor of AED-related PSEs (PSEs occurred in 23% of patients with a psychiatric history vs 12% of patients without such a history, P<0.001). The relative rates of AED-related PSEs were similar when controlling for non-AED predictors and when analyzing only patients on monotherapy. CONCLUSIONS: There are significant differences between the newer AEDs in terms of their PSE profiles. Patients taking levetiracetan experience significantly more PSEs than average, and patients taking gabapentin and lamotrigine experience significantly fewer PSEs. Even with the medication with the highest rate of PSEs (levetiracetam), less than 10% of patients discontinued it because of PSEs. A past psychiatric condition is the most significant nondrug predictor of AED-related PSEs.

Psychiatric and behavioral side effects of anti-epileptic drugs in adolescents and children with epilepsy.

PURPOSE: The objective of the study was to compare the psychiatric and behavioral side effect (PBSE) profiles of both older and newer antiepileptic drugs (AEDs) in children and adolescent patients with epilepsy. METHOD: We used logistic regression analysis to test the correlation between 83 non-AED/patient related potential predictor variables and the rate of PBSE. We then compared for each AED the rate of PBSEs and the rate of PBSEs that led to intolerability (IPBSE) while controlling for non-AED predictors of PBSEs. RESULTS: 922 patients (≤18 years old) were included in our study. PBSEs and IPBSEs occurred in 13.8% and 11.2% of patients, respectively. Overall, a history of psychiatric condition, absence seizures, intractable epilepsy, and frontal lobe epilepsy were significantly associated with increased PBSE rates. Levetiracetam (LEV) had the greatest PBSE rate (16.2%). This was significantly higher compared to other AEDs. LEV was also significantly associated with a high rate of IPBSEs (13.4%) and dose-decrease rates due to IPBSE (6.7%). Zonisamide (ZNS) was associated with significantly higher cessation rate due to IPBSE (9.1%) compared to other AEDs. CONCLUSION: Patients with a history of psychiatric condition, absence seizures, intractable epilepsy, or frontal lobe epilepsy are more likely to develop PBSE. PBSEs appear to occur more frequently in adolescent and children patients taking LEV compared to other AEDs. LEV-attributed PBSEs are more likely to be associated with intolerability and subsequent decrease in dose. The rate of ZNS-attributed IPBSEs is more likely to be associated with complete cessation of AED.

The effect of flavonoids on aerobic glycolysis and growth of tumor cells.

Certain bioflavonoids inhibit the glycolysis of variety of tumor cells by interfering with the generation of adenosine diphosphate and inorganic phosphate which are required for glycolysis. Tetra- and pentahydroxy flavones with hydroxyl groups as 3, 3', 4', 5, and 7 (e.g., quercetin) are the most active. They inhibit the activity of isolated Na+-K+-adenosinetriphosphatase of the plasma membrane and of mitochondrial adenosinetriphosphatase, but under appropriate conditions do not interfere with the ion transport increase the the translocation efficiency of the ion pump. It was shown that in several tumor cells loosely coupled ion pumps are responsible for the high rate of aerobic glycolysis, the effect of quercetin on the growth of several cell lines was examined. Since bicarbonate and serum albumin were found to counteract the effect of quercetin, the cells were grown in tissue cultures at low concentrations of these compounds. Pronounced inhibition of growth was observed at 5 to 20 mug of quercetin per ml of growth medium.

Interstitial hypertension in carcinoma of uterine cervix in patients: possible correlation with tumor oxygenation and radiation response.

