[CANCER RESEARCH 33, 895-901, April 1973]

The Effect of Dose Interval on the Survival of L1210 LeukemicMice Treated with DNA Synthesis Inhibitors1G. L. Neil and E. R. Homan

With the technical assistance of T. E. Moxley, R. C. Manak, L. G. Gray and S. L. Kuentzel2

Cancer Research, The Upjohn Companv, Kalamazoo, Michigan 49001 /G. L. N.], and the National Cancer Institute, NIH, Bethesda, Maryland20014 ¡E.R.H.I

SUMMARY

The survival of L1210 leukemic C57BL/6 X DBA2 F! micehas been studied as a function of the time interval betweentwo i.p. doses of DNA synthesis inhibitors (1-0-D-arabino-

furanosylcytosine, hydroxyurea, sodium camptothecin, and5-hydroxy-2-formylpyridine thiosemicarbazone). In general,life-span increases that were observed with single doses ofthese agents (i.e., a time interval of zero) were consistent withthe cell kill expected (approximately 66 to 86%) for anS-phase-specific agent with a short effective contact time. Thetherapeutic effect appeared to be a function of the intervalbetween the two doses. When optimal time intervals betweentwo doses were used, cell kills in excess of 99% were achieved.In some cases, enhanced therapeutic ratios (increase inlife-span compared to host weight loss) were achieved byjudicious choice of the dose interval. It was possible in mostcases to correlate the optimal interval for maximal cell killwith the time required for the maximal recovery of the abilityof the peritoneal ascites LI210 cells to incorporate tritiatedthymidine into DNA after the initial dose of the agent. Theresults are discussed in terms of a simple model based oncell-cycle kinetic parameters.

eradicate the entire LI210 cell population with only 2optimally spaced doses of an S-phase-specific agent (i.e., anagent that kills cells only when the cells are exposed to itduring the DNA-synthetic phase of the cell cycle). An initialexposure to the agent should kill the S-phase cells of theasynchronous population. One could then wait an appropriatetime until the partially "synchronized" surviving cells (i.e.,

those in the GÌ, G2, and M phases during the initial exposure)had entered S phase and then "complete" the cell kill with a

2nd exposure. The studies reported below represent in part anattempt to test such a hypothesis.

In L1210 leukemic mice, the optimal interval for the 2-doseadministration of an S-phase-specific agent can, of course, bedetermined empirically. However, we also investigated thepossibility that changes in a pertinent biochemical parameter(i.e., total tumor-cell DNA synthesis) might correlate with thisoptimal dose interval. Such a correlation would prove of valuein predicting (prior to therapy) the optimal interval for drugadministration and might have important implications withrespect to treatment of human neoplasia.

Analogous studies with LI210 cells growing in culture havebeen performed, and those results are presented in acompanion paper (3).

INTRODUCTION

The choice of treatment schedules (i.e., the temporalrelationship between successive doses) can greatly influencethe results of chemotherapy of neoplastic disease in experimental animals. The work of Skipper et al. (19) and Kline etal. (9) with ara-C3 in the treatment of LI210 leukemic mice

represents a striking example of this.With L1210 leukemia in mice, the S (DNA-synthetic) phase

comprises over 50% of the total cell-cycle time (19, 21, 22).Provided that cell-cycle time and phase lengths are homogeneous, it should be possible (at least theoretically) to totally

'This study was supported in part by Contract PH43-NCI-68-1023with Chemotherapy, Drug Research and Development, National CancerInstitute, NIH, Bethesda, Md. 20014.

'All affiliated with Cancer Research, The Upjohn Company,Kalamazoo, Mich. 49001.

3The abbreviations used are: ara-C, l-/3-D-arabinofuranosy Icytosine(cytarabine, NSC 63878); HU, hydroxyurea (NSC 32065); 5-HP,5-hydroxy-2-formylpyridine thiosemicarbazone (NSC 107392); camptothecin, camptothecin sodium salt (NSC 100880); TdR-3H, tritiatedthymidine; PCA, perchloric acid.

Received August 29, 1972;accepted January 18, 1973.

MATERIALS AND METHODS

Agents Used

ara-C (Cytosar) was supplied by The Upjohn Company,Kalamazoo, Mich.; HU, 5-HP, and camptothecin were obtainedfrom Drug Research and Development, National CancerInstitute, Bethesda, Md.; and TdR-3H (2.2 Ci/mmole) was

obtained from New England Nuclear, Boston, Mass. Injectionsolutions of ara-C, HU, and camptothecin were prepared in0.9% NaCl solution. We prepared injection solutions of 5-HPby first dissolving the agent in NaOH and then adjusting thepHto 10.0 with HC1.

