[CANCER RESEARCH 35, 1530-1535,June1975]

SUMMARY

The stem-cell response of the Lewis lung carcinoma tosingle doses of cyclophosphamide has been studied by threeassay techniques: in vitro colony formation, lung colonyformation, and the end-point dilution assay. These threetechniques have given comparable results, and the end-pointdilution results showed that 50% takes could be achievedwith as few as 1 to 3 cells. Studies have been made of thegrowth of small i.m. implants and of the time followingimplantation at which they could be eradicated by cyclophosphamide. The results were compared with the curabilitythat would be expected on the basis of cell survival studies.It was found that older (and therefore larger) implants werecured than might have been expected. It seems unlikely thatthis discrepancy was due to additional cell kill caused by animmune response. An alternative explanation, that thesurviving fraction following a dose of cyclophosphamidewas lower in small implants than in larger i.m. tumors, wassupported by studies of cell survival in dissectable lungcolonies.

INTRODUCTION

The response of experimental tumors to irradiation or tochemotherapeutic agents may be assessed by a variety ofend points, which can be divided into the main categories oftumor inhibition and regrowth studies, tumor cure ratestudies, and studies of cell survivial. While regrowth andtumor control studies have the advantage of directness ofapplication, the cell survival studies offer the hope ofidentifying some of the cellular mechanisms involved intumor response. In particular they can sometimes show thatsuccessful treatment is frustrated by a minority of resistanttumor stem cells (21). The major drawback to cell survivalstudies is, however, their artificiality. The assays involvesubjecting tumors to the additional insult of trypsinizationor mechanical homogenization, thereby removing each cellfrom its in vivo relationship with other neoplastic orconnective tissue cells. It is important, therefore, to examinedirectly the relationship between the results of cell survivalstudies and tumor regrowth and control (12, 13), partly inorder to substantiate the cell survival data but also toinvestigate the mechanisms involved in the regrowth of

1 This work was partly supported by National Cancer Institute Contract

N0l-CM-237 17.ReceivedSeptember4, 1974;acceptedFebruary 6, 1975.

tumors from small numbers of surviving cells. The projectdescribed in this report was designed to investigate therelationship between cell survival and tumor control in theLewis lung tumor. The results also have a bearing on theimportance of tumor size in determining therapeutic response.

MATERIALS AND METHODS

Lewis Lung Carcinoma. A specimen of Lewis lungcarcinoma was obtained from Dr. K. Hellman of theImperial Cancer Research Fund, London, England, inDecember 1971. It has since been maintained within C57BLmice of the Institute of Cancer Research colony. Itshistological and macroscopic characteristics are consistentwith those reported by Sugiura and Stock (20). Thetransplantation procedure has been to chop the tumor tissuefinely with crossed scalpels, make an homogenate by forcingit through a syringe with 10 volumes of balanced saltsolution, and then inject 0.02 ml bilaterally into thegastrocnemius muscles of recipient mice. No such implantshave failed to take.

Tumor Size Measurement. Measurement of the size ofi.m. tumors was made by a calibration curve technique (19).Mice bearing single i.m. tumors that had been produced bythe implantation of between 102 and l0@cells were killedover a range of tumor sizes. Each tumor-bearing leg wasshaved and its external diameter at the tumor site wasmeasured with calipers. Both hind legs were then disjointedat the knee, the skin was removed down to the ankle, andthe legs were weighed. The difference in weight between thetumor-bearing and control legs was taken as an estimate oftumor weight. The assumption in this was that edema due tothe presence of the tumor did not seriously affect the result.Chart 1shows the calibration curve relating tumor weight toleg diameter. The scatter in the points is reasonably smallalthough, in view of the assumption that has just beenmentioned, it is possible that the data in this chart may besubject to systematic error.