Elevated tumor interstitial fluid pressure (IFP) is believed to be responsible, at least in part, for the poor penetration and heterogeneous distribution of blood-borne therapeutic agents and nutrients in solid tumors. Using the wick-in-needle technique, IFP was measured in human patients with squamous cell carcinoma of the uterine cervix at the initial and final stages of fractionated external beam radiotherapy. Mean IFP values ranged from 10 to 26 mm Hg with an overall mean of 15.7 +/- 5.7 (SD) mm Hg in stage IIB and IIIB tumors (n = 12) and from 0 to 3 mm Hg in normal cervix (n = 3). IFP decreased in some patients with therapy while in others it increased. The changes in IFP values agree well with the clinical response to radiotherapy (n = 7, P less than 0.05). Oxygen tension, measured in selected tumors (n = 3) with polarographic oxygen microelectrodes, inversely correlated with IFP. These results show for the first time that the IFP in human cervical carcinomas is elevated, and that it can be lowered in some tumors using fractionated radiation therapy. These findings also suggest that IFP values may provide an indication of tumor oxygenation and that IFP modifications could be prognostic indicators of radiation response.

EBV-specific cytotoxic T lymphocytes protect against human EBV-associated lymphoma in scid mice.

Scid mice develop human EBV-positive B cell tumors after injection with EBV-transformed human B cells or peripheral blood lymphocytes from EBV-seropositive donors. Injection of cytotoxic T cell (CTL) lines, selectively expanded in vitro to recognize autologous EBV-transformed B cells, protected mice against developing human lymphoproliferative disease. CTL protection required a single dose of polyethylene-glycol-conjugated IL2. Mice were not protected by CTL that did not specifically recognize EBV-transformed B cells or by mitogen-activated lymphocytes.

Serum lactate dehydrogenase-3 isoenzyme in chronic granulocytic leukemia.

Abnormal levels of serum lactate dehydrogenase-3 (LD-3) activity were observed in 92% of patients (35 of 38) with active chronic granulocytic leukemia (CGL), in 40% of patients (4 of 10) in partial remission, and in 13% of patients (1 of 8) in complete remission. In evaluating the electrophoretic LD isoenzyme patterns of these patients, three criteria were used. In criterion-1 elevations, the LD-3/total LD ratio, expressed as a fraction of serum total LD, and LD-3 value, expressed in absolute units, were greater than the upper limit of the reference range. In criterion-2 elevations, only the LD-3/total LD ratio was greater than the upper limit of the reference range. In criterion-3 elevations, only the absolute LD-3 activity exceeded the upper limit of the reference range, and these specimens showed isomorphic elevation of all five LD isoenzymes. Use of the last of these criteria increased the clinical sensitivity of serum LD-3 elevations in active CGL from 82% to 92%. The mean serum LD-3 absolute value and serum total LD activity usually showed statistically significant differences (P less than 0.05) among patients with active CGL, those in partial remission, and those in complete remission, but did not distinguish between subgroups of individuals with active CGL. Elevation of the serum LD-5/total LD ratio in 16 of 58 patients was due to hepatic injury or methodologic imprecision, showing analytically insignificant, borderline abnormalities in all cases of active CGL. In 10 of 62 patients in complete remission or partial remission, however, such elevation was unexplained. Our results indicate that the evaluation of serum LD-3 values in both absolute and relative terms slightly increases the clinical sensitivity of LD-3, and, therefore, suggest that LD-3 might be a useful marker for CGL.

Variability in the chemistry of estuarine plants and its effect on feeding by Canada geese.

We investigated the influence of interspecific and seasonal variations in plant chemistry on food choices by adult and gosling Canada Geese, Branta canadensis, on Cape Cod, Massachusetts. The geese fed primarily on the abundant marsh grasses, Spartina spp., and rushes, Juncus gerardi, early in the growing season and switched to a greater dependence on eelgrass, Zostera marina, later. Forbs were generally avoided all season even when growing within patches of abundant species. The avoidance of forbs was related to their low abundance and their high concentrations of deterrent secondary metabolites. Differences in plant chemistry also determined the switch from marsh graminoids to Z. marina during the growing season. Marsh grasses were higher than Z. marina in nitrogen, particularly in the spring when the nitrogen requirement of geese is especially high. Z. marina was a better source of soluble carbohydrates and was the preferred food during the summer when the need to build up energy reserves may be more critical to geese than protein intake. Goslings, which require a diet higher in nitrogen than do adults, fed on marsh graminoids later into the growing season than the adults. The nitrogen content of the diets of goslings was significantly higher than that available to them in the plants, indicating that they selected for introgen. The diets of non-breeding adults in the spring and all geese in mid summer closely reflected the nutrient content of the plants. The diet of breeding adults was more similar to that of their goslings than to that of non-breeding adults. The effects of plant chemistry and the nutritional needs of geese on food choices were modified by the need to select a safe feeding site.