Survival and Toxicity Studies

The parent line of LI 210 cells was obtained from Mr. I.Wodinsky (Arthur D. Little, Inc., Cambridge, Mass.) and wasmaintained by weekly passage (i.p. inoculation) of ascites cellsin female DBA/2 mice (The Jackson Memorial Laboratories,Bar Harbor, Maine). Therapeutic studies were conducted infemale C57BL/6 X DBA/2 F, (hereafter called BDF,) mice

APRIL 1973 895

on July 20, 2017. © 1973 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

G. L. Neil and E. R. Homan

from the same source, weighing 19 ±2 g. Mice were inoculatedon Day 0 with 1 X 10s L1210 cells per mouse. All of the

agents were administered i.p. Therapy was initiated 48 hr later,on Day 2. The interval between doses was varied from 0 to 12hr (or from 0 to 24 hr) in 1- or 2-hr increments. Eight or 10mice per group were used. Mice were observed and their deathswere recorded twice daily. Mean survivals ±S.D. werecalculated. Animals were weighed on Day 1 and again on Day5, and average weight changes (g/mouse) were determined. Theapproximate percentage cell kill achieved was calculated fromsurvival data, by the method of Schabel (17). Experimentswere terminated after 45 days, and 45-day survivors were

considered cured.

Recovery of DNA Synthesis.

BDFi mice (same sex and weight as those used in therapystudies) were inoculated i.p. on Day 0 with 1 X IO6 L1210

cells/mouse. On Day 3 (when the total tumor load wasapproximately 5 X IO7 cells/mouse), mice were dosed i.p.with drug and, at various times thereafter, TdR-3H (10

juCi/mouse, 2.2 Ci/mmole) was administered (0.5 ml volume).Four to 6 mice/group were used. In addition, control mice(not treated with drug) were dosed with the same amount ofTdR-3H. Control groups (2 or 3 groups, at different times

during the experiment) contained 8 to 12 mice/group. Thirtymin after the administration of TdR-3H, mice were sacrificed

by cervical dislocation. Heparinized 0.9% NaCl solution (2.5ml) was injected into the peritoneal cavity. This cavity wasmassaged, and approximately 1.5 to 2.0 ml of ascites fluidwere withdrawn and added immediately to 0.1 ml ofnonradioactive thymidine (10 mg/ml) in 0.9% NaCl solution.A 1:400 dilution was prepared and used for electronic particlecounting (Coulter counter; Coulter Electronics, Hialeah, Fla.),and incorporation into DNA was determined by 1 of thefollowing methods.

Test Tube Method. Cells were sedimented by centrifugation;macromolecules were then precipitated with 5 ml cold 10%trichloroacetic acid. Two washes with 5 ml of trichloroaceticacid were performed, followed by 2 washes (5 ml each), 1 withethanol:ether (3:1) and 1 with ether. After air drying, theDNA was extracted from the cell pellet with 1 ml of 0.5 NPCA at 70°for 20 min. After cooling, the suspension was

centrifuged, and the radioactivity of 0.1-ml aliquots of thePCA supernatant was measured by scintillation techniques,with a Packard Tri-Carb Model 574 liquid scintillationspectrometer (Packard Instrument Co., Downers Grove, 111.).

Millipore Disc Technique. Two 0.5-ml portions of ascites cellsuspensions were pipetted onto Millipore discs (0.45-jum poresize; 25-mm diameter) on a filter apparatus. The liquid wasremoved by suction filtration. Discs were washed 3 times with4-ml portions of 0.2 N cold PCA and once with absoluteethanol. The discs were then placed in scintillation-countingvials, 0.5 ml of 0.5 N PCA was added, and DNA was extractedby heating at 70°for 20 min. Fifteen ml of Diotol were then

added, and radioactivity was determined as described above.Specific incorporation (cpm/106 cells) was determined.

Incorporation rates (DNA synthesis) were compared to thoseobserved in control animals and are expressed as percentage ofcontrol value.

RESULTS

Charts 1 through 4 show the effects of the length of theinterval between 2 doses of ara-C, HU, 5-HP, and campto-thecin (respectively) on the survival and weight change ofL1210 leukemia. These results are compared with the effectsof the agent on DNA synthesis of L1210 ascites cells at varioustimes after a single administration of the agent in question.