Cell Suspension Technique. Mice with i.m. tumors measuring 10 to 15 mm in diameter were killed by cervicaldislocation. The tumors were dissected out using sterileprecautions, freed from the muscle, and weighed. They werethen finely chopped with crossed scalpels in a small amountof a mixture of phosphate-buffer saline, NaC1, 8.0 g; KC1,0.2 g; Na2HPO4, 1.15 g; KH2PO4, 0.2 g; phenol red,0.02 g; and distilled water, 1.0 liter. A cell suspension was

I530 CANCERRESEARCHVOL.35

Stem-Cell Survival and Tumor Control in the Lewis LungCarcinoma'

G. Gordon Steel and Kay Adams

Departments of Biophysics and Radiotherapy Research, Institute of Cancer, Belmont, Surrey, England

Association for Cancer Research. by guest on September 3, 2020. Copyright 1975 Americanhttps://bloodcancerdiscov.aacrjournals.orgDownloaded from

Ce!! Survival and Tumor Control

prepared by trypsinization in phosphate-buffered salinecontaining 2.5% by volume of Bacto-Trypsin, reconstitutedas recommended by Difco Laboratories, Inc., Detroit,Mich. Brief shaking of the tumor fragments in warm trypsin-free medium at the end of trypsinization greatly increased the yield of viable tumor cells. After the cells werewashed in Eagle's basal medium (Biocult Laboratories,Paisley, Scotland) they were filtered through a 400 meshstainless steel gauze under gravity. The cell viability asjudged by phase-contrast microscopy usually exceeded90%.

High transplantation efficiency could be obtained only byadding HR2 to the inocula. These were usually prepared in aseparate but simultaneous digestion. They were brought to afinal total cell concentration of 2 x 107/ml and were given adose in excess of 10,000 rads of 60Co ‘-y-raysat a dose rate ofabout 3,000 rads/min. Every experiment included controlsgiven 106 HR alone, and growth from such inocula wasobserved only in I experiment, which was abandoned.

For i.m. implantation, the viable cell concentration wasadjusted to give the required number of cells in 0.05 ml.Equal volumes of viable suspension and HR suspensionwere mixed and from this mixture volumes of 0. 1 ml wereimplanted, thus containing the viable cell dose plus 106 HR.

End-Point Dilution Assay. Graded inocula of viable cellswere implanted within a range that was designed to includethe TD50. The cell concentrations usually differed by afactor of 3.16 (i.e., ‘@,/Th)so that 2 dilutions just covered a10-fold range. Recipient mice were observed for 50 daysafter implantation and new tumors rarely appeared after 30days, irrespective of the treatment that the cells hadreceived. The results were analyzed by a computer programthat applied the method of Porter and Berry (I I). Thefraction of tumor cells surviving treatment was calculated asthe ratio of the TD50 estimate for untreated tumors to theTD50 estimate for treated tumors. Since it was possible toimplant a total cell number of over 106 into each implantation site, this type of assay could be used to measuresurvivals down to approximately 106.

Lung Colony Assay. The method was as described by Hilland Bush (6). Viable tumor cells were mixed with 106 HRand 106plastic l5-@tmmicrospheres, and they were injectedi.v. in a volume of0.3 ml. Lungs were removed and fixed inBouin's fixative at 21 days after injection, and the colonieswere counted under a dissecting microscope. In theseexperiments the yield of colonies was in the range of Icolony/ l0@ to l0@ untreated viable cells. The fraction ofsurviving tumor cells was calculated as the ratio of thecloning efficiency (colonies per 106 viable cells) ofcells fromtreated tumors to the cloning efficiency of cells fromuntreated tumors. With the present line of Lewis lungtumor, this method could be used to measure survivals downto lO_2or@

In Vitro Colony Assay. Colony formation in vitro wasstudied by the method developed for this tumor by Courtenay (manuscript in preparation, but referred to in Ref.16). The method was a double-layer soft agar technique

2 The abbreviations used are: HR. heavily irradiated tumor cells; TD,0,

viable cell inoculum for 50%tumor takes.

0,

.2'

13 15 17 19

Leg Diameter (mm)

Chart 1. Calibration curve relating the estimatedtumor weight to legdiameter. Points, results on 1 mouse.

using a gas phase containing 5% oxygen. For the Lewis lungtumor it was possible to measure survivals down to about 5x l0@ with this technique.

