Relationship between Dihydropyrimidine Dehydrogenase Activity and Plasma5-FluorouraciI Levels with Evidence for Orcadian Variation of Enzyme
Activity and Plasma Drug Levels in Cancer Patients Receiving5-Fluorouracil by Protracted Continuous Infusion1
Barry E. Harris, Ruiling Song, Seng-jaw Soong, and Robert B. Diasio2
Department of Pharmacology and Comprehensive Cancer Center. Division of Clinical Pharmacology' [B. E. H., R. S., K. B. D.J. and Comprehensive Cancer Center,Bimlalistics I 'nit ¡S-j.SJ, University of Alabama at Birmingham, Birmingham, Alabama 35294
ABSTRACT
The activity of dihydropyrimidine dehydrogenase (DPD) in peripheralblood mononudear cells and plasma concentration of 5-fluorouracil(FUra) were simultaneously determined in cancer patients receiving FUraby protracted continuous infusion (300 mg/m!/day). Blood samples were
drawn every 3 h over 24-h period and the resulting DPD and FUra valuesanalyzed for circadian periodicity. In the seven patients studied, a circa-dian rhythm of DPD activity was observed (P < 0.00001, Cosinoranalysis) with the peak of activity at 1 a.m. (0.197 ±0.007 nmol/min/mg) and the trough at a 1 p.m. (0.113 ±0.007 nmol/min/mg). In addition,a circadian rhythm was observed for the plasma concentrations of FUraobtained over a 24-h period (/' < 0.00001, Cosinor analysis) with peak
values (27.4 ±1.3 ng/ml) occurring at 11 a.m. and trough values (5.6 ±1.3 ng/ml) occurring at 11 p.m. The ratio of the maximum concentrationof Ilia to the minimum concentration observed was almost 5-fold. Thisstudy demonstrates a circadian variation of DPD activity in humanperipheral blood mononuclear cells and a circadian variation of FUraplasma levels in patients receiving FUra by protracted continuous infusion. An inverse relationship between the circadian patterns of DPDactivity and FUra plasma levels was also noted, suggesting that anassociation may exist between DPD activity and FUra plasma concentration. Further evidence of an association between DPD activity in peripheral blood mononuclear cells and plasma FUra concentration was demonstrated by a linear relationship between the two parameters in allpatients (r = —¿�0.627)and within individual patients (—0.978< r <—¿�0.742).With the recent advent of programmable pumps, information
on the circadian pattern of FUra and/or DPD may be useful in planningcontinuous infusion schedules in order that optimal plasma drug concentration may be maintained over a 24-h cycle, thereby enhancing thetherapeutic efficacy of FUra administered by continuous infusion.
INTRODUCTION
Several studies have shown that plasma drug levels varysignificantly during continuous infusion of FUra3 (1-4). In
addition. Petit et al. (3) reported that the variability followed apredictable pattern, suggesting that there was a circadian variation in plasma FUra levels during continuous infusion at aconstant rate. The biochemical mechanism for the observedvariability and/or circadian periodicity of FUra during continuous infusion has not been determined.
DPD is the initial enzyme of pyrimidine catabolism with anestimated 80% (or greater) of an administered dose of FUrabeing rapidly degraded by this route (5). The importance ofcatabolism and, particularly, DPD in FUra chemotherapy hasbeen demonstrated in studies with competitive inhibitors (6, 7)
Received 6/19/89: revised 9/26/89; accepted 10/3/89.The costs of publication of this article were defrayed in part by the payment
of page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1Supported in part by USPHS Grants CA-40530 and CA-13148.: To whom requests for reprints should be addressed: P. O. Box 600 Volker
Mall. Division of Clinical Pharmacology, University of Alabama at Birmingham,DAB Station. Birmingham. Alabama 35294.
and in a recent report of a patient with a complete deficiencyof DPD activity (8). These studies indicate that catabolism hasan important role in determining the availability of FUra foranabolism which, in turn, determines the cytotoxic effects ofFUra. We have previously demonstrated that DPD exhibits acircadian periodicity both in rat liver (9) and human peripheralblood mononuclear cells (10).
