Correlation Between Response to Chemotherapy of
`
`Human Tumors in Patients and in Nude Mice
`
`BEF’PINO C. GIOVANELLA. PHD, JOHN S. STEHLIN JR, MD. FACS.
`RANDALL C. SHEPARD, AND LEO J. WILLIAMS JR. MD
`
`Human tumors serially heterotransplanted in nude mice have been tested for their response to che-
`motherapeutic agents. Fourteen melanomas. l4 colorectal carcinomas. and 14 breast carcinomas have
`been used. Each tumor originated in a different patient. The tumors were maintained by serial sub-
`cutaneous transplantation in nude mice. For the experiments in this study. each neoplasm was trans-
`planted under the kidney capsule of 60 to 100 adult nude mice. The areas of the individual tumor
`implants were precisely measured immediately after insertion using a stereo microscope equipped with
`a micrometric ocular. The animals were then divided into groups of six to ten animals each. One group
`was injected daily with saline and served as controls. The mice in the remaining groups were injected
`daily for eight days with one of the following chemotherapeutic agents—Adriamycin (doxorubicin).
`S-fluorouracil. methotrexate. Cytoxan'(cyclophosphamide). Alkeran (melphalan). vincristine. vinblas-
`tine. methyl-CCNU. or BCNU—at optimum doses (the maximum dose tolerated that causes less than
`10% weight loss). Treatment was initiated when the implants were well established, having roughly
`doubled their initial mass. The animals were then sacrificed and the tumors measured again. A drug
`was rated effective only if it inhibited growth of the tumor by 99% or more. The results so obtained
`were compared with the published results of various clinical trials. When the sensitivity of the human
`tumors in the mice was compared with the sensitivity of tumors of the same type that had been treated
`in human patients. a close correlation was found. The panel study detected nine of ten efiective drugs.
`giving only two false-positive results. Our data strongly support the validity of heterotransplants of
`human tumors in the nude mouse as a predictive system for testing new anticancer agents and in
`determining optimal treatment schedules and combinations of known drugs.
`Cancer 52zl146—l 152. 1983.
`
`IN recent years. human tumors heterotransplanted in
`immunosuppressed animals. particularly nude mice.
`have been widely used to assess the antitumor activities
`of various chemotherapeutic agents. Such model sys-
`tems have been most often used to find new anticancer
`chemicals. “3
`
`From the St. Joseph Hospital Laboratory for Cancer Research and
`the Departments ofSurgery and Pathology. St. Joseph Hospital. Hous-
`ton Texas.
`Supported by Contract NOl—67073-TQ from the National Cancer
`institute. and by the Stehlin Foundation for Cancer Research.
`Address for reprints: Beppino C. Giovanella. PhD. Stehlin Foundation.
`777 St. Joseph Professional Building. Houston. TX 77002.
`The authors thank D. R. Coil. A. J. Kozielski. J. T. Mendoza. C.
`Quian. J. A. Robertson. M. O. Smith. and D. M. Vardeman for their
`technical assistance: R. Hada forthe illustrations: P. Robie for editorial
`assistance: B. Harris for helping prepare the manuscript: Dr. Relda
`Cailleau for having supplied us with breast cell lines MDA-MD 23]
`(KlE) and MDA-MB 26l (HUR) and with pleural effusion VAN: Dr.
`Jorgen Fogh for having supplied us with cell line HT-29: Col. Albert
`Leibovitz for having supplied us with cell lines SW-48 (CAS) and SW
`(KEN): and Dr. Robinson. who supplied us with line COO through
`Dr. Caiileau.
`Accepted for publication July 2. 1982.
`
`A large body of results is accumulating. However. the
`validity of such results depends upon the reliability of
`the model employed. In other words. does a particular
`tumor treated with a given chemical respond in the same
`way in the patient and in the nude mice? A valid answer
`to this question can be obtained in two main ways. The
`first. and most direct. method consists oftreating human
`malignancies with an anticancer agent in the patient of
`origin and as heterotransplants growing in the nude
`mice. By comparing the results obtained on the same
`tumor in the two systems.