Antigenicity of the Lewis Lung Tumor. The immunological status of the tumor growing in our colony of mice wasassessed by measuring the TD50 for i.m. implantation afterattempted immunization. Three main types of immunization were used. In a series of experiments performed over aperiod of 18 months, HR were injected either i.m. into theleft hind leg or at 5 sites close to lymph nodes, (2 axillary, 2inguinal, plus i.p.). The i.m. TD50 in untreated miceremained below 3 cells, while in the immunized groupssuccessive estimates of the TD50 were 26, 92 (hind legimmunization), 99, 94, and 54 (para-lymph node immunization). The 3rd type of immunization was to allow i.m.implants in the left hind leg to grow for 7 to 10 days afterwhich the local tumor was eradicated by irradiation with3000 to 4500 rads of 60Co ‘y-irradiation.A single dose ofcyclophosphamide, 300 mg/kg, was also given in order tocontrol the metastases existing at this time and to produce asituation that was comparable with that in mice in which thecurability of artificial lung metastases was to be studied. Atboth 7 and 83 days after this combined treatment the i.m.TD50 (in untreated legs) was indistinguishable from con

RESULTS

Transplantability. For s.c. implantation the TD50 wasover l0@, and the addition of up to l0@ HR only reducedthis to 5 x l0@. However, i.m., the TD50 with viable cellsalone was 600 cells and the addition of 106 HR reducedthis to less than 3 cells. Chart 2 indicates the results of 1experiment using cells from untreated tumors. The fullline is a cumulative Poisson distribution. Its fit to the dataimplies uniformity of the implantation sites and that evensingle Lewis lung tumor cells could establish positive takesunder these conditions.

Growth Curve for i.m. Lewis Lung Tumors. The growth ofthe tumors during the “silentinterval― before detection wasexamined by a technique that has been used on experimental

trols.

1531June 1975

Association for Cancer Research. by guest on September 3, 2020. Copyright 1975 Americanhttps://bloodcancerdiscov.aacrjournals.orgDownloaded from

G. G. Steel and K. Adams

each leg it was possible to dissect out approximately 0.5 gof tumor. Cyclophosphamide was injected i.p., usually intogroups of 3 to 4 mice, and the animals were killed bycervical dislocation usually 3 hr afterwards. Cell suspensions were prepared and assayed by the 3 methods describedabove.

The results are shown in Chart 4. Measurements by theend-point dilution assay were made down to a survival ofl06, and over this range the survival curve is probablyindistinguishable from an exponential. There is evidence ofan initial shoulder equivalent to a cyclophosphamide dose ofabout 20 mg/kg. Agreement between the 3 assays is goodalthough the 2 points obtained by the lung colony assay areinsufficient to provide a precise comparison. The shape ofthe survival curve above a dose of 300 mg/kg has not beenexplored because of host toxicity.

Regrowth after Single Doses of Cyclophosphamide. Tumors growing i.m. were treated when they reached a legdiameter of 8 mm (a size of about 0.2 g as judged by Chart1); the regrowth data are shown in Chart 5. In this chart theestimated tumor sizes are relative to the size of the tumor atthe time of treatment. Also indicated (on the ordinate) arethe levels of cell survival associated with each cyclophosphamide dose, derived from the data of Chart 4. It isapparent that these levels of cell kill are much lower thanwould have been anticipated from the regrowth data alone.For instance, a backward extrapolation of the tumor weightdata for the dose of 300 mg/kg might indicate a survival ofabout lO_2, compared with the measured value of 106. Thesituation is reminiscent of the work of Hermens andBarendsen (5), who found that the volume response of anexperimental rhabdomyosarcoma to 2000 rads of X-irradiation was much less than was indicated by in vitro estimatesof the survival of clonogenic cells. We have not attempted tomeasure the regrowth of clonogenic cells in the Lewis lungtumors, but the broken lines in Chart 5 indicate theapproximate rate at which regrowth must have taken place.These lines have been drawn at their upper ends to the point

i;::

id7 ‘@ 200@ 3&DDoseof Cyclophosphamide(mqJkg)