The purpose of this study was to simultaneously determineDPD activity in peripheral blood mononuclear cells and plasmaFUra concentrations in patients receiving low dose "flat" con
tinuous infusion of FUra to determine if either exhibited acircadian pattern and if any relationship existed between DPDactivity and plasma FUra levels.
MATERIALS AND METHODS
Patient Characteristics. Seven patients (five men, two women) participated in this study. The age of the patients ranged from 48 to 69 years.All seven patients had biopsy-proven gastrointestinal malignancies (e.g.,gastric, colorectal, or pancreatic) and were receiving protracted continuous infusion of FUra. Characteristics of the patients including hepaticand renal function are listed in Table 1.
Protocol. All seven patients received 300 mg FUra/m2/day adminis
tered by an ambulatory pump (Cormed II, Cormed, Inc., Murray, NJ;Infumed 300. Medfusion Systems, Inc., Norcross, GA) via a Hickmancatheter placed in the lower portion of the external jugular vein. Pumpswere pretested and no difficulties were observed during the study.Patients received continuous infusion FUra for a minimum of 2 weeksprior to the study. For blood sampling, patients were admitted to theGeneral Clinical Research Center of the University of Alabama Hospital on an institutionally approved protocol with each patient givinginformed consent. Every effort was made to maintain the patients on anormal light/dark cycle. Time of awakening was between 6 and 7 a.m.,and time of retiring was between 10 and 11 p.m. Blood samples weretaken at the following times over a 2-day period: Day 1-9 a.m., 12p.m., 6 p.m., 12 a.m.; Day 2-3 a.m., 6 a.m., 3 p.m., and 9 p.m.
Collection of Blood Samples. A heparin-lock was placed in a peripheral arm vein so that blood samples could be drawn with minimaldiscomfort and disturbance to the patient during sampling. Two bloodsamples were taken at each of the time points described above. Onesample was used for the DPD assay and the other for the FUra assay.
Assay of DPD Activity. Blood samples (25 ml) were drawn from theheparin-lock into a 30-ml syringe containing 5 ml of heparin (1000units/ml) and immediately added to 15 ml of Hanks' balanced salt
solution in a 50-ml centrifuge tube. Plasma was removed by centrifu-gation and the pelleted cells were loaded onto a Ficoll-Hypaque discontinuous gradient to separate mature erythroid blood cells from peripheral blood mononuclear cells. The peripheral blood mononuclear cellfraction was removed and mixed with 1 ml of 135 mivi sodium phosphate buffer (pH 7.5). The cells were placed in an ice bath and thenlysed by sonicating five times for periods of 10 s. After removal ofdebris by centrifugation (20,000 x g for 3 min) the supernatant wasassayed for DPD using a modification (8) of a method originallydescribed by Marsh and Perry (11). The supernatant was incubated at37°Cfor varying intervals up to 30 min in the presence of 250 /¿M
a Normal adult range. 0-1.2 mg/dl.* Normal adult range, 25-115 units/liter.f Normal adult range. 0.6-1.6 mg/dl.d I = Infumed 300. Medfusion Systems, Inc.; C : Cormed II. Cormed. Inc.
NADPH and 20 //M |6-'H)FUra (3.5 ßd/nmo\). At the end of the
incubation time an aliquot (300 »I)was added to an equal volume ofice-cold ethanol and then (within 3 h) analyzed for the presence ofFUra and FUra catabolites by a HPLC method described previously( 12). Protein concentration was determined by the method of Lowry etal. (13). The enzyme activity was expressed as nanomoles of totalcatabolites formed/min/mg protein.
Assay of Fl'ra Plasma Concentration. In addition. 10-12 ml of blood
were drawn from the heparin lock at each of the time points intoheparinized vacutainer tubes and quickly centrifuged. Plasma was removed and stored at -20°C until FUra concentration was determined
by HPLC (w¡thin7 days following sampling). Quantitation of FUra byHPLC followed the technique of Buckpitt and Boyd (14) with a limitof sensitivity of approximately 2 ng/ml. A separate set of FUra plasmastandards was used for each set of assays with minimal interpatientvariation.
