Medial dose summary for
Anticancer activity for
X-ray crystallography studies with
A mechanism of action has been proposed for
The pharmacokinetics, metabolic fate and toxicology of
Cognitive/behavioral studies have been conducted in rats and dogs and are reviewed in depth.
2. Materials and methods
2.1. Drug formulation and chemicals
2.2. High pressure liquid chromatography (HPLC) analysis of 1 and a metabolite, 2
HPLC analysis was performed using an Agilent Technologies (New Castle, DE) 1200 model HPLC fitted with a diode array UV detector set at a wavelength of 244 nm (λmax for
The mobile phase for
Plasma and erythrocyte samples were stored at-74oC until analyzed. Standard solutions were prepared by dissolving 6 mg of
Control dog and rat whole blood and plasma samples were spiked with
Verification of the HPLC assays included calibration curves derived from the assay of five (5) erythrocyte and eleven (11) plasma standards in duplicate prepared each with
2.3. P-glycoprotein (P-gp) transport studies
Adult Sprague-Dawley mice [Crl: CD1(ICR) BR] (males 20-25 g and female 18-25 g) and rats [Crl: CD1(ICR) BR] (males 300-350 g and female 225-250 g) were obtained from Harlan Industries (Indianapolis, IN), housed in groups of three-five per cage in light-controlled (12 h/day) and temperature-controlled (24oC) animal isolators with filtered vents and exhausts. They were fed a diet of Purina Laboratory Chow (Purina Feed) and received tap water
Adult male and female beagle dogs (6.5-7.5 mo. of age, 6.5-9.49 kg) were raised and maintained at MPI Research (Mattawan, MI). They were fed a diet of Purina Dog Chow (Purina Feed), received tap water
All mice, rat and dog studies were conducted at MPI Research (Mattawan, MI) under GLP regulations as described in the Guide for the Care and Use of Laboratory Animals, Office for Laboratory Animal Welfare, NIH, Bethesda, MD.
For mouse toxicity studies,
For the acute behavioral studies, rats received drugs and controls to identify/verify gross behavioral patterns employing a Morris modified water maze (4’ x 3’ x 1.5’) with a single layer of white polyethylene peanuts that floated (in 6’’ of water) and covered a single mounting stage [13, 14]. Adult female rats (Hsd:SD, 175-225 g.) were grouped 3-6 animals per drug arm. The test agents were dissolved in or suspended in 5% aqueous TweenR 80-hydroxypropyl cellulose in 5% saline and administered intraperitoneally (IP). Controls received the vehicle only. Swimming trials began – 1, 2, 3 and 20 hours post-dosing. For each time period post-dosing, the rats were challenged on six (6)-back-to-back swim trial events to find the platform. The daily swimming times and ranges were compared to vehicle controls
2.5. Pharmacokinetic studies
Groups of rats (5/sex) were administered 100, 200, or 300 mg/kg of
Adult beagle dogs (8 M) were administered
2.6. Animal pharmacokinetic data analysis
Model parameters were estimated using Micropharm software and nonlinear least squares regression was performed using Simplex and Gauss Newton algorithms . An open two-compartment model provided the best fit. Clearance, volume of distribution and half-lives were derived from estimates of the model parameters.
2.7. Data analysis
Data analysis was performed on all plasma and
3.1. Stability studies
Bulk drug product,
To further document product stability a 50:50 mixture of
Acute oral and IV toxicity study results for
Clinical signs generally reflecting the deteriorating state of both mice and rats post dosing were observed in both sexes for
3.3. Acute single dose intravenous studies in mice
Adult male and female mice, 10 animals per sex per dose level, were intravenously dosed with 50, 100, 200, 400 and 600 mg/kg. No animals died at 0 or 100 mg/kg, 1 animal died at both 50 and 200 mg/kg doses, 7 of 10 animals died at 400 mg/kg and 8 of 10 animals died at 600 mg/kg (Table 1). Various clinical signs reflecting treatment-related effects were noted in both sexes, oftentimes in a generally dose-dependent manner. These clinical signs included decreased activity, rapid/difficult/slow/shallow breathing, limbs splayed, tremors and skin cold to touch. No seizures or loss of coordination was noted. The deaths at 400 and 600 were of a very immediate nature, occurring within minutes or less post-dose, with no clinical signs exhibited prior to death. While transient incidences of rapid breathing were also noted in a couple of control animals, a definitive relationship to the vehicle was unclear. No definitively clear treatment-related body effects were noted in those mice surviving the 14-day observation period when compared with controls. No macroscopic findings were noted in any animal at necropsy.