`it
`is possible to assess the
`similarity ofresults and determine to what extent human
`results can be predicted from results in the mice. We are
`conducting such a study: however. it is necessary to wait
`for the assessment of clinical results in the patient (for
`example. the reappearance of a removed neoplasm) be
`fore this comparison can be made. The second method.
`postulated by Bellet e! at.“ consists of assessing the re-
`sponses to a given anticancer agent ofa panel of human
`tumors of the same histologic type that have been het-
`erotransplanted in nude mice. lfthe response of human
`
`0008—543X/83/IOOI/l I46 $l.15 © American Cancer Society
`
`1146
`
`Genentech 2092
`
`Celltrion v. Genentech
`
`|PR2017-01122
`
`Genentech 2092
`Celltrion v. Genentech
`IPR2017-01122
`
`

`

`No. 7
`
`CHEMOTHER RESPONSE OF TUMORS IN PATIENTS AND NUDE MICE - Giovanella et al.
`
`1 147
`
`Melanoma:
`
`Tumor
`Sex/age
`Location
`histologic type
`
`Previous treatment
`
`Nude
`passages
`
`Melanin
`
`Refs.
`
`TABLE I. Melanomas
`
`
`EBE
`WlL
`FRA
`STE
`BOW
`HOF
`N15
`FOS
`TR]
`VIC
`BAG
`GIL
`FlN
`BRO
`
`F/49
`F/30
`M/43
`F/32
`M/36
`F/50
`M/54
`F/54
`F/45
`M/38
`M/Zl
`M/55
`M/Sl
`M/25
`
`Left lung
`Right forearm
`Left thigh
`Femoral lymph node
`Skin of back
`Skin of leg
`Lymph node. neck
`lliac lymph node
`Skin of shoulder
`Left thigh
`Right foot
`Left groin
`Axillary lymph node
`Left shoulder
`
`Metastatic
`Recurrent
`Recurrent
`Metastatic
`Recurrent
`Recurrent
`Metastatic
`Metastatic
`Primary
`Metastatic
`Primary
`Metastatic
`Metastatic
`Primary
`
`AL. BCG. C. parvum, DTlC
`None
`None
`AL
`BCG. C'. parvum
`BCG. AL
`None
`None
`None
`Rx
`None
`AL. C. purvmn, DTlC
`BCG. Rx
`None
`
`A
`6
`A
`25
`M
`7
`A
`5
`A
`29
`A
`20
`M
`15
`A
`I
`l
`A
`14
`A
`l
`A
`6
`A
`12
`A
`4
`
`25 A
`
`(17)
`
`(17)
`
`A: amelanotic: M: melanotic; AL: Alkeran (melphalan): BCG: Bacillus Calmette Guerin: DTIC: dacarbazine: Rx: radiation therapy.
`
`tumors of the type being studied is known, the response
`of the mice can be compared to the human response.
`By statistical methods. such correlation can be quanti-
`tated. In this article, we are reporting our results using
`this second approach. Gehan5 has demonstrated that if
`at least one in a panel of 14 tumors of the same type
`responds to a given treatment. there is a 95% probability
`that 20% or more of such tumors will respond to the
`agent in question. We have used three panels of 14 tu-
`mors each: one panel of melanomas, one of colorectal
`carcinomas. and one of breast carcinomas.
`
`transplanting human bioptic material directly from the
`patient to the nude mice. The origins and main char-
`acteristics of each tumor are summarized in Tables 1
`
`through 4.
`
`Animals
`
`Swiss nude mice bred and maintained in our labo-
`
`ratory under pathogen-free conditions" were used
`throughout this study. The average age of the mice used
`in the experiments was 3 months.
`
`Materials and Methods
`
`Experimental Procedure
`
`Tumors
`
`The tumors used for these experiments are all human
`tumors. the majority of which were obtained by hetero-
`
`Human tumors serially transplanted in nude mice
`were used. Small fragments of each tumor were trans-
`planted under the kidney capsules of nude mice using
`our modification of the technique of Bogden 01 a/.' Each
`
`Tumor
`
`Sex/age
`
`Previous
`Nude
`
`Location
`Histologic type
`treatment
`passages
`References
`
`TABLE 2. Colorectal Carcinomas
`
`F/63
`F/64
`?/?