Statistical Analysis. The data obtained from the DPD and FUraassays were analyzed by the "Cosinor" method (15). Due to the large
interpaticnt variability of observed values for both DPD activity andFUra plasma concentration, comparisons between patients were carriedout by expressing the data as a percentage of the 24-h mean for thatpatient. The values obtained were fitted to a cosine wave by regressionanalysis utilizing the method of least squares (16). Three parameterswere quantitated in this analysis. These parameters include the mesor(i.e.. the rhythm-adjusted mean), the amplitude (i.e., maximum orminimum value from the mean), and the acrophase (i.e., time ofmaximum or minimum value from a given phase of reference). Inaddition, relationship of DPD activities and FUra concentrations wasevaluated by linear regression analysis (17).
RESULTS
Variation of DPD Activity. In the seven patients studied, therewas a circadian variation in DPD activity of each patient (Table2). Fig. \A illustrates the pattern of DPD activity from one ofthe seven patients (patient 6, B. B.) studied with DPD activityexpressed as nmole catabolites formed per minute per milligramprotein. Most of the patients had a peak of DPD activitybetween 10 p.m. and 4 a.m., while one patient (patient 3, W.J.) had a peak at 11 a.m. (Table 2). Due to the interpatientvariability in DPD activity and the time of maximum or minimum activity, the data for each patient was normalized in orderto establish an overall pattern for all patients. The data wasnormalized in two ways: (a) DPD activity was expressed as apercentage of the 24-h mean (i.e., mesor) and (b) the data foreach patient was adjusted along the time scale to a commonreference point (i.e., the DPD activity peak). The normalizeddata was analyzed by Cosinor analysis and the overall valuesfor all patients are shown in Table 3. The peak of DPD activityfor the entire group of patients was at approximately 1 a.m.(0.197 ±0.007 nmol/min/mg) and the trough at approximately1 p.m. (0.113 ±0.007 nmol/min/mg). For the normalized data(% 24-h mean) maximum DPD activity exceeded minimum
activity by approximately twofold (Table 3).Variation of Plasma FUra Concentration. The plasma concen
tration of FUra (ng/ml) varied significantly both within a singlepatient and between patients. All seven patients exhibited asignificant circadian rhythm with respect to plasma FUra concentration over a 24-h period (Table 2). Fig. \B represents theplasma FUra concentration (ng/ml) in one of the seven patients(patient 6, B. B.) in this study. As with the DPD activity, FUraplasma concentrations for each patient were normalized to apercentage of the 24-h mean (i.e., mesor) and adjusted alongthe time scale to a common reference point, the DPD activitypeak for that patient. The Cosinor analysis of these values isshown in Table 3. The peak FUra concentration for the entiregroup of seven patients was at approximately 11 a.m. (27.4 ±1.3 ng/ml) and the trough at approximately 11 p.m. (5.6 ±1.3ng/ml). For the normalized data (% 24-h mean) maximumFUra concentration exceeded minimum concentration by almost 5-fold (Table 3).
Correlation between DPD Activity and FUra Concentration.As shown in Fig. 2, as DPD activity decreases, FUra plasmaconcentration appears to increase and vice versa. In support ofthis observation, comparison of DPD activity with FUra concentration for any sample from any patient yielded an overallcorrelation between the two parameters (Fig. 3, r = -0.627). It
is interesting to note that if consideration was given to individual patients only, the correlation of DPD activity and FUraplasma concentrations would be much better (-0.978 < r <—¿�0.742)than the overall correlation for all patients combined(r = —¿�0.627).Since it is clear that interpatient variability is
greater than intrapatient variability, the mean DPD activity foreach patient was examined with the corresponding mean valuefor the FUra concentration in the same patient. The correlationof these two parameters (Fig. 3, inset) further suggests thatthere is an inverse relationship (r = -0.973) between DPD
activity in peripheral blood mononuclear cells and plasma FUraconcentrations in patients receiving FUra by protracted continuous infusion.