Neither aplastic bone marrow nor splenic depletion of lymphocytes was noted.
|Mice||36 M||50-600||IV||132/385||2-14 days||Deaths were erratic|
|Mice||30 M||800-2000||Oral –||0.8 – 2.0 g/d x||21 days|
|30 F||Gavage||5d – no deaths||No deaths|
|Rat||73 M||IV||100||1-Rat died @ 100 mg/kg; cholesterol – elevated, but acceptable|
|73 F||50-300||Infusion||(LD10)||15 days|
|8 F||10-30||Bolus||>30.0||10 days||No deaths|
|4 F||10-30||Bolus||>30.0||10 days||No deaths|
Based on the conditions and findings of this study, the intravenous LD10 for
3.4. Sub-chronic oral mouse toxicity (Table 1)
A study was conducted in groups of 10 male/10 female mice per dose. The study evaluated
3.5. Acute single dose intravenous studies in rats (Table 1)
A rat study was conducted with the objectives to evaluate and characterize acute toxicity, maximum tolerated dose (MTD), and evaluate pharmacology (including pharmacokinetic parameters) of
Initially, a dose range finding
The main study phase consisted of a control group (10-M/10-F) that each received the vehicle only and three groups (10-M/10-F) that each received a single IV infusion of
The rats were divided into 2-groups that were euthanized either on Day-2 or Day-15. Complete macroscopic/microscopic examinations and complete clinical chemistry, hematology, coagulation studies and urinalysis were completed on all animals.
There were no meaningful hematological effects noted. On Day-2, erythrocytes, hemoglobin, and hematocrit tended to be higher in the 300 mg/kg/dosed group. These changes were most likely a result of fluid imbalances relative to reduced water intake. Monocytes were increased in both sexes at 200 and 300 mg/kg/dose and lymphocytes were decreased in males at 300 mg/kg/dose. Neutrophils were elevated in all groups on Days-2 and 15 and were attributed to stress and/or route of administration. All other changes were resolved by Day-15 and all values returned to normal pre-drug limits.
There were no test article-related effects on coagulation parameters or on urine analysis values.
The most significant findings were
|Cholesterol||2||50/ 49.4||106.4a /77.8||206.4b/157.4 b||188b/165.6 b|
|15||26.6/51.8||32.3/51.2||35.6 a /40||37.8 b/49.6|
The triglycerides also increased in the 200 and 300 mg/kg groups on Day-2 post-dosing, which resolved by Day 15. The females demonstrated the most significant elevations in both triglycerides and LDL-cholesterol. On Day 15 the profiles for both total cholesterol and triglycerides had returned to WNL. Table 2 reviews cholesterol and triglyceride trends.
Alanine aminotransferase (ALT) in males, and γ-glutamyl transferase (GGT), and alkaline phosphatase in females were minimally to mildly elevated at the 300 mg/kg/dose on Day 2. All of the findings noted on Day-2 had resolved by Day-15.
Various clinical signs reflecting treatment-related effects were noted, mostly lethargy that cleared. No behavioral alterations were noted.
Macroscopic/microscopic examinations revealed increased sizes of the livers and spleens. On Day-2, the 200 and 300 mg/kg groups possessed vacuolated macrophages in the liver (Kupffer cells) and spleen. By Day-15 the macrophages contained smaller oil aggregates and clusters within hepatic sinusoids. The findings in both the spleen and the liver showed trends toward resolution by Day-15 with both biliary hyperplasia in the liver and splenic focal necrosis resolving; vacuoles were smaller and cell cytoplasm had a more eosinophilic tint seen in both livers and spleens, albeit the vacuoles still expanded the cytoplasm of the cells. The latter changes are artifacts resulting from the extraction of drug/lipids/cholesterol from hepatic/splenic macrophages during fixation/preparation of tissues for microscopic examination. The changes seen on Day-2 in the spleens and the livers trended towards resolution by Day-15. Controls (vehicle alone) did not demonstrate the above changes.