Chart 1 shows the results with ara-C. In this case, theintervals between doses ranged from 0 to 24 h. Twosimultaneous i.p. doses of ara-C, each 100 mg/kg, (equivalent,of course, to a single dose of 200 mg/kg) yielded a survival of9.4 ±1.6 days, as compared to 7.9 ±0.8 days in the control.This survival represents a cell kill of approximately 85%.However, if the 2 doses were administered sequentially, meansurvival increased rather regularly as the interval between thedoses increased from 0 to approximately 10 hr. With an

o<tXuh-X •¿�caUJ

£

-2Sä-1

I Oâ„¢¿�+1

+2°Control

ara-C dose: 100 mg/kg x 2 -

99.9

99 i

90 "

O 4 8 12 16 20 24INTERVALBETWEENTWO DOSES(HOURS)

Ú 15

¡=2

10

! e1se

~ 5

ara-C dose:100 mg/kg

04 8 12 16 20 24INTERVALBETWEENARA-C DOSEAND

TdR 3HPULSE(HOURS)Chart 1. Dose-interval studies with ara-C in L1210 leukemic mice;/!

and B, survival studies. LI210 leukemic BDF, mice were treated 2 daysafter tumor inoculation with 2 i.p. doses (100 mg/kg each) of ara-C.The interval between the doses was varied. A, average weight change(g/mouse; Days 0 to 5); B, mean survival times (±S.D.).Cell kill wascalculated by the method of Schabel (17). C, DNA synthesis recoverystudies. L1210 leukemic BDF, mice were treated with a single i.p. doseof ara-C (arrow). At various times, TdR-3 H (10 |uCi/mouse; 2.2

Ci/mmole) was injected i.p. Incorporation into DNA of peritonealascites cells (cpm/106 cells) was determined 30 min later. The intervalshown does not include the 30-min labeling period. Results areexpressed as percentage of incorporation in control (untreated) mice.Results of 3 separate experiments are shown.

896 CANCER RESEARCH VOL. 33

on July 20, 2017. © 1973 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Dose Interval Effects

interval of 9 hr, a mean survival of 12.7 ±0.8 days(corresponding to >99.5% cell kill) was achieved. However, aninterval of 12 hr was considerably less effective (mean survival,10.3 days), and there appeared to be some indication of a 2ndoptimal interval at approximately 16 hr.

Although the optimal therapeutic interval for 2 doses ofara-C was approximately 8 to 10 hr, weight loss was alsoparticularly severe at this interval. It appeared that, with someintervals, the therapeutic index was more favorable. Forexample, at an interval of 3 hr, although weight loss wasapproximately equal to that with 2 simultaneous doses, thetherapeutic effect was somewhat enhanced. Apparently,judicious choice of dose interval can result in an increasedtherapeutic index.

Further evidence that the toxicity of ara-C is dependent ondose interval is shown by the data in Table 1. Although a totaldose of 4000 mg/kg i.v. (in 2 equal doses) could beadministered without lethality to male albino mice if theinterval between the doses was less than approximately 4.5 hr,2 doses of 400 mg/kg, with an interval of 9 hr between them,resulted in 60% mortality. Two doses of 2000 mg/kg, with aninterval of 9 hr, were lethal to all mice.

After the administration of a single 100-mg dose of ara-Cper kg to LI 210 leukemic mice, ascites-cell DNA synthesis (asmeasured by TdR-3H incorporation) is rapidly inhibited and

remains at less than 1% of control level for approximately 4hr. DNA synthesis recovers after this time, reaching amaximum (of approximately 15 to 25% of control) afterabout 10 hr. This is approximately the same interval at whicha 2nd dose of ara-C gives optimal therapeutic effects.

Results of similar studies with HU are shown in Chart 2. Inthis and the following cases, the intervals studied ranged from0 to 12 hr. With HU, 2 simultaneous doses of 250 mg/kgyielded very little therapeutic effect (mean survival, 8.4 ±0.3days compared with 8.0 ±0.4 days for the control). Thiscorresponded to less than 50% cell kill. However, at an intervalof 4 to 5 hr, survivals of approximately 12 days were observed(corresponding to approximately 99.5% cell kill). Survival thendecreased as the interval increased (to approximately 7 hr). Atan interval of 10 to 12 hr, another peak of activity wassuggested. No particular pattern was observed with regard to

Table 1Lethal toxicity of ara-C in albino mice (effect of dose interval).

Male albino mice (10 per treatment group) were treated (i.v.) at zerotime with ara-C at the doses indicated. At an appropriate time, a 2nddose (same route and size) was administered. The dose interval wasvaried from 1.5 to 12 hr.

% mortality by day 12°with

ca=<.

Dose interval(hr)1.534.567.5910.512ara-C,

400mg/kg/dose000203060200ara-C,

900mg/kg/dose001003080500ara-C,2000mg/kg/dose00030101009040

<oc ~y

£2c7>

O+1+2

12

10

8

6

100

75'•O

1 A

|cÌ50i "o

] ^=i* 25

TControlHU dose: 250 mg/kg x 2

99 _,•¿�jae

90 iLUC9

O 2 4 6 8 10 12INTERVALBETWEENTWO DOSES(HOURS)

HU dose: 250 mg/kg

" Ail but 2 of the 57 deaths occurred by Day 7.