DISCUSSION
The present study demonstrates that there was a significantvariation in the plasma level of FUra during protracted continuous infusion in cancer patients. This observation has also beenmade by others (1-4) although no apparent mechanism hasbeen elucidated. In an attempt to determine the biochemicalmechanism, we examined FUra catabolism in patients receivingFUra by continuous infusion and observed a significant variation of DPD activity in peripheral blood mononuclear cells.Although it has been apparent for some time that metabolismis very important in determining the effectiveness of the fluo-
DIHYDROPYRIMIDINE DEHVDROGENASK AND PLASMA FUra LEVELS
Table 2 Summary of lìP1) activity in peripheral hliwtl mononuclear cells and Fl'ra plasma concentration* in patients receiving Fl 'ra hy protracted continuous infusion
Fig. I. DPD activity in peripheral blood mononuclear cells (A) and FUraplasma concentration (if) in one (patient 6. B. B.) of the seven patients receivingprotracted continuous infusion of FUra (300 mg/nr/day). The Cosinor model(16. 17) was used to fit this data to a cosine wave ( ).
ropyrimidines, most of the attention has been focused onanabolism rather than catabolism. The paucity of informationon catabolism is surprising, particularly since it has been estimated that more than 80% of an administered dose of FUra israpidly catabolized (5) suggesting that catabolism determinesthe availability of FUra for anabolism (18). In addition, severalrecent studies have emphasized the importance of catabolism,particularly DPD, in fluoropyrimidine chemotherapy (6-8, 18).
Two potential sources of error that could account for theobserved variation of DPD and FUra in this study are: (a) avariation in the infusion rate by the ambulatory pump or (b) avariation in the assays for DPD or FUra. Mechanical variationsof infusion rates with either pump are unlikely to exceed 5%(pump specifications provided by Medfusion, Inc. and Cormed,Inc.). A variation in either assay is also unlikely since both theDPD assay (11) and the FUra assay (14) utilized methods withhigh sensitivity and reproducibility. Further, the 2- and 5-foldvariation that was observed in DPD activity and plasma FUralevels, respectively, is not likely to result from experimentalerror in either assay.
In the present study, we report the simultaneous analysis ofDPD activity and plasma FUra concentration and demonstratea circadian variation of DPD activity (P < 0.00001, Cosinoranalysis) in human peripheral blood mononuclear cells and acircadian variation in the plasma FUra level (P < 0.00001,Cosinor analysis) in patients receiving FUra by protractedcontinuous infusion at 300 mg/nr/day. Although a circadianpattern is evident for both DPD activity and plasma FUralevels, a general trend that is characteristic of all patients is notdemonstrated. The time of maximum or minimum values foreach patient varied as much as 7 h for six of the seven patients(9:30 p.m. to 4:30 a.m.) with one patient (patient 3, W. J.)demonstrating times of maximum and minimum values thatwere up to 12 h out of phase with the other patients. It isinteresting to note that this patient (patient 3, W. J.) hadworked a night shift for several years. While a circadian patternwas evident for each of the seven patients, the striking inter-patient variation in the time of maximum/minimum values forDPD and FUra suggests that caution should be used in developing generalized time-programmed infusion schedules as pre
viously suggested (19). However, for individual patients, monitoring DPD activity and/or plasma FUra levels of individual
DIHYDROPYRIMIDINE DEHYDROGENASE AND PLASMA FUra LEVELS
Table 3 Rhylhmometric summary of single Cosinor analysis ofDPD activity and FUra concentration in seven patients receiving protracted continuous infusion of FUra(300mg/nr/dayf
Mcsor ±SE*
(%)Amplitude ±SE(%)DPD
activity100.1 + 2.6 30.4±3.7FUra
plasma concentration99.1 ±4.4 65.0 ±6.3Maximum
±SEC(%)130.6
±4.5
164.1 ±7.7Time
ofmaximum''
(h)1.0
±0.7
11.4 ±0.5Minimum
±SEr(%)69.7
±4.5
34.0 ±7.7Time
ofminimum''
(h)13.0
±0.7
23.4 ±0.5Maximum/
minimum1.94.8R*0.930
0.955P<0.00001<0.00001" Cosinor analysis was carried out on normalized data expressed as %24-h mean.* Mesor = Rhythm-adjusted mean.c Maximum/minimum = mesor ±amplitude.äTime is expressed on a 24-h scale.