Transitory changes in the hepatic profile are considered 2o to stasis of
Although the above findings resolved by Day-15, they must be considered adverse – based on the degree of elevation in triglycerides (3-fold) and LDL-cholesterol (30-fold in females) seen in some groups. The control group received the vehicle alone – soybean oil and egg yolk lecithin – both rich in triglycerides and did not demonstrate abnormal lipid profiles or the liver/spleen changes.
Based on the conditions and findings of this study, the intravenous LD10 of
3.6. Acute dog IV toxicity (Tables 1, 3)
A single IV dose study was performed in adult Beagle dogs, which consisted of
No treatment related effects on survival, hematology, urinalysis, or macroscopic and microscopic evaluations were noted during the study. Numerous clinical signs reflecting treatment-related effects were noted in both sexes of all groups, including the control group, and exhibited no dose-dependent pattern, clearly suggesting that the effects were attributable to the 0.3% Klucel+1.92% Tween 80 vehicle rather than
|IV once||0||2M||2 F||No deaths|
|10||2 M||2 F||No deaths|
|20||2 M||2 F||No deaths|
|30||2 M||2 F||No deaths|
There were marked elevations of alanine aminotransferase (ALT), and sorbitol dehydrogenase (SDH) in males in both controls and all treated animals on Day 2. Increased aspartate aminotransferase (AST) was observed in one (1) male with remarkably increased values for AST, ALT and SDH. The other male in this group exhibited only mildly increased values on Day 2 for these parameters. Similar trends were noted in females. ALT and SDH were highest on Day 2 for all dose levels, including controls. This acute and transient effect on liver enzymes exhibited no dose-dependent pattern and attributable to the vehicle. The latter finding was not observed to this extent in the rat study which used a different vehicle. The above was verified in a second group of 4M/4F.
The dog hematological studies confirmed that the 10-30 mg/kg doses do not produce myelosuppression, thrombocytopenia or anemia.
Treatment-related neurotoxicity was not observed following the single IV bolus administration of
3.7. Summary Median Lethal Dose (Single Dose)
Table 1 summarizes the toxic effects of single IV dose administrations of
3.8. Acute rat behavioral studies
Rats in groups (5-females) received
|Vehicle (Controls)||0.5 mL||Alert; normal behavior||Alert; normal behavior||Alert; normal behavior|
|Mk-801 (control)||0.05 once||Lethargic||Lethargic||Lethargic|
|5-FU (Chemo Control)||78||Lethargy, eyes closed||More alert||Normal behavior|
|400-800 once||Eyes closed; spastic; lethargy||Less lethargy; spastic||Normal behavior|
|350-600 once||No acute toxicity; no spasms||Normal behavior||Normal behavior|
|100 & 300 once||No acute toxicity; no spasms||Normal behavior||Normal behavior|
Impaired learning behavior has not been observed for
A control memory impairment agent, MK-801, is included to demonstrate complete impalement. In contrast,
|Agent||Dose (mg/kg/ dose||#Rats||Schedule||1 Hr||2 Hr||3 Hr||20 Hr||1Hr→||2Hr→||3Hr→||1 Hr →|
|2 Hr||3 Hr||20 Hr|
|Fold ↑ Time (sec)|
|MK-801||0.05||3||QD X 1||0**||0**||0**||0**||0||0||0||0||0|
|5-FU||78||6||QD X 1||1.7||1.3||1.7||8.2||1.0||0.6||1.3||0.7||-13.5|
|3||400||6||QD X 1||0.8||1.3||0.6||1.2||1.4||1.7||0.6||1.0||-1.2|
|2||135||6||QD X 1||6.7||1.5||1.7||2.4||4.0||1.0||1.0||3.5||42.1|
|1||100||6||QD X 1||3.2||1.3||1.8||3.8||3.2||0.6||1.5||2.4||23.6|
3.9. Brain/tumor penetration
Adult male mice
The intact frozen brains (~1.3 g) were coronal sliced into three sections in a mouse brain blocker (Kopf). The encapsulated gliomas were easily identified and separated readily from normal brain tissue with a scalpel and using microscopic ‘touch finger printing’ – separation verified. The tumor tissues were weighed, pooled and homogenized in 10 mL saline at 5OC. This process was repeated for the normal brain tissue. The cold homogenates were separately extracted with 10 mL dichloromethane, the organic layer separated and evaporated to dryness.