Chart 4. Cell survival in approximately 500mg i.m. tumors following asingle dose of cyclophosphamide. The measurements were made by in vitroassay (0), lung colony assay (A), or end-point dilution assay (El).

leukemia (15, 18). A single-cell suspension was preparedand implants, ranging in decade steps from I to 106 cells,were made into both hind legs of groups of 10 mice. HR.l0, were added to each implant in all except the 106 viablecell group. The times at which the leg diameters reached 10mm were recorded and the median times are plotted inChart 3. The results of 2 separate experiments are shown.Positive takes were observed in all implant sites except thosemade with a mean of 1 cell/implant, where there were 7 of10 takes. Although there is some interexperiment variationin the TD50, and therefore in the proportion of viable cellsthat actually grow, this result suggests that in this experiment the TD50 was close to its ultimate value of0.7 and thatalmost every viable cell contributed to growth. With theassumption that this holds for all sizes of inoculum, thestraight line that fits these data may be taken to representthe growth curve from small implant sizes. It has a doublingtime of 1.02 days and there is no evidence for growthretardation up to a size of 106cells.

Survival Curve for Single Doses of Cyclophosphamide.Mice bearing i.m. Lewis lung tumors in both hind legs wereselected for treatment on Day I3 after the implantation of0.02 ml of a tumor homogenate (1:10 tumor in Eagle's basalmedium). At this time the legs were swollen from theirnormal width of about 5 mm to about 10 mm, and from

% @%% %

I―@i1ij,i7:@T.ãoi@ 1.0@

Mean cells per dose

.Chart 2. End-point dilution assay results for the Lewis lung tumorimplanted i.m. Points,proportion oftakes observedwhenthegivennumberof viable cells were mixed with l0 HR. The actual numbersof implantsand takes are indicated aboveeachpoint. Full line, cumulative Poissondistribution; bar, 95%confidencelimits on the TD,0 value.

C0

E

.2>

Mediantime to @mmleg diameterChart 3. The relation betweenthe numberof viablecells implantedand

the time it took the implants to reach a standard 10-mm leg diameter.Symbols, results of 2 separate experiments.

1532 CANCER RESEARCH VOL.35

Association for Cancer Research. by guest on September 3, 2020. Copyright 1975 Americanhttps://bloodcancerdiscov.aacrjournals.orgDownloaded from

‘I

I I •‘I / •/

I, /I@ /

I •,,I /

I ,i

II I

I

Ce!! Survival and Tumor Contro!

By multiplying the implant size by the relative survivalfactor, we may calculate the average number of clonogeniccells that should have survived treatment. If we then assumethat any 1 surviving clonogenic cell is capable of initiatingregrowth after treatment, then the probability of regrowthcan be calculated from a cumulative Poisson distribution (asin Chart 2, but displaced such that the TD50 is 0.7). Thiscalculational procedure gives the broken lines shown inChart 6.

The 1st point of interest in this result is that thetheoretical curves have the same shape as those defined bythe experimental data. The difference is only a displacementin time. Since the form (as opposed to the position) of thetheoretical curves depends only upon the statistics of thePoisson distribution, this implies that the shape of theexperimental tumor control curves is also dominated by thestatistics of the survival of individual clonogenic cells.

The 2nd conclusion is that the displacement of thetheoretical curves from the data is such as to indicate thatlarger implants were being controlled than might have beenexpected. The displacement is small for the 300-mg/kg dosebut large for the other 2 doses. This result is surprising inview of the natural expectation that tumor cure will beharder to achieve than one might predict and perhapsfrustrated by small numbers of resistant cells (21). Anexplanation of the discrepancy may be.sought in 2 generalhypotheses: (a) mechanisms other than cyclophosphamidetoxicity were also killing cells; and (b) cell kill due tocyclophosphamide was greater than is indicated by the datain Chart 4.