`M/29
`F/83
`
`F/44
`
`M/6O
`
`Cecum primary
`Liver metastasis
`?
`Rectum primary
`Colon (cell line)
`primary
`Colon primary
`
`Colon primary
`
`M/50
`F/4l
`M/50
`
`Mod diff adeno. cecum
`Mod to poor diff adeno. rectum
`Mod diff adeno
`Mod difi adeno. rectum
`Very poorly diffadeno
`
`Mucinous (mod to poor diff)
`adeno. colon
`Mucinous (mod difl") adeno.
`colon
`Mucinous (mod diff) adeno
`Mod to poor diff adeno. colon
`Poorly diff adeno
`
`008
`KLO
`FEI
`PEY
`KEN
`
`HT-29
`
`BEG
`
`SQU
`KON
`CAS
`
`ANZ
`MOR
`NOV
`WAL
`
`None
`None
`?
`None
`None
`
`7
`
`None
`
`20
`13
`6
`19
`34
`
`32
`
`16
`
`(18. 19)
`
`(20)
`
`?
`5FU
`None
`
`16
`8
`7
`
`1
`
`(19)
`
`Diaphragm metastasis
`Ovary metastasis
`Colon (cell line)
`primary
`1
`None
`Mod to poor diff adeno colon
`Colon primary
`F/81
`4
`None
`Mod diff adeno. colon
`Liver metastasis
`F/70
`3
`None
`Poorly diff adeno rectum
`Femoral lymph node
`F/64
`
`
`
`
`Anorectal squamous cell ca NoneVaginal spreadF/48 9
`
`Mod: moderately: diff: difl‘erentiated: adeno: adenocarcinoma; ca: carcinoma.
`
`

`

`1 148
`
`CANCER October 1
`
`1983
`
`Vol. 52
`
`TABLE 3. Breast Carcinomas
`
`
`Nude
`
`Tumor
`Sex/age
`Location
`Histologic type
`Previous treatment
`passages
`Refs.
`
`CLO
`KIE
`WAR
`VAN
`
`F/30
`F/Sl
`F/62
`F/49
`
`Left breast primary
`Cell Line pleural elfusate
`Chest wall recurr.
`Pleural effusate
`
`Inf. duct cell ca
`Inf. duct cell ca
`lnf. duct cell ca
`lnf. duct cell ca
`
`None
`5-FU
`Rx
`CTX. AL. Adria. 5-FU.
`MTX
`
`53
`22
`24
`25
`
`(2|)
`
`DRE
`ELL
`HIG
`C00
`ALL
`HUR
`MUR
`JAM
`
`F/65
`F/32
`M/78
`F/?
`F/66
`F/63
`F/40
`F/43
`
`Right breast primary
`Breast primary
`Right chest wall metastasis
`Cell line
`Left breast primary
`Brain metastasis cell line
`Left breast primary
`Left ax. lymph node
`
`Inf. duct cell ca
`lnf. duct cell ca
`lnf. duct cell ca
`Inf. duct cell ca
`Medullary ca
`Inf. duct cell ca
`lnf. duct cell ca
`Inf. duct cell ca
`
`None
`None
`’?
`7
`None
`Multiple chemother
`None
`C‘. parvum, S-FU. CTX.
`MTX. VC, Adria
`8
`None
`Inf. duet cell ca
`Right breast primary
`F/56
`SAW
`
`
`
`Breast primaryF/48WlS 3 Medullary ca None
`
`
`
`(21)
`
`(3)
`
`17
`24
`27
`12
`8
`5
`13
`4
`
`lnf: infiltrating; ca: carcinoma: S-FU: S-fluorouracil: Rx: radiation
`therapy; CTX: Cytoxan (cyclophosphamide): AL: Alkeran (mel-
`
`phalan); Adria: Adriamycin; MTX: methotrexate; VC: Vincristine.
`
`fragment was carefully measured immediately after
`being positioned under the kidney capsule. A micro-
`metric ocular inserted in a stereo microscope was used
`for this operation. The inoculated animals were then
`divided into groups of six to ten animals. Each day one
`control group was injected intraperitoneally with saline,
`0.1 m]. The other groups were treated daily with che-
`motherapeutic agents at the doses shown in Table 5.