The residues were dissolved in dichloromethane and underwent preparative TLC on silica gel plates (Sigma-Aldrich, Milwaukee, WI) with a mobile phase – hexane/dichloromethane:10/30. The respective spots for
3.10. Normal brain penetration
Adult male rats [Crl: CD1(ICR) BR] (325-350 g wt) in groups of 5 animals were dosed intraperitoneally with 50 mg/kg of DM-CHOC-PEN in 0.3% Klucel/Tween80/saline daily x 2 days. On the 3nd day the rats were sacrificed and the intact brains removed (~1.9 g) and each homogenized in 10 mL saline at 5OC. To the cold homogenates, 20 mL dichloromethane was added and shaken for 30 minutes. The organic layer was removed and evaporated to dryness under vacuum at room temperature. The residues were dissolved in 1 mL of tetrahydrofuran and 100 μL chromatographed on silica gel plates with hexane:dichloromethane (1:1) as solvent. DM-CHOC-PEN was identified at Rf 0.74 with an additional spot – Rf 0.51. All spots were cut out, extracted with THF and analyzed by HPLC. DM-CHOC-PEN and a polar metabolite were identified by HPLC (see below).
DM-CHOC-PEN was calculated to be present-100 ng/g (avg.) of whole brain. The more polar peak (Rf 0.51) was not DM-PEN and present at 20 ng/g of whole brain. 1H-NMR of the latter fraction identified a pair of peaks at δ 5.68 & δ 5.75-consistent with loss of a methylene chlorine and binding to an NH-group, possible adduct. The material possessed a cholesteryl carbonate moiety. Normal brain tissue was used as a control.
3.11. Pharmacokinetic studies in rats and dogs (Table 5)
Plasma concentration-time profiles for
The shifts noted in the bioavailability for
|10 mg/kg||Dog (4)||1.23 (Mean)||21.6 (Mean)||0.42 (Mean)||328.8 (Mean)|
|0.53 (SD)||16.00 (SD)||0.17 (SD)||221.2 (SD)|
|30 mg/kg||Dog (4)||0.63 (Mean)||18.7 (Mean)||1.12 (Mean)||346,8 (Mean)|
|0.09 (SD)||10.7 (SD)||0.09 (SD)||54.5 (SD)|
|100 mg/kg||Rat (5)||0.51 (Mean)||2.48 (Mean)||1.05 (Mean)||30.4 (Mean9.99|
|0.05 (SD)||0.8 (SD)||0.53 (SD)||(SD)|
|200 mg/kg||Rat (5)||0.25 (Mean)||6.94 (Mean)||3.46 (Mean)||16.9 (Mean)|
|0.1 (SD)||2.1 (SD)||0.46 (SD)||4.04 (SD)|
|300 mg/kg||Rat (5)||0.12 (Mean)||4.0 (Mean)||5.17 (Mean)||19.40 (Mean)|
|0.06 (SD)||1.2 (SD)||1.74 (SD)||9.13 (SD)|
3.12. P-glycoprotein (P-gp) transport (Table 6)
The results of incubating Rho in the presence or absence of Vpml and/or
Six reaction conditions (
The results summarized in Table 7 for the 3 sensitive cell lines are coherent: The rate of incorporation of Rho is lower when cells are treated by the mixture of Vpml and DM-CHOC-PEN or Vpml alone but not when the cells are treated with DM-CHOC-PEN alone (mean fluorescence intensity of 6 roughly the same for control cells). This is interpreted as meaning that DM-CHOC-PEN has no effect on the function of P-gp transport.
The rationale for the pre-clinical development of
We report here the results of acute toxicity and pharmacology studies with single intravenous injections of
The IV LD10 single-dose value for mice (sexes combined) was calculated as 139 mg/m2. The mouse study generally displayed a typical dose-response effect (with the exception of one death at 50 mg/kg), with
A sub-chronic oral mouse toxicity study was conducted at MPI Research, Mattawan, MI, under GLP conditions in male/female mice. The study evaluated
Adult rats were treated once with single IV infusions of
The most significant abnormal findings were the statistically increased plasma values for cholesterol and triglycerides in the 200 and 300 mg/kg treated groups. LDL-cholesterol was significantly elevated in females – increased from 5.4 to 142 and 156 mg/dL for the 200 and 300 mg/kg groups, resp. This elevation is significant and considered a SLT (CTEP.v4). The triglycerides were increased by 4-fold in the 300 mg/kg group females, however, they return to normal values by Day-15. Hepatic and splenic deposits of fats were also noted on gross and microscopic examinations which cleared by Day-15.