The idea that some additional cell kill was taking placeraises the question of the immune status of the tumor. Ourdata on this are summarized above and they show that slightantigenicity could be detected in mice immunized by theinjection of killed tumor cells. Tests performed on micewithin which i.m. tumors were eradicated by treatment withlocal irradiation and a single systemic dose of cyclophosphamide proved negative. Since the anomaly between thetumor control and cell survival data is in mice that had beentreated with cyclophosphamide, it seems likely that thistreatment may also have reduced or abolished an alreadylow immunological response to the tumor.

The other general hypothesis that might explain thediscrepancy between the expected and the actually observedage at which i.m. implants could be eradicated is that thedata in Chart 4 underestimate the amount of cell kill due tocyclophosphamide. This hypothesis is plausible because the

00a:

ic

1151

iO@

iO5

166

C0

8U-

0,C>>

(I)

0 5 10 15 20 25Daysafter CycLophosphamide

Chart 5. Regrowth of Lewis lung tumors following single doses ofcyclophosphamide, 50 to 300 mg/kg. Volumes are relative to the tumorsize at the time of drug administration. Broken lines are drawn from thesurviving fraction measured at 3 hr (Chart 4) to the point at which the

tumor volume had regrown to 4 times the treatment volume.

at which the regrowing tumors reached a leg diameter of 12mm (a tumor weight that is 4 times the treatment size).Measurements of TD50 in such tumors gave values that wereindistinguishable from the value in untreated tumors. Theslopes of the broken lines in Chart 5 can only giveapproximate estimates of the doubling times of the clonogenic cell population during regrowth, but the estimates (indays) are as follows: 1.2 (50), 0.8 (100), 0.75 (150), 0.8 (250),and 0.95 (300), the cyclophosphamide dose (in mg/kg) beinggiven in parentheses. These estimates of doubling time arevery short, but they are not surprising in view of the1.02-day doubling time of the growth curve in Chart 3 andthe work of Simpson-Herren and Lloyd (17) and Hill andStanley (unpublished observations) which has shown thatthe average intermitotic time of cells in small Lewis lungtumors is below 18 hr.

Curability of Small i.m. Implants of Lewis Lung Tumor.For these studies implants of 100 viable cells plus 108 HRwere made into the left hind legs ofgroups of 10 mice. Eachgroup was treated with a single i.p. injection of cyclophosphamide at up to 24 days later. Mice were killed at 60 daysafter implantation or when the tumors exceeded 10-mmdiameter, and the proportion of controlled tumors wascalculated. The results are shown in Chart 6 for cyclophosphamide doses of 100, 200, or 300 mg/kg. The tumor agesat which 50% control was achieved are 4, 11.5, and 16 days,respectively.

These results may be compared with the expected age atwhich tumors should be cured on the basis of the data shownin Charts 3 and 4. For any chosen time after implantation of100 viable cells, we may calculate from Chart 3 thesubsequent size of the implant, having allowed for the factthat a TD50 of approximately 2 cells implies that only 100(ln 2)/2 = 34.6 cells initially take. For any chosen dose ofcyclophosphamide, Chart 4 gives the relative cell survival.

\\ \@@

C ‘0 2 4 6 B@ 12 14 16 lB 20 22

Daysafter implontation

Chart 6. The proportion of 100 cell implants that were cured by a singledose of cyclophosphamidegiven at various times thereafter. Drug doseswere 300 mg/kg (0), 200 mg/kg (A), or 100 mg/kg (El. Broken lines.theoretical expectation (see text).

June 1975 I 533

Association for Cancer Research. by guest on September 3, 2020. Copyright 1975 Americanhttps://bloodcancerdiscov.aacrjournals.orgDownloaded from