0 2 4 6 8 10 12INTERVALBETWEENHU DOSEAND TdR-3H PULSE

(HOURS)

Chart 2. Dose-interval studies with HU in L1210 leukemic mice; Aand B, survival studies. L1210 leukemic BDF, mice were treated 2 daysafter tumor inoculation with 2 i.p. doses of HU. The interval betweenthe doses was varied. A, average weight change (g/mouse; Days 0 to 5);B, mean survival times (±S.D.).Cell kill was calculated by the methodof Schabel (17). C, DNA synthesis recovery studies. L1210 leukemicBDF, mice were treated with a single i.p. dose of HU (arrow). Atvarious times, TdR-3 H (10 ¿iCi/mouse;2.2 Ci/mmole) was injected i.p.Incorporation into DNA of peritoneal ascites cells (cpm/106 cells) was

determined 30 min later. The interval shown does not include the30-min labeling period. Results are expressed as percentage ofincorporation in control (untreated) mice.

toxicity (as measured by weight change), again indicating thatthe therapeutic index could be improved by the judiciouschoice of interval. With HU, the optimal interval for 2 doses ofthe agent correlated extremely well with the interval at whichDNA synthesis recovered to its maximum after the initialinhibition, following 1 dose of HU. A 2nd recovery of DNAsynthesis was also observed at approximately 10 to 12 hr,corresponding to what appears to be another good interval fortherapeutic results.

Results with 5-HP are shown in Chart 3. Two simultaneousdoses of 250 mg/kg yielded less than 60% cell kill (meansurvival, 8.6 ±0.5 days compared with 8.0 ±0.4 days in thecontrol group). Survival was again a function of the doseinterval, with a peak (mean survival, 11.8 ±0.7 days; cell kill,approximately 99.3%) occurring at approximately 6 hr. Littlechange in toxicity (as measured by weight loss) was observedas a function of dose interval. With 5-HP, recovery of DNA

APRIL 1973 897

on July 20, 2017. © 1973 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

G. L. Neil and E. R. Homan

synthesis appeared to lag significantly behind the optimalinterval for 2 doses.

Results with camptothecin are shown in Chart 4. In thiscase, there was no obvious relationship between the doseinterval and the survival of LI210 leukemic mice. Two20-mg/kg doses administered simultaneously were quiteeffective (mean survival, 13.0 ±0.7 days, compared with 8.8 ±0.6 days with controls; cell kill, approximately 99.3%) andgave results similar to those observed when the intervalbetween doses ranged from 1 to 8 hr. A slight decrease inactivity was observed when the interval was 9 to 12 hr.However, with this agent, toxicity appeared to increase ratherregularly as the interval between the 2 doses increased. Thedecrease in therapeutic activity observed at intervals of 9 to 12hr may reflect this rather severe toxicity. No recovery of DNAsynthetic capacity was observed after a single dose ofcamptothecin at intervals up to 12 hr.

The extent and time course of DNA synthesis recovery ofascites L1210 cells is shown asa function of ara-C dose (Chart5). In this chart, results are plotted semilogarithmically toemphasize the extent of inhibition of DNA synthesis. At a

"lS2LU

+2

12

- 10M

I >>

I8

6

D Control

Control5-HP dose: 250 mg/kg x 2

99 _,2ae

90-LU

o

O 2 4 6 8 10 12INTERVAL BETWEEN TWO DOSES (HOURS)

0 2 4 6 8 10 12INTERVAL BETWEEN 5-HP DOSE AND TdR-3H PULSE

(HOURS)Chart 3. Dose-interval studies with 5-HP in LI210 leukemic mice;,4

and B, survival studies. L1210 leukemic BDF, mice were treated 2 daysafter tumor inoculation with 2 i.p. doses of 5-HP. The interval betweenthe doses was varied. A, average weight change (g/mouse; Days 0 to 5 ; Bmean survival times (±S.D.).Cell kill was calculated by the method ofSchabel (17). C, DNA synthesis recovery studies. L1210 leukemicBDF, mice were treated with a single i.p. dose of 5-HP (arrow). Atvarious times, TdR-3H (10 juCi/mouse; 2.2 Ci/mmole) was injected i.p.Incorporation into DNA of perintoneal ascites cells (cpm/10' cells) was

determined 30 min later. The interval shown does not include the30-min labeling period. Results are expressed as percentage ofincorporation in control (untreated) mice.