`The treatment was initiated when the implants were well
`established, having roughly doubled their initial mass.
`This was determined by killing inoculated animals every
`second day until doubling was confirmed. The mice
`were then injected daily for eight days, killed, and the
`tumors measured again. The tumor mass was calculated
`according to the formula
`
`TABLE 4. Summary of Patients/Tumors in Study
`
`Melanomas
`
`Colorectal
`carcinomas
`
`Breast
`carcinomas
`
`5] (30—78)
`57(41—8l)
`39 (21—55)
`Age (avg/range)
`M: I. F: 13
`M: 8, F: 6
`M28. F26
`Sex
`
`
`
`1-25 2—36Passages 2—29
`
`TABLE 5. Drug Doses in Panel Studies
`
`Adriamycin
`5-FU
`Methotrexate“
`Cytoxan
`Alkeran
`Vincristine
`Vinblastine
`Methyl-CCNU
`BCNU
`
`1 mg/kg/day X 8
`20 mg/kg/day X 8
`8 mg/kg/day X 8
`20 mg/kg/day X 8
`l mg/kg/day X 8
`0.2 mg/kg/day X 8
`0.3 mg/kg/day X 8
`0.4 mg/kg/day X 8
`0.4 mg/kg/day X 8
`
`" Initially MTX was given in doses of 24 mg/kg/day X 8. This was
`gradually scaled down to 8 mg/kg/day X 8 due to toxicity.
`
`Weight (mg) =
`
`
`aXb2
`2
`
`where a is the length of the tumor and b is the width.
`Tumor growth inhibition was calculated by comparing
`the mass of the treated and untreated tumors. A strong
`inhibition (++) was arbitrarily assessed when the treated
`tumor mass was reduced 99% or more in comparison
`with the untreated tumor which served as a control. A
`
`moderate (+) inhibition was assessed with a reduction
`in tumor mass of more than 80% but less than 99%. Any
`reduction in tumor mass of less than 80% was considered
`
`a negative (—) result. Examples of results obtained in
`two typical experiments are shown in Figures 1 and 2.
`
`Results
`
`Our results are reported in Tables 6 through 8. Be-
`cause the clinical reports on the effectiveness of a drug
`in the treatment of a given malignancy do not always
`agree among themselves, we decided to report more than
`one clinical series for each type of tumor studied. We
`have, accordingly, compared our results with the results
`reported in some recent clinical compilations chosen on
`the basis of the number of patients reported and on the
`quantitation of the reports. It must be remembered that
`we are considering effective only those anticancer agents
`that when administered alone give a positive response
`in 20% or more of the treated patients. Consequently.
`we cannot use for comparison any reports that do not
`give the exact percentage of positive responders.
`The results of our series and of the clinical series we
`
`used for comparison are tabulated side by side in Tables
`9 through 11. When the results were borderline, either
`in our series or in the clinical series. the data are ex-
`
`

`

`No. 7
`
`CHEMOTHER RESPONSE OF TUMORS IN PATIENTS AND NUDE MICE - Giovanel/a et a].
`
`1149
`
`BOW P5, V0 = 2.5 Mg
`
`
`VAN P14, V0 = 4.5 Mg
`
`
`vvvvvvvv
`TWO
`
`
`CONTROL
`
`/ ADRIAMYCIN
`
`
`
`............................................... ALKERAN
`————————————————————— VlNCR/S T/NE
`
`
`
`
`
`TWO vvvvvvvvv
`
`°
`
`CONTROL
`
`
`
`
`//
`
`VINCRISTWE
`
`
`
`DAYS
`
`DAYS
`
`FIG. 1. Results obtained in typical experiment for melanoma.
`
`FIG. 2. Results obtained in typical experiment for breast carcinoma.
`
`pressed as the real percentage in parentheses and are
`mentioned as being borderline in the general summary
`(Table 12).
`This is not the place for a morphologic description
`of the tumors used. However‘ some observations made
`
`during this study may help one understand the results
`4, the mel-
`obtained. As reported in Tables 1 through
`anomas and the colorectal carcinomas used i
`n the panels
`did not differ appreciably from the average tumor of
`these types found at random in the human
`population.