G. G. Steel and K. Adams

survival data were obtained on i.m. tumors weighing about0.5 g that already contained areas of necrosis; it isconceivable that cyclophosphamide was less available tosome cells within such tumors than within i.m. implants ofup to only l0@cells. An attempt to test this hypothesis hasbeen made by performing cell survival studies on small lungtumors. These were produced by the simultaneous i.v.injection of l0@ viable tumor cells plus 106 HR plus 106plastic microspheres. On Day 16 after injection, the micewere divided into groups and either left untreated or givenan i.p. injection of cyclophosphamide. Three or 16 hr laterthe mice were killed, the lungs were removed, and lungcolonies of about I- to 2-mm diameter were removed andprocessed for a single-cell suspension as indicated above.Each group of 4 mice yielded about 30 colonies. Theestimates of cell survival obtained in this way are shown inChart 7. The data predominantly fall below the line that wasdetermined for the larger i.m. tumors (Chart 4). Also shownin Chart 7 are the cell survival values that may be calculatedfrom the tumor cure data. From the data in Chart 6 one caninterpolate the tumor age at which 50% tumor control wasachieved. Chart 3 gives the size of the tumors at this age andthe estimate of cell survival comes from the ratio of this sizeto the number of cells that may be assumed to give 50%regrowth probability. For instance, the dose of cyclophosphamide, 200 mg/kg, cured 50% of tumors at 11.5 daysafter the 100 cell implants. Taking a TD50 of 2.0 we assumethat 34.6 cells actually grew, and from Chart 3 it can be seenthat by 11.5 days they would have produced 8 x l0@cells.The further assumption that the TD50 for regrowth liesbetween 0.7 (the ultimate TD50 for the regrowth of allsurviving cells) and 2.0 gives a calculated range of survivalsfrom 8.8 x 10-6 to 2.5 x l0-@.

These calculations show that the tumor cure data areconsistent with a steeper survival curve for cells within small

8U-

2'>

U,

Chart 7. Cell survival in I- to 2-mm diameter lung coloniesfollowingsingle dosesof cyclophosphamide.®,A, results of 2 separategroups ofinvestigators using the in vitro colony assay, 3 experiments in all. Solid lineis taken from Chart 4. Bars, cell survival that would be expected on thebasis of the results shown in Chart 6.

lung colonies than has been determined on the largerdissectable i.m. tumors. We may also conclude that, withinthe rather wide limits of precision of these calculations, acell survival curve determined shortly after drug administration gives a realistic indication of the number of cells thatwill retain the capacity to contribute to tumor regrowth.There is no evidence that subsequent cell kill due toimmunological or other mechanisms was important.

DISCUSSION

The numerous studies on the Lewis lung tumor during thepast few years have largely concentrated upon the responseof its metastases. Humphreys and Karrer (7) and Karrer eta!. (9) described the efficacy of chemotherapy in eliminating metastatic disease after the surgical removal of an i.m.implant; Johnson (8) performed similar experiments usingirradiation treatment of the primary implant. The chemotherapeutic response of s.c. implants was studied by Mayoet al. (10), and the inhibition of metastases by ICRF 159,Triton, and cyclophosphamide has been described byHellman and Burrage (4), Salsbury et a!. (14), and Franchiand Garratini (3). The 1st reports of cell survival studieswith the Lewis lung tumor were by DeWys (1 , 2) who used aquantitative transplantation assay. His results indicatedthat the extent of cell kill decreased with increasing age (orsize) of tumors treated with cyclophosphamide.

The initial objective of this work has been to developreliable methods for measuring cell survival following invivo treatment of Lewis lung tumors. The use of theend-point dilution assay has yielded the important conclusion that TD50 values as low as 1 to 3 cells can be achievedwith this tumor. Bearing in mind that the cells that areassayed have been subjected to the trauma of trypsinizationand that it is difficult to reduce the standard error of TD60values below about 20% of the mean, these results areconsistent with a majority of the extracted cells with thecapacity to initiate tumor growth in the i.m. site. Inconcluding this, it must be recalled that the cell separationprocedure has a yield of only about 10%. The cells that arebrought into suspension may or may not be representativeof the total cell population of the tumor. The greatsensitivity of the end-point dilution assay has allowed us tomeasure surviving fractions after cyclophosphamide treatment down to the limit that is set by host toxicity (a survivalof about lO_6). At this level there was little evidence of a

resistant component in the survival curve of the typeobserved by Valeriote et a!. (21). The 3 assays that we haveused gave cyclophosphamide survival data that were indistinguishable, although the lung colony assay was notadequately tested. We have not pursued this comparison inview of the excellent agreement between in vitro and lungcolony assay estimates of irradiation survival in the Lewislung tumor that has been observed in this laboratory byShipley et a!. (16). In the B16 melanoma we have alsoobserved good agreement between all 3 assays in the studyof survival following cyclophosphamide (G. G. Steel and K.Adams, manuscript in preparation).