Acute-phase proteins in patients with head and neck cancer treated with interleukin 2/interferon alfa.

Circulating acute-phase proteins may mediate adverse reactions in patients receiving biologic response modifiers, including inhibition of immune responsiveness and clinical toxic effects. Nine patients with unresectable head and neck squamous cell carcinoma were prospectively examined for levels of acute-phase proteins during interleukin 2/interferon alfa immunotherapy and for clinical toxic effects. Simultaneous determination of the in vitro immunomodulatory capacity of autologous serum on the induction of lymphokine-activated killer cells was assessed in 4-hour chromium release assays. Of the seven acute-phase proteins analyzed, haptoglobin and C-reactive protein levels were elevated before therapy was started. Toxic events leading to cessation of interleukin 2/interferon alfa therapy had a high correlation with elevated C-reactive protein and lowered C3 component of complement levels. No relationship was noted between serum levels of acute-phase proteins and induction inhibition of lymphokine-activated killer cell cytotoxicity. The role of C-reactive protein and complement degradation products in mediating interleukin 2/interferon alfa toxicity requires further investigation.

A comparison of post-mortem ethanol levels obtained from blood and subdural specimens.

Post-mortem subdural ethanol levels have been proposed as a useful test in certain forensic cases involving head trauma, particularly when the time interval from injury to death may have caused a lowering of the blood ethanol concentration to insignificant or undetectable levels. This study of 75 autopsied persons from whom both blood and subdural ethanol levels were obtained, shows the usefulness of the subdural ethanol level, especially where there is a prolonged or unknown post-traumatic time interval. Use of such a test is recommended in these situations.

Immunomodulation of the induction phase of lymphokine-activated killer activity by acute phase proteins.

Effective treatment of head and neck cancer with biologic response modifiers may be benefitted by an understanding of in vivo factors capable of modulating the lymphokine-activated killer (LAK) cell phenomenon. Eighteen patients with squamous cell carcinoma of the head and neck were studied. Killer cells from each patient, activated by recombinant interleukin-2 (10 U/ml), were induced in either complete medium alone or complete medium plus 10% autologous serum solution and analyzed. Cytotoxicity against both K562 and squamous cell carcinoma (MDA686-Ln) cell lines was determined by use of standard chromium-release assays. The immunomodulatory capacity of serum was correlated with levels of various acute phase proteins. Autologous serum significantly inhibited the induction phase of the LAK phenomenon in 61% of patients and stimulated it in 22%. No patients with early stage I or II disease had significant inhibition of induction. No direct correlation between inhibition and serum acute phase protein levels were seen. An inverse relationship was seen between the C3 component of complement and induction inhibition (r = -0.6). These findings suggest that advances of in vivo immunomodulatory therapy will require elucidation of mechanisms of serologic inhibition of the induction phase of the LAK phenomenon. Such studies may lead to serologic modification to enhance treatment efficacy of biologic response modifiers.

Point dose variations with time during traditional brachytherapy for cervical carcinoma.

In traditional brachytherapy for carcinoma of the cervix, doses are often prescribed to specifically chosen points (A and B) and the normal tissue tolerance calculated at specific reference points in the bladder and rectum. These tolerance doses are often used to modify the brachytherapy treatment plan. It is inherently assumed that the position of the brachytherapy applicator does not change in relation to the relevant anatomical structures over the time-course of an implant. To assess the accuracy of this assumption, 2 sets of localization films were obtained for each implant in 28 patients, 1 prior to loading and another after the removal of the radioactive sources. Significant applicator movement and, consequently, significant dose variations were ob: served. Therefore, isolated one-time dose measurements to normal critical structures should not be used as the sole basis for making therapeutic decisions. The magnitude of dose variations and their clinical significant are discussed.