-2

£2 +2

14

"

GControl

PControlCamptothecin dose: 20 mg/kg x 2

0 2 4 6 8 10 12INTERVAL BETWEEN TWO DOSES (HOURS)

A

99.9

_J

99 *

LU

U90 ae

0

B

100

Si 75

>- £50

25

Camptothecin dose.-20 mg/kg

0 2 4 6 8 10 12INTERVAL BETWEEN CAMPTOTHECIN DOSE AND

TdR3H PULSE (HOURS)

Chart 4. Dose-interval studies with camptothecin in L1210 leukemicmice; A and B, survival studies. L1210 leukemic BDF, mice weretreated 2 days after tumor inoculation with 2 i.p. doses (20 mg/kgeach) of camptothecin. The interval between the doses was varied. A,average weight change (g/mouse; Days 0 to 5); B, mean survival times(±S.D.).Cell külwas calculated by the method of Schabel (17) C, DNAsynthesis recovery studies. LI210 leukemic BDF, mice were treatedwith a single i.p. dose of camptothecin (arrow). At various times,TdR-3H (10 ¿iCi/mouse;2.2 Ci/mmole) was injected i.p. Incorporation

into DNA of peritoneal ascites cells (cpm/10 cells) was determined 30min later. The interval shown does not include the 30-min labelingperiod. Results are expressed as percentage of incorporation in control(untreated) mice.

dose of 10 mg ara-C per kg, recovery of DNA synthesis wassomewhat more rapid and achieved higher levels than at dosesof 30 or 100 mg/kg. The peak of this recovery, however, stilloccurred approximately 8 to 10 hr after drug administration.

DISCUSSION

All of the agents examined in this study (ara-C, HU, 5-HP,and camptothecin) inhibit DNA synthesis of L1210 cellsgrowing in culture more than they inhibit either RNA orprotein synthesis (3). Bhuyan et al. (4) have shown withsynchronized DON Chinese hamster cells that all of theseagents are S-phase specific; that is, they are maximallycytotoxic to cells in the S (DNA-synthetic) phase of the cellcycle. Results with asynchronous LI210 cells in culture (3) are

848 CANCER RESEARCH VOL. 33

on July 20, 2017. © 1973 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Dose interval Effects

also consistent with with the S-phase specificity of theseagents.

If one makes a number of assumptions, it is possible toconstruct a very simple model to describe the cell kill onemight expect with 2 short exposures of LI 210 cells toS-phase-specific agents. The assumptions involved are thefollowing, (a) The time of exposure to cells to the agent(contact time) is short with respect to the cell-cycle time, (b)All S-phase cells (but no non-S-phase cells) are killed (i.e.,rendered incapable of further proliferation) by an instantaneous exposure to some minimally effective level of theagent in question, (c) The LI210 cell population is homogeneous with respect to cell-cycle times and the lengths of thevarious phases, (d) The exposure to the agent does not affectthe progression of non-S-phase cells into the S phase (i.e., theagents are not "limiting").

The phase and cell-cycle lengths for L1210 cells in vivo havebeen determined (G. L. Neil and T. E. Morley, unpublishedresults, our laboratories)4 and are listed below with estima

tions of the fractions of cells in any particular phase. Thesefractions are roughly approximated by division of the phaselength by the total cell-cycle time (rc).5

100

50

Phase

Gj +MTotal cycle

Length (hr)

1.9 (To, )6.6 (Ts)l-5(rGa +M

10.0(7c)

Approximate % oftotal no of cells ina particular phase

196615

100

Using the above assumptions and cell-cycle parameters, onemay calculate that the fractional cell kill (expressed as afraction of those cells present at the time of administration)expected for a single administration of an S-phase-specificagent would equal (!F8 + length of effective exposure time)/rc= (6.6 + length of effective exposure time)/10.0. For the ideal

case in which exposure time is infinitely short, the cell killachievable with a single dose would simply equal the fractionof cells in the S phase (approximately 66%). For an effectiveexposure of 2 hr, the cell kill would be approximately 86%.

The cells unaffected by the initial exposure (i.e., those inGÌ,G2, or M during that time) will then become mostsensitive to a 2nd exposure when they have all progressed intothe S phase. The optimal time for a 2nd dose will correspondto the time when all of these cells are in the susceptible Sphase. For an agent with infinitely short contact time, thisdose may be administered 3.4 (TC-TS) to 6.6 (Ts) hr after the1st dose. After 6.6 hr, the cells will begin to enter the G2phase and the population will become less sensitive to theagent. For an agent with an effective contact time of 2 hr, a2nd dose would be optimal anywhere from 3.4 to 8.6 hr afterthe 1st dose.