200 -
150 -
100 -
50 -
0.5
1 2 18 24
TIME FOLLOWING DPD PEAK (MRS)Fig. 2. Circadian variation of OPD activity (•)in peripheral blood mononu-
clear cells and FUra plasma concentration (A) in seven patients receiving protracted continuous infusion of FUra (300 mg/nv/day). Each value represents themean ±SD. Statistical significance of circadian periodicity was determined byCosinor analysis (P < 0.00001).
patients may be useful in planning FUra continuous infusionchemotherapy.
The present study also demonstrates an apparent relationshipbetween DPD activity in peripheral blood mononuclear cellsand plasma FUra levels. Comparison of DPD activity and FUraplasma concentrations for all patients at all sampling timesyielded a moderate correlation (r = -0.627, Fig. 3). However,
if the comparison of DPD activity and FUra plasma levels waslimited to those values within a single patient the correlationwas much better (-0.978 <r< -0.742, Fig. 3) indicating that
interpatient variation was responsible for the weaker overallcorrelation and not intrapatient variation. Therefore, althoughdrug concentrations cannot be accurately predicted for all patients treated with a specific regimen, FUra plasma concentrations potentially can be predicted within individual patientsfrom representative samples taken at various times of the day.Further support for the relationship of these two parameters issuggested by the correlation between mean DPD activity andmean FUra plasma concentration in the seven patients in thisstudy (Fig. 3, inset).
o>E
0.3
S 0.2O
oo.Q
0.1
0.0
l .
0.4
> t» 0.3PE
DE 0.2Q «
5¡0.1W _s
0.00 10 20
MEAN FUra CONC(ngrnl)
AP a +A A A
A
A oA.
*
1 O 20 30 40 SO
FUra CONCENTRATION (ng/ml)
Fig. 3. Comparison of simultaneously determined values of DPD activity inperipheral blood mononuclear cells and FUra plasma concentration in patientsreceiving protracted continuous infusion of FUra (300 mg/mVday). Six to eightsamples were taken at various times of the day from each of the seven patients (n= 52). Correlation was evaluated by linear regression fitted by the method of leastsquares (r = —¿�0.627).Values from individual patients also exhibited significantcorrelations (patient l (O), r = -0.742; patient 2 (•),r = -0.834; patient 3 (D),r = -0.978; patient 4 (A), r = -0.808; patient 5 (+), r = -0.809; patient 6 (A), r- -0.978; and patient 7 (•).r = -0.916). Inset, comparison of mean DPDactivity with the corresponding mean FUra plasma concentration in the samepatient in all seven patients (identified by patient number, i.e., 1-7). Correlationwas evaluated by linear regression fitted by the method of least squares (r
Clearance of FUra primarily results from catabolism withinthe liver (20), although extrahepatic sites may also be involved(21). DPD has previously been demonstrated to have the highestactivity in liver and peripheral blood mononuclear cells (22)with minimal activity in kidney, spleen, lung, colon, pancreas,breast tissue, and bone marrow cells (23). It has been suggestedthat the activity of numerous liver enzymes in drug metabolismexhibits circadian rhythms (24) leading to the hypothesis proposed here that a circadian variation in DPD might be responsible for the variation in plasma drug levels during constant-rate continuous infusion of FUra. The importance of DPD indetermining the clinical pharmacokinetics of FUra has previously been demonstrated in a patient with a complete deficiencyof DPD activity (8).