The role of fibroblast Tiam1 in tumor cell invasion and metastasis.

The co-evolution of tumors and their microenvironment involves bidirectional communication between tumor cells and tumor-associated stroma. Various cell types are present in tumor-associated stroma, of which fibroblasts are the most abundant. The Rac exchange factor Tiam1 is implicated in multiple signaling pathways in epithelial tumor cells and lack of Tiam1 in tumor cells retards tumor growth in Tiam1 knockout mouse models. Conversely, tumors arising in Tiam1 knockout mice have increased invasiveness. We have investigated the role of Tiam1 in tumor-associated fibroblasts as a modulator of tumor cell invasion and metastasis, using retroviral delivery of short hairpin RNA to suppress Tiam1 levels in three different experimental models. In spheroid co-culture of mammary epithelial cells and fibroblasts, Tiam1 silencing in fibroblasts led to increased epithelial cell outgrowth into matrix. In tissue-engineered human skin, Tiam1 silencing in dermal fibroblasts led to increased invasiveness of epidermal keratinocytes with pre-malignant features. In a model of human breast cancer in mice, co-implantation of mammary fibroblasts inhibited tumor invasion and metastasis, which was reversed by Tiam1 silencing in co-injected fibroblasts. These results suggest that stromal Tiam1 may have a role in modulating the effects of the tumor microenvironment on malignant cell invasion and metastasis. This suggests a set of pathways for further investigation, with implications for future therapeutic targets.

Cross-sensitivity of skin rashes with antiepileptic drug use.

OBJECTIVE: To determine rates of cross-sensitivity of rash among commonly used antiepileptic drugs (AEDs) in patients with epilepsy. METHODS: The incidence of AED-related rash was determined in 1875 outpatients (> or =12 years), taking carbamazepine (CBZ), clobazam (CLB), felbamate (FBM), gabapentin (GBP), levetiracetam (LEV), lamotrigine (LTG), oxcarbazepine (OXC), phenobarbital (PB), phenytoin (PHT), primidone (PRM), tiagabine (TGB), topiramate (TPM), vigabatrin (VGB), valproic acid (VPA), or zonisamide (ZNS). We compared rates of rash for each AED in patients with vs those without a rash to 1) another specific AED; 2) any other AED; 3) any two other AEDs; and 4) any non-epilepsy medication. RESULTS: A total of 14.3% (269/1,875) of patients had a rash attributed to at least one AED; 2.8% had a rash to two or more AEDs. Of patients who had a rash to CBZ and were also prescribed PHT (n = 59), 57.6% had a rash to PHT (abbreviated as CBZ --> PHT: 57.6%); of patients who had a rash to PHT and were also prescribed CBZ (n = 81), rate of rash was 42% (i.e., PHT --> CBZ: 42%). Other results: CBZ --> LTG: 20% (n = 50); LTG --> CBZ: 26.3% (n = 38); CBZ --> OXC: 33% (n = 15); OXC --> CBZ: 71.4% (n = 7); CBZ --> PB: 26.7% (n = 30); PB --> CBZ: 66.7% (n = 12); LTG --> PHT: 38.9% (n = 36); PHT --> LTG: 18.9% (n = 74); PB --> PHT: 53.3% (n = 15); PHT --> PB: 19.5% (n = 41); OXC --> LTG: 37.5% (n = 8); LTG --> OXC: 20% (n = 15). There was evidence of specific cross-sensitivity between CBZ and PHT, and between CBZ and PB. CONCLUSION: Cross-sensitivity rates between certain antiepileptic drugs (AEDs) are high, especially when involving carbamazepine and phenytoin. Specific cross-sensitivity rates provided here may be useful for AED selection and counseling in individual patients.

Comparison and predictors of rash associated with 15 antiepileptic drugs.