Cholesterol is released during metabolism of
Although the above cholesterol and triglyceride findings resolved by Day-15, they must be considered adverse – triglycerides (3-fold) and LDL-cholesterol (30-fold in females). The control group received the vehicle alone – soybean oil and egg yolk lecithin – both rich in triglycerides and did not demonstrate abnormal lipid profiles.
Alanine aminotransferase (ALT) in males, and γ-glutamyl transferase (GGT), and alkaline phosphatase in females were minimally to mildly elevated in the 300 mg/kg group on Day-2. All of these findings on Day-2 resolved by Day-15. Transient elevations in transaminases were considered to be 2o to hepatic clearance of the drug. Neither gross nor microscopic evidence of toxicity (other than hepatic cysts) was noted at autopsies, including CNS. Table 7 compares calculated starting therapeutic doses for humans .
A single IV dose administration study was performed in adult beagle dogs employing single doses (10 – 30 mg/kg) of
No hematological deficiencies were noted in any group. Drug-related neurotoxicity was not observed. This was confirmed by second opinion (RT), who conducted silver stains and confirmed MPI’s observation that there were no microscopic pathological CNS changes present in the brains of dogs treated with
Based on the conditions and findings of this study, a single bolus intravenous injection of
Pharmacokinetic studies were conducted in two species – rats and dogs. Parameters were obtained from Gauss Newton algorithm modeling . The values are compared in Table 6. No statistical differences between male and female rats were noticed. The differences in half life can be explained in reference to administration routes – dog-IV bolus
In the learning/cognitive screening study, rats treated with
The drug penetrated human glioblastoma tumor tissue growing IC in mice, with none detectable in the normal tissue. This only reinforced our interest in using the drug to treat patients with cancers involving the CNS.
Thus, preclinical studies, conducted under GLP guidelines are presented as support for
|DM-CHOC-PEN||Mouse||136 mg/kg/d||39 mg/m2/d (10% of LD10)|
|DM-CHOC-PEN||Rat||100 mg/kg/d||60 mg/m2 (10% of LD10)|
|DM-CHOC-PEN||Dog||>30 mg/kg/d||>100 mg/m2/d**|
Supported by NCI/SBIR grants – 5R44CA85021 and 1R43CA132257
Morgan LR, Struck RF, Waud WR, LeBlanc B, Rodgers AH, Jursic BS. Carbonate and carbamate derivatives of 4-demethylpenclomedine as novel anticancer agents. Cancer Chemotherapy and Pharmacology, 64: 829-836, 2009.
Morgan LR, Struck RF, Waud WR, Jursic BS, Serota D, Papagiannis C, Rodgers AH. Carbonate and carbamate derivatives of 4-demethylpenclomedine as novel anticancer agents. Molecular Targets and Cancer Therapeutics, 23, 231, 2007.
Morgan LR, Struck RF, Rodgers AH, Serota DG. Preclinical Toxicity of 4-Demethyl-4-cholesteryloxylcarbonylpenclomedine (DM-CHOC-PEN). Proc. Amer. Assoc. Res., 48: abst. 5614, 2007.
Waud WR, Tiwari A, Schmid SM, Shih T-W, Strong JM, Hartman NR, O’Reilly S,
Struck RF. 4-Demethylpenclomedine, an antitumor-active, potentially non-neurotoxic metabolite of penclomedine. Cancer Res, 57:815-817, 1997.
Berlin J, Stewart JA, Storer B, Tutsch KD, Arzoomanian RZ, Alberti D, Feierabend C, Simon K, Wilding G (1998) Phase I clinical and pharmacokinetic trial of penclomedine using a novel, two stage trial design for patients with advanced malignancy. J Clin Oncol, 16:1142-1149.
O’Reilly S, Grochow L, Donehower RC, Bowling K, Chen TL, Hartman N, Struck R, Rowinsky EK. Phase I and pharmacologic studies of penclomedine, a novel alkylating agent in patients with solid tumors. J Clin Oncol, 15: 1974-1984, 1997.