Our investigations of regrowth after the administration of

50 1@0@0@0250300Doseof Cycbphosphamide(mg/kg)

CANCER RESEARCH VOL.351534

Association for Cancer Research. by guest on September 3, 2020. Copyright 1975 Americanhttps://bloodcancerdiscov.aacrjournals.orgDownloaded from

Cell Survival and Tumor Control

cyclophosphamide have led to the conclusion that repopulation by clonogenic cells must be very rapid. Although wehave not actually measured the recovery of the clonogenicfraction in this tumor, it would seem that the posttreatmentdoubling time cannot be much greater than I day.

One of the primary objectives of this project was toexamine the relationship between measurements of clonogenic cell survival and the size of microscopic tumors thatcan be cured. In Charts 6 and 7 are shown the theoreticalestimates of the age of implants that should be cured byvarious doses of cyclophosphamide and the estimates of cellsurvival that are implied by the observed tumor cure data.The results show that older (and therefore larger) implantswere cured than would have been expected on the basis ofthe survival curve obtained on dissectable i.m. tumors.Although it is difficult to rule out the possibility that animmune response may have been partly responsible for thiseffect, it is largely explained by the fact that the cyclophosphamide survival curve is steeper in small tumor foci than inthe larger i.m. implants. Evidence on the antigenicity of theLewis lung tumor from other laboratories is of limitedrelevance to these studies, since both the tumor and the micemay show immunological variation. However, both Johnson (8) and Humphreys and Karrer (7) found that whencured mice were reinoculated with live tumor cells theimplants grew as well as in nonimmunized mice.

It is not possible, on the basis of the results reported here,to decide whether the increased effectiveness of cyclophosphamide within small tumor nodules is due to a greatercellular sensitivity or to better drug availability. However,in concurrent experiments in this laboratory Shipley et a!.(16) have observed that the sensitivity to -y-irradiation of thecells within lung nodules of Lewis lung tumor is alsoconsiderably higher than in larger tumors. For nodules up to2 mm in diameter the hypoxic fraction was below l0@ andcells removed and irradiated under well oxygenated conditions in vitro showed a lower D0 than for cells taken fromlarger s.c. tumors.

ACKNOWLEDGMENTS

We acknowledge with gratitude the support of Professor L. F.Lamerton and the adviceof Dr. W. U. Shipley.The in vitro colony-forming assaywas developedby V. D. Courtenay who, together with J. A.Stanley, producedthe in vitro assaydata shownin Charts 4 and 7.

REFERENCES

I. DeWys,W. D. A Quantitative Model for the Study of theGrowth andTreatment of a Tumor and Its Metastaseswith Correlation betweenProliferative StateandSensitivity to Cyclophosphamide.CancerRes.,32: 367-373, 1972.

2. DeWys,W. D. Studies Correlating the Growth Rate of a Tumor andIts Metastasesand Providing Evidencefor Tumor-related SystemicGrowth-retarding Factors.Cancer Res.,32: 374—379,1972.

3. Franchi, G., and Garratini, S. Selective Chemotherapy of CancerMetastases with Triton WR 1339. European J. Cancer, 7: 579—580,1971.

4. Hellmann, K., and Burragc, K. Control of Malignant Metastases by

ICRF 159.Nature,224: 273-275, 1969.5. Hermens,A. P., and Barendsen,G. W. Changesof Cell Proliferation

Characteristics in a Rat Rhabdomyosarcomabefore and after XIrradiation. EuropeanJ. Cancer,5: 173-189, 1969.