Cells surviving the 1st dose will of course be least sensitiveto a 2nd dose when they have all progressed through the S

4These results are in reasonable agreement with other published data

(19,21,22).5The actual percentages in each phase, calculated by the more

rigorous method of Okada (15), are: G2 + M, 11%; G,, 23%; S, 66%.For simplicity, the estimates based directly on phase length (see above)are used in "Discussion."

o 10cet-

O tO 3

u-O

vtLU

S iz>-

< 0.5

0.1

ara-C dose (mg/kg)

•¿�10

o 30

•¿�100

O 2 4 6 8 10 12INTERVAL BETWEEN ARA-C DOSE AND TdR-3H PULSE

(HOURS)Chart 5. Dose-interval studies with ara-C in L1210 leukemic mice;

dose response and DNA synthesis recovery. L1210 leukemic BDF,mice were treated with a single i.p. dose of ara-C (arrow). At varioustimes, TdR-3H (10 fiCi/mouse; 2.2 Ci/mmole) was injected i.p.Incorporation into DNA of peritoneal ascites cells (cpm/106 cells) was

determined 30 min later. The interval shown does not include the30-min labeling period. Results are expressed as percentage ofincorporation in control (untreated) mice.

phase and returned to the nonsusceptible phases. This willoccur 10 hr (equivalent to the cell-cycle time) after the initialexposure to an agent with an infinitely short contact time(10.0 to 12.0 hr after an agent with an effective contact timeof2hr).

Before discussing how well the data fit this simple model,we offer the assumptions involved in its construction forexamination, as follows.

Assumption 1. The time of exposure of target cells to theagent will be determined by the pharmacokinetic behavior ofthe agent. The pharmacokinetic behavior of ara-C (5, 14, 16),HU (13), camptothecin (7), and 5-HP (20) in mice has beenstudied. All of these drugs have relatively short plasmahalf-lives and, after the administration of doses similar to thoseused in our study, effective levels would be expected to persistin the plasma for no more than 1 to 3 hr. In some cases,however, body fluid levels may not accurately reflect the levelsin the target cells themselves. Metabolsim (both anabolic andcatabolic) within the cell itself is often of utmost importance.For example, phosphorylation of ara-C (to its correspondingnucleotides) can serve as a concentration or "trapping"

mechanism which may not be reflected in body fluid levels.As mentioned above, for a nonlimiting agent (i.e., one that

does not affect the progress of non-S-phase cells inttt the Sphase), an increase in the effective contact time will increasethe period of time during which administration of a 2nd dose

APRIL 1973 899

on July 20, 2017. © 1973 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

G. L. Neil and E. R. Homan

will be optimal. However, the earliest time (3.4 hr) at which a2nd dose will be optimal will not be changed. With a limitingagent, the optimal dose interval may be longer than thatdescribed above and will depend on the contact time of theagent and on the time required for the non-S-phase cells torecover their ability to progress into the S phase.

Assumption 2. Cells at different points in their progressionthrough the S phase may well have different sensitivitiestowards agents interfering with DNA synthesis. There is, ofcourse, also some finite exposure time necessary so that thecytotoxic activity of an agent can manifest itself. This isparticularly likely for an agent such as ara-C, which requiresanabolic transformation to mainfest its lethal activity. Although ali of the agents studied may rightfully be consideredS-phase specific, this specificity is probably not absolute, andkilling of non-S-phase cells may be possible at appropriateconcentrations. This appears to be particularly true forcamptothecin; with this agent (although maximal cytotoxicityis seen for S-phase cells), cells in GÌ, and G2, and M are alsokilled (4).

Assumption 3. The cell population is, of course, notcompletely homogeneous with respect to cell cycle and phaselengths; indeed, the cell-cycle parameters referred to aboverepresent only average values.

Assumption 4. It is well documented that agents such asara-C (6) and HU (1, 18), for example, are self-limiting tosome extent, in that they affect the progression of non-S-phasecells through the rest of the cell cycle and, in particular, slowthe progression of GÃŒcells into the susceptible S phase. This isdiscussed in more detail in a companion article (3).

The simple model described above would also predict thatDNA synthesis (TdR-3H incorporation) should increase afterthe initial inhibition, as non-S-phase cells enter S. The time formaximal recovery of DNA synthetic capacity would thencoincide with the time for optimal administration of a 2nddose. It has been shown by Sinclair (18), however, that for atleast 1 of the agents studied (HU), S-phase cells, destined todie after exposure to the drug, do recover (to a limited extent)the ability to incorporate TdR-3H. The magnitude and kinetics

of such incorporation may well perturb any expectedcorrelation of optimal dose interval and time for recovery ofDNA synthetic capacity.

In view of the many uncertainties involved in theassumptions forming the basis of the simple model described,it it perhaps somewhat presumptuous to use it to predict (orexplain) the data obtained. However, in some cases, theagreement was surprisingly good. With 3 of the agents (ara-C,HU, and 5-HP), the cell kill achieved with a single dose(approximately 50 to 85%) is consistent with that predictedfor S-phase-specific agents of short duration.