OBJECTIVE: To determine predictors and relative incidence of antiepileptic drug (AED)-related rash in patients taking all common AEDs. METHODS: We reviewed 1,890 outpatients. Eighty-one variables were tested as potential predictors of rash. We compared the rate of rash attributed to each AED (AED rash) with the average rate of rash attributed to the other AEDs in all adults (aged > or =16 years; n = 1,649) when taking carbamazepine (CBZ), clobazam (CLB), felbamate (FBM), gabapentin (GBP), lamotrigine (LTG), levetiracetam (LEV), oxcarbazepine (OXC), phenobarbital (PB), phenytoin (PHT), primidone (PRM), tiagabine (TGB), topiramate (TPM), vigabatrin (VGB), valproate (VPA), or zonisamide (ZNS). We repeated this analysis for patients with and without the identified nondrug predictors of AED rash. RESULTS: The average rate of AED rash was 2.8%. The only nondrug predictor significant in multivariate analysis was occurrence of another AED rash (odds ratio 3.1, 95% CI 1.8 to 5.1; p < 0.0001); the rate of rash in this subgroup was 8.8%, vs 1.7% in those without another AED rash. Higher AED rash rates were seen with PHT (5.9% overall, p = 0.0008; 25.0% in those with another AED rash, p = 0.001), LTG (4.8%, p = 0.00095; 14.4%, p = 0.025), and CBZ (3.7%, not significant; 16.5%, p = 0.01). Lower rates were seen with LEV (0.6% overall; p = 0.00042), GBP (0.3%, p = 0.00035), and VPA (0.7%, p = 0.01). Rash rates were also low (<1% overall) with FBM, PRM, TPM, and VGB (not significant). These AED differences remained similar in patients with and without other AED rashes. There were four cases of Stevens-Johnson syndrome involving four AEDs. CONCLUSIONS: The rate of an antiepileptic drug (AED) rash is approximately five times greater in patients with another AED rash (8.8%) vs those without (1.7%). Rash rates were highest with phenytoin, lamotrigine, and carbamazepine and low (<1%) with several AEDs.

Development and evaluation of a self-administered version of the ALSFRS-R.

We evaluated the reliability and sensitivity to change over time of a newly developed self-administered version of the ALS functional rating scale-revised (ALSFRS-R) in 60 consecutive patients from an ALS clinic. The self-administered ALSFRS-R showed excellent reliability (intraclass correlation = 0.93, 95% CI: 088 to 0.96) and similar sensitivity to change over time vs the standard evaluator-administered ALSFRS-R.

A two-stage design for a phase II clinical trial of coenzyme Q10 in ALS.

BACKGROUND: The combination of a small pool of patients at any given time with the availability of many potential neuroprotective agents to be tested in ALS requires efficient phase II trial designs. OBJECTIVE: To describe the design of the Clinical Trial of High Dose Coenzyme Q10 (CoQ10) in ALS (QALS study)--a phase II, randomized, placebo-controlled, double-blind, multicenter clinical trial. METHODS: The study design features two stages. The first stage (dose selection) identifies which of two doses of CoQ10 (1800 mg or 2700 mg) is preferred using a selection procedure rather than a formal hypothesis test. The second stage (early efficacy test) compares the preferred dose of CoQ10 against placebo using a non-superiority or futility design. Data from patients assigned to the preferred dose of CoQ10 in the first stage are also used in the second stage. The primary outcome measure is the decline in Amyotrophic Lateral Sclerosis Functional Rating Scale-revised (ALSFRSr) score from baseline to 9 months. RESULTS: The total sample size required is 185 patients, as compared to a much larger sample size estimated to be necessary using a conventional superiority design (total: 852 patients). The authors report a bias correction made necessary by the inclusion of patient data from the first stage in the second stage. CONCLUSIONS: Several features of the Clinical Trial of High Dose Coenzyme Q10 in ALS study design promote efficiency. These features may be beneficial in phase II trials in amyotrophic lateral sclerosis and other fields.