6. Hill, R. P., and Bush, R. S. A Lung Colony Assay to Determine theRadiosensitivity of the Cells of a Solid Tumor. Intern. J. RadiationBiol., 15:435-444, 1969.

7. Humphreys, S. R., and Karrer, K. Relationship of Dose Schedules tothe Effectivenessof Adjuvant Chemotherapy.Cancer ChemotherapyRept., 54: 379-393, 1970.

8. Johnson,R. E. Combined Chemotherapyand Irradiation in Ewing'sSarcoma. Frontiers Radiation Therap. Oncol. 4: 195-202, 1969.

9. Karrer, K., Humphreys,S. R., and Goldin, A. An ExperimentalModel for Studying FactorsWhich InfluenceMetastasisof MalignantTumors. Intern. J. Cancer,2: 213-223, 1967.

10. Mayo, J. G., Laster,W. R., Andrews,C. M., and Schabel,F. M.Successand Failure in the Treatment ofSolid Tumors. III. “Cure―ofMetastatic Lewis Lung Carcinoma with Methyl CCNU and SurgeryChemotherapy.CancerChemotherapyRetp., 56: 183-195, 1972.

I 1. Porter, E. H., and Berry, R. J. The Efficient Designof TransplantableTumor Assays.Brit. J. Cancer, 17: 583-585, 1963.

12. Reinhold, H. S., and DeBree,C. Tumor Cure Rate and Cell Survivalof a TransplantableRat Rhabdomyosarcomafollowing X-irradiation.EuropeanJ. Cancer,4: 367-374, 1968.

13. Rockwell, S., andKallman, R. F. Cellular RadiosensitivityandTumorRadiation Responsein the EMT6 Tumor Cell System. RadiationRes.,53:281-294,1973.

14. Salsbury, A. J., Burrage, K., and Hellmann, K. Inhibition ofMetastatic Spread by ICRF 159:SelectiveDeletion of a MalignantCharacteristic. Brit. Med. J., 4: 344—346,1970.

15. Schabel,F. M. In Vivo LeukaemicCell Kill Kinetics and“Curability―in Experimental Systems. In: The Proliferation and Spread ofNeoplastic Cells, M. D. Anderson Hospital and Tumor Institute atHouston, pp. 379—408.Baltimore: The Williams & Wilkins Co., 1968.

16. Shipley. W. U., Stanley,J., Courtenay,V. D., and Field, S. B. Repairof Radiation Damagein Lewis Lung Carcinoma Cells following InSitu Treatment with Fast Neutrons and -@‘-Rays.Cancer Res., 35:932-938, 1975.

17. Simpson-Herren, L., and Lloyd, H. H. Kinetic Parameters andRegrowth Curves for ExperimentalTumor Systems.CancerChemotherapy Rept., 54: 143-174, 1970.

18. Skipper, H. E., Schabel, F. M., and Wilcox, W. S. ExperimentalEvaluation of Potential Anticancer Agents. XXI. Scheduling ofArabinosylcytosine to Take Advantage of Its S-Phase Specificityagainst LeukaemiaCells. CancerChemotherapyRept., ii: 125—165,1967.

19. Steel, G. G., Adams, K., and Barrett, J. C. Analysis of the CellPopulation Kinetics of Transplanted Tumours of Widely-DifferingGrowth Rate. Brit. J. Cancer,20: 784-800, 1966.

20. Sugiura, K., and Stock, C. C. Studiesin a Tumor Spectrum. III. TheEffect of Phosphoramideson the Growth of a Variety of Mouse andRat Tumours. Cancer Res., 15: 38—51,1955.

2 1. Valeriote, F. A., Bruce, W. R., and Meeker, B. E. Synergistic Actionof Cyclophosphamideand l,3-Bis(2-Chloroethyl)-l-Nitrosourea on aTransplantedMurine Lymphoma. J. NatI. Cancer Inst., 40: 935-944,1968.

June1975 1535

Association for Cancer Research. by guest on September 3, 2020. Copyright 1975 Americanhttps://bloodcancerdiscov.aacrjournals.orgDownloaded from