Dose intervals are also found at which a 2nd administrationof the agent greatly increases the cell kill. Cell kills in excess of99.5% were achieved. With HU and 5-HP, the optimal doseinterval (5 and 6 hr, respectively) is consistent with thatpredicted (approximately 3.4 to 6.8 hr) by the simple model.The optimal interval for a 2nd dose of ara-C is somewhatlonger (approximately 9 hr) than the predicted value. Similarresults were observed in studies in vitro (3). Karon andShirakawa (8) showed, with DON cells, that after a finite

exposure to ara-C, subsequent progression of GÃŒcells into Swas slowed (8), a phenomenon that might explain the resultsobserved in our studies.

With ara-C, HU, and 5-HP, DNA synthesis did recover afteran initial dose, and the time for maximal recovery with ara-Cand HU corresponded closely to the optimal time foradministration of a 2nd dose. The lag in DNA synthesisrecovery with 5-HP may reflect DNA synthesis resumption byS-phase cells that are destined to die.

Results obtained with camptothecin are entirely inconsistent with the model presented. Single doses (20 mg/kg)apparently killed more than the S-phase population, no doseinterval effect was observed, and DNA synthesis did notrecover. Results of studies mentioned above (4), indicatingthat camptothecin kills a significant number of cells in phasesother than S, may represent at least a partial explanation forthis discrepancy.

The synchronization of mammalian cells in vivo bytreatment with DNA synthesis inhibitors has been reportedpreviously. Bertalanffy and Lindsey-Gibson (2) observedpartial synchrony with ara-C in B16 murine melanoma.Madoc-Jones and Mauro (13) described partial synchrony ofmouse lymphoma cells with HU. Using a spleen colony assay,they studied the effects of the interval between a dose of HUand either a 2nd dose of HU, a dose of a Vinca alkaloid, or adose of X-rays, on the survival of the lymphoma cells.Life-span data were not reported, however.

The results of this study have some clinical implications.Judicious choice of dose interval in the treatment of humanneoplasia with agents such as those used in our study mayresult in enhanced therapeutic effects. Measurements of thetime course of recovery of DNA synthesis of leukemia cells,for example, after an initial dose of 1 of these agents shouldhave some predictive value in determining such intervals.Indeed, Lampkin et al. (10-12) studied the recovery of DNAsynthesis in leukemic cells of patients after treatment withara-C and found evidence for a considerable degree ofsynchronization. These authors used a more rigorous measureof DNA synthesis (labeling index) which allows one todetermine the number of cells actively incorporating TdR-3H.

The incorporation method used in our study measures onlythe total rate of DNA synthesis. However, because of the greattime savings effected, this latter method might prove moreuseful in a clinical situation wherin autoradiographic techniques may be impractical. Lampkin et al. (12) have alsostudied the effects of several phase-specific chemotherapeuticagents (including ara-C) administered to patients after initialsynchronization with ara-C, and they concluded that a greatertherapeutic advantage was achieved over that seen withadministration of the 2nd agents themselves.

The determination of optimal schedules for the clinicaltreatment of neoplastic disease with chemotherapeutic agentsis an exceedingly important and difficult task. Hopefully, thestudies described above, particularly those relating changes in apertinent biochemical parameter (DNA synthesis recovery) tooptimal dose intervals in an experimental animal system, willencourage further studies of a similar nature at the clinicallevel.

900 CANCER RESEARCH VOL. 33

on July 20, 2017. © 1973 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Dose Interval Effects

ACKNOWLEDGMENTS

We thank Dr. B. K. Bhuyan for helpful discussions and suggestionsand Ms. Elizabeth L. Clark and Mrs. N. Lallinger for editorial andsecretarial assistance.

REFERENCES

1. Bacchetti, S., and Whitmore, G. F. The Action of Hydroxyurea onMouse L Cells. Cell Tissue Kinet., 2: 193-211, 1969.

2. Bertalanffy, F. D., and Lindsay-Gibson, M. H. The in Vivo Effectsof Arabinosylcytosine on the Cell Proliferation of Murine B16Melanoma and Ehrlich Ascites Tumor. Cancer Res., 31: 66-71,1971.

3. Bhuyan, B. K., Fräser,T. J., Gray, L. G., Kuentzel, S. L., and Neil,G. L. Cell-Kill Kinetics of Several S-phase Specific Drugs. CancerRes., 33. 889-894, 1973.

4. Bhuyan, B. K., Scheldt, L. G., and Fraser, T. J. Cell-Cycle PhaseSpecificity of Antitumor Agents. Cancer Res., 32: 398-407, 1972.

5. Borsa, J., Whitmore, G. F., Valeriote, F. A., Collins, D., and Bruce,W. R. Studies on the Persistence of Methotrexate, CytosineArabinoside, and Leucovorin in Serum of Mice. J. Nati. CancerInst., 42: 235-242, 1969.

6. Graham, F. L., and Whitmore, G. F. The Effect of 1-0-D-Arabino-furanosylcytosine on Growth, Viability, and DNA Synthesis ofMouse L-Cells. Cancer Res., 30: 2627-2635, 1970.