Although the liver is the primary site of pyrimidine catabolism leading to the degradation of FUra, it is not feasible todirectly study DPD activity in liver tissue of patients. Therefore,in the present study, DPD activity was examined in peripheral
DIHYDROPYRIMIDINE DEHVDROGENASE AND PLASMA FUra LEVELS
blood mononuclear cells with the assumption that the DPDpatterns and/or activity in these cells may be related to that ofDPD in liver. This assumption is supported by previous studiesfrom our laboratory which demonstrate that DPD exhibits acircadian pattern in rat liver (9) as well as in human peripheralblood mononuclear cells (10).
In summary, the present study demonstrates a highly significant circadian rhythm in both DPD activity in peripheral bloodmononuclear cells and plasma FUra levels during protractedcontinuous infusion of FUra with an apparent inverse relationship between these two parameters. Since FUra plasma levelsare associated with drug-induced side effects (25), knowledgeof the circadian variation of plasma drug levels in individualpatients may allow more precise planning of time-modifiedschedules of drug delivery which, in turn, may result in increased effectiveness of continuous infusion regimens. The recent availability of programmable infusion pumps which permitmodulation of the infusion rate on a 24-h scale has greatlyincreased the ease with which this information may be utilizedin planning infusion schedules. Although further evaluation ofboth host and tumor tissues is needed, this study suggests thatDPD may be a major determinant of the periodicity of FUraplasma levels when this drug is administered by protractedcontinuous infusion and should be considered in planning infusion schedules.
REFERENCES1. Erlichman, ( '.. Fine, S., and Elhakim, T. Plasma pharmacokinetics of SFU
given by continuous infusion with allopurinol. Cancer Treat. Rep.. 70: 903-904. 1986.
2. Kawai, M., Rosenfeld. J.. McCulloch, P., and Hillcoat, B. L. Blood levels of5-fluorouracil during intravenous infusion. Br. J. Cancer. 36:346-347.1976.
3. Petit. E.. Milano. C., Levi. F., Thyss, A., Bailleul, F.. and Schneider, M.Circadian rhythm-varying plasma concentration of 5-fluorouracil during afive-day continuous venous infusion at a constant rate in cancer patients.Cancer Res.. 48: 1676-1679, 1988.
4. Poplin, I .. Chabot, G., Rutkowski, K., and Baker, L. Continuous infusion(CI), low-dose 5-fluorouracil (FUra): plasma concentrations. Proc. Am. As-soc. Cancer Res.. 29: 190. 1988.
5. Woodcock, T. M.. Martin, D. S., and Damin, L. E. M. Clinical trials withthymidine and fluorouracil: a phase 1and clinical pharmacological evaluation.Cancer (Phila.). 45: 1135-1143. 1980.
6. Martin. D. S.. Nayak. R., Sawyer. R. C. Stolti, R. L.. Young. C. W.,Woodcock. T., and Spiegelman. S. Enhancement of 5-fluorouracil chemotherapy with emphasis on the use of excess thymidine. Cancer Bull., JO: 219-224. 1978.
7. Wasternack, C. Degradation of pyrimidincs and pyrimidine analogs—pathways and mutual influences. Pharmacol. Ther., 8: 629-651. 1980.
8. Diasio, R. B., Beavers, T. L., and Carpenter, J. T. Familial deficiency ofdihydropyrimidine dehydrogenase: biochemical basis for familial pyrimidi-nemia and severe 5-fluorouracil-induced toxicity. J. Clin. Invest., 81:47-51,1988.
9. Harris, B. E., Song. R., He, Y-j.. Soong, S-j.. and Diasio. R. B. Circadianrhythm of rat liver dihydropyrimidine dehydrogenase: possible relevance tofluoropyrimidine chemotherapy. Biochem. Pharmacol.. 37: 4759-4762.1988.
10. Harris, B. E., Song, R., He, Y-j., and Diasio, R. B. Circadian periodicity ofdihydropyrimidine dehydrogenase (DPD) activity in humans: possible relevance to toxicity from fluoropyrimidine (FP) drugs with implications for FPinfusion by programmable pumps. Proc. Am. Soc. Clin. Oncol.. 7: 77, 1988.