Jodrell DI, Bowman A, Stewart M, Dunlop N, French R, MacLellan A, Cummings J, Smyth JF. Dose-limiting neurotoxicity in a phase I study of penclomedine (NSC 388720, CRC 88-04), a synthetic alpha-picoline derivative, administered intravenously. Brit J Cancer, 77: 808-811, 1998.
Berlin J, Wahamaki A, Tutsch KD, Alberti D, Feierabend C, Binger, K, Arzoomanian, RZ, Volkman, J, Karca J, Mornocha, R, Stewart J, Wilding G. Phase I, pharmacokinetic and bioavailability study of oral penclomedine administered daily x 5 every four weeks. Proc. Amer. Assoc. Cancer Res, 40: 92, 1999.
O’Reilly S, O’Hearn E, Rowinsky EK, Struck RF, Molliver ME. Neuroanatomic studies of the cerebellar effects of penclomedine. Proc. Amer. Assoc. Cancer Res, 37:
O’Reilly S, O’ Hearn E, Struck RF, Rowinsky EK, Molliver ME. The alkylating agent penclomedine induces degeneration of purkinje cells in the rat cerebellum. Invest New Drugs, 21:269, 2003.
Morgan LR, Struck RF, Waud W, Jursic BS, Serota DG, Papagiannis C, Rodgers AH, Thornton, ME, GS LaHoste. Comparative pharmacokinetics and intermediary metabolism of 4-demethyl-4-cholesteryl-oxycarbonylpenclomedine (DM-CHOC-PEN), EORTC/ AACR/NCI (Berlin), Abst 871, 2010.
Pletsas D, Wheelhouse RT, Pletsa V, Nicolaou A, Jenkins TC, Bibby MC. Kytopoulas SA. Polar, functionalized guanine-O6 derivatives resistant to repair by O6-alkylguanine-DNA alkyltransferase: implications for the design of DNA-modifying drugs. Eur. J. Med Chem., 11: 1-10, 2006.
Morris, R. Development of water-maze procedure for studying spatial learning in the rat. J. Neuroscience Methods, 11: 47-60, 1984.
Saab, BJ, Saab, AMP, Roder, JC. Statistical and theoretical considerations for the platform re-location water maze. J. Neurosci. Meth. 198:44-52, 2011.
Weiner RS, Friedlander P, Gordon C, Mahmood T, Ware ML, Bastian G, Rodgers AH, Urien S, Morgan, LR. A first-in-humans: phase I clinical for 4-Demethyl-4-cholesteryl-oxycarbonylpenclomedine (DM-CHOC-PEN). Proc Amer. Assoc. Cancer Res., 54, 1173, 2013.
Dietrich J, Han R, Yang Y, Mayer-Proschel M, Noble M. CNS progenitor cells and oligodendrocytes are targets of chemotherapeutics in vitroand in vivo. J Biol. 5:22-32, 2006.
Konat GW, Kraszpulski M, James I, Zhang HT, Abraham J. Cognitive dysfunction induced by chronic administration of common cancer chemotherapeutics in rats. Metabolic Brain Diseases, 23:325–333, 2008.
Common chemo drug can cause memory problems. In: the Times-Picayune News Paper, New Orleans, LA, B-3, pg 3, 2008 and ref cited.
Morgan LR. A Phase I Trial: Safety and tolerance of 4-demethyl-4-cholesteryloxy-carbonylpenclomedine (DM-CHOC-PEN)-a potential neuro-alkylating agent for glioblastoma and metastatic cancers of the central nervous system, FDA IND 68,876, 2010.
Morgan LR, Rodgers AH, Bastian G, Benes E, Waud WR, Jursic BS, Struck RF, LaHoste G, Thornton M, Luttrell M, Stevens E, Thompson R. Comparative preclinical pharmacology and toxicology for 4-demethyl-4-cholesteryloxycarbonylpenclomedine (DM-CHOC-PEN): A potential neuro-alkylating agent for glioblastoma and metastatic cancers of the central nervous system. ECCO in European J Cancer, Abst. 57, 2011.
Lund-Pero, M, Jeppson, B, Arneko-Nobin, B, Sjogren, HO, Holmgren, K, Pero, RVV. Non-specific steroidal esterase activity and distribution in human and other mammalian tissue. Clin. Chim Acta, 224: 9-20, 1994.