7. Guarino, A. M., Hart, L. G., Call, J. B., and Oliverio, V. T. Studiesof the Physiological Disposition of the Antitumor Agent Campto-thecin Sodium (NSC-100880) in Rodents. Proc. Am. Assoc. CancerRes., 10: 33,1969.

8. Karon, M., and Shirakawa, S. Effect of 1-0-D-Arabinofuranosyl-cytosine on Cell Cycle Passage Time. J. Nati. Cancer Inst., 45:861-867, 1970.

9. Kline, I., Venditti, J. M., Tyrer, D. D., and Goldin, A.Chemotherapy of Leukemia LI210 in Mice with 1-0-D-Arabino-furanosylcytosine Hydrochloride. I. Influence of Treatment Schedules. Cancer Res., 26: 853-859, 1966.

10. Lampkin, B. C., Nagao, T., and Mauer, A. M. Drug Effect in AcuteLeukemia. J. Clin. Invest., 48: 1124-1130, 1969.

11. Lampkin, B. C., Nagao, T., and Mauer, A. M. Synchronization ofthe Mitotic Cycle in Acute Leukaemia. Nature, 222: 1274, 1969.

12. Lampkin, B. C., Nagao, T., and Mauer, A. M. Synchronization andRecruitment in Acute Leukemia. J. Clin. Invest., 50: 2204-2215,1971.

13. Madoc-Jones, H., and Mauro, F. Age Responses to X-Rays, VincaAlkaloids, and Hydroxyurea of Murine Lymphoma Cells Synchronized in Vivo.}. Nati. Cancer Inst. ,45: 1131-1143, 1970.

14. Neil, G. L., Moxley, T. E., and Manak, R. C. Enhancement byTetrahydrouridine of 1-0-D-Arabinofuranosylcytosine (Cytarabine)Oral Activity in L1210 Leukemic Mice. Cancer Res., 30:2166-2172,1970.

15. Okada, S. A Simple Graphic Method of Computing the Parametersof the Life Cycle of Cultured Mammalian Cells in the ExponentialGrowth Phase. J. Cell Biol., 34: 915-916, 1967.

16. Pittillo, R. F., and Hunt, D. E. Cytosine Arabinoside Sensitivity inActinobolin-Resistant Streptococcus faecalis: The Basis of aUtilitarian Microbiological Assay. Proc. Soc. Exptl. Biol. Med.,124: 636-640, 1966.

17. Schabel, F. M., Jr. Improvement of Therapeutic Activity ofAntileukemic Agents in Experimental Systems by Varying DoseSchedules and by Combination Chemotherapy. In: T. C. Hall andK. Kairer (eds.), Proceedings. Fifth International Congress ofChemotherapy, International Society of Chemotherapy. CancerChemotherapy, Dose Schedules, and Combinations, pp. 65-82.Vienna: Verlag der Wiener Medizinischen Akademie, 1967.

18. Sinclair, W. K. Hydroxyurea: Differential Lethal Effects onCultured Mammalian Cells during the Cell Cycle. Science, 750:1729-1731,1965.

19. Skipper, H. E., and Schabel, F. M. Jr., and Wilcox, W. S.Experimental Evaluation of Potential Anticancer Agents. XXI.Scheduling of Arabinosylcytosine to Take Advantage of its S-PhaseSpecificity against Leukemia Cells. Cancer Chemotherapy Rept.,51: 125-165, 1967.

20. Tomchick, R., Weinstein, E. P., and Mead, J. A. R., ThePhysiological Distribution of 14C-Labeled 5-Hydroxypyridine-2-carboxaldehyde Thiosemicarbazone (NSC-107392) in Mice.Biochem. Med., 6: 282-293, 1972.

21. Wheeler, G. P., Bowdon, J. Wilkoff, L. J., and Dulmadge, E. A. TheCell Cycle of Leukemia LI210 Cells in Vivo and in Vitro. Proc.Soc. Exptl. Biol. Med., 726: 903-906, 1967.

22. Yankee, R. A., DeVita, V. T., and Perry, S. The Cell Cycle ofLeukemia L1210 Cells in Vivo. Cancer Res., 27: 2381-2385,1967.

APRIL 1973 901

on July 20, 2017. © 1973 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1973;33:895-901. Cancer Res   G. L. Neil, E. R. Homan, T. E. Moxley, et al.   Mice Treated with DNA Synthesis InhibitorsThe Effect of Dose Interval on the Survival of L1210 Leukemic

  Updated version

  http://cancerres.aacrjournals.org/content/33/4/895

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  .permissions@aacr.orgDepartment at

To request permission to re-use all or part of this article, contact the AACR Publications

on July 20, 2017. © 1973 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from