11. Marsh, J. C., and Perry, S. The reduction of thymine by human leukocytes.Arch. Biochem. Biophys.. 104: 146-149, 1964.
12. Sommadossi, J. P., Gewirtz, D. A., Diasio. R. B.. Aubert. C.. Cano, J. P..and Goldman. 1. D. Rapid catabolism of 5-fluorouracil in freshly isolated rathepatocytes as analyzed by high performance liquid chromatography. J. Biol.Chem., J57: 8171-8176, 1982.
13. Lowry, O. H.. Rosebrough, N. J., Fair. A. L.. and Randall. R. J. Proteinmeasurement with the Folin phenol reagent. J. Biol. Chem., 193: 265-275,1951.
14. Buckpitt, A. R., and Boyd, M. R. A sensitive method for determination of 5-fluorouracil and 5-fluoro-2'-deoxyuridine in human plasma by high-pressureliquid chromatography. Anal. Biochem.. 106:432-437, 1980.
15. Halberg, F., Johnson, E. A., Nelson, W., Runge, W., and Sothern, R.Autorhythmometry procedures for physiologic self-measurements and theiranalysis. Physiol. Teacher, 1:1-11. 1972.
16. Reinberg. A., and Smolensky. M. H. Investigative methodology for chrono-biology. In: A. Reinberg and M. H. Smolensky (eds.). Biological Rhythmsand Medicine: Cellular, Metabolic, Physiopathologic. and PharmacologieAspects, pp. 23-46. New York: Springer-Verlag, 1983.
17. Draper, N. R., and Smith. H. Applied Regression Analysis. New York: JohnWiley & Sons. 1981.
18. Heggie, G. D., Sommadossi, J. P., Cross, D. S., Huster, W. J., and Diasio,R. B. Clinical pharmacokinetics of 5-fluorouracil and its metabolites inplasma, urine, and bile. Cancer Res., 47: 2203-2206, 1987.
19. Roemeling, R. v., and Hrushesky. W. J. M. Circadian pattern of continuousFUDR infusion reduces toxicities. In: J. E. Pauly and L. E. Scheving (eds.).Advances in Chronobiology, pp. 357-373. Boston: Alan R. Liss. Inc., 1987.
20. Ensminger. W. D., Rosowsky, A., Raso, V., Levin, D. C.. Glode, M., Come,S., Steele, G., and Frei, E., III. A clinical pharmacological evaluation ofhepatic arterial infusion of 5-fluoro-2'-deoxyuridine and 5-fluorouracil. Cancer Res.. 38: 3784-3792, 1978.
21. Myers, C. E. The pharmacology of fluoropyrimidines. Pharmacol. Rev., 33:1-15, 1981.
22. Naguib, F. M. N., el Kouni, M. H., and Cha, S. Enzymes of uracil catabolismin normal and neoplastic human tissues. Cancer Res., 45: 5405-5412, 1985.
23. Ho, D. H., Townsend, L.. Luna. M. A., and Bodey, G. P. Distribution andinhibition of dihydrouracil dehydrogenase activities in human tissues using5-fluorouracil as a substrate. Anticancer Res., 6: 781-784, 1986.
24. Radzialowski, F., and Bousquet, W. Daily rhythmic variation in hepatic drugmetabolism in rat and mouse. J. Pharmacol. Exp. Ther., 163:229-238,1968.
25. Seifen, P., Baker, L. H.. Reed. L. M., and Vaitkevicius, V. K. Comparisonof continuously infused 5-fluorouracil with bolus injection treatment ofpatients with colorectal adenocarcinoma. Cancer (Phila.). 36:123-128, 1975.
1990;50:197-201. Cancer Res Barry E. Harris, Ruiling Song, Seng-jaw Soong, et al. Continuous Infusionin Cancer Patients Receiving 5-Fluorouracil by Protracted Circadian Variation of Enzyme Activity and Plasma Drug LevelsActivity and Plasma 5-Fluorouracil Levels with Evidence for Relationship between Dihydropyrimidine Dehydrogenase