`
`TABLE 6.
`Melanomas
`
`
`
`BCNU
`CTX
`MTX
`5-FU
`Adria
`Methyl-CCNU
`VB
`VC
`AL
`
`EBE
`7
`7
`7
`7
`WlL
`+
`++
`+
`++
`FRA
`++
`++
`7
`7
`STE
`7
`—
`7
`7
`BOW
`7
`—
`7
`7
`HOF
`-
`7
`7
`7
`N18
`7
`7
`7
`7
`F08
`7
`7
`7
`+
`TRl
`7
`7
`7
`7
`VIC
`7
`—
`—
`7
`BAG
`7
`7
`7
`7
`GIL
`7
`7
`—
`7
`FlN
`—
`—
`—
`7
`BRO
`7
`7
`7
`
`+
`
`Note: Circle indicates that the patient was treated with the drug in
`question before or after biopsy was taken.
`
`Adria: Adriamycin; MXT: methotrexate: CTX: Cytoxan; AL: Alk-
`eran: VC: vincristine: VB: vinblastine.
`
`

`

`1150
`
`CANCER October I
`
`1983
`
`Vol. 52
`
`TABLE 7. Colorectal Carcinomas
`
`MTX
`
`CTX
`
`AL
`
`VC
`
`VB
`Methyl-CCNU BCNU
`
`
`
`GOB
`KLO
`FE]
`PEY
`CAS
`HT-29
`BEG
`sou
`KON
`KEN
`ANZ
`MOR
`7
`—
`NOV
`—
`7
`7
`7
`WAL
`
`
`7
`+
`
`—
`
`—
`
`—
`
`—
`
`—
`7
`
`7
`
`7
`|
`
`7
`—
`
`7
`
`l
`
`+
`l
`
`—
`++
`
`—
`
`l
`l
`l
`
`Note: Circle indicates that the patient was treated with the drug in
`question before or after biopsy was taken.
`
`Adria: Adriamycin; CTX: Cytoxan; AL: Alkeran; VC: vincristine:
`VB: vinblastine.
`
`TABLE 8. Breast Carcinomas
`
`
`
` AL VC VB Methyl-CCNU BCNU
`
`
`
`
`
`
`++
`++
`++
`—
`7
`++
`++
`++
`7
`7
`+
`+
`+
`7
`—
`++
`—
`++
`7
`7
`
`
`
`CTX
`
`+|
`
`|l||+|++
`
`7
`_
`G)
`7
`—
`
`Adria
`I
`
`5-FU
`
`+
`l
`
`MTX
`l
`l
`
`7
`
`+ll+l
`
`7
`7
`_
`G
`7
`7
`
`+l
`
`7
`
`l++
`
`—
`—
`—
`7
`6)
`7
`7
`
`l||++l
`
`—
`—
`—
`—
`6
`—
`—
`
`CLO
`KlE
`WAR
`VAN
`DRE
`ELL
`HIG
`coo
`ALL
`HUR
`MUR
`JAM
`SAW
`wrs
`
`Note: Circle indicates that the patient was treated with the drug in
`question before or after biopsy was taken.
`
`Adria: Adriamycin: MTX: methotrexate: CTX: Cytoxan: AL: Alk—
`eran: VC: vincristine: VB: vinblastine.
`
`TABLE 9. Results of Panel Studies (Nude Mice/Clinical Series)
`of Melanomas
`
`
` Comis. Carter”Nude mice Luce '4
`
`|
`l
`
`Adriamycin
`5-FU
`Methotrexate
`Cytoxan
`Alkeran
`Vincristine
`Vinblastine
`Methyl-CCNU
`BCNU
`
`at
`
`+I++++I++
`
`|++|
`
`* 96—98% inhibition.
`1“ 19% responders
`i 18% responders.
`A positive (+) result on this table is the equivalent of a ++ result
`on Tables 6—8 and indicates 99% or more inhibition. with the excep-
`tion noted.
`
`That is. these tumors were no more and no less differ—
`
`entiated than 14 tumors of each type chosen at random
`in the human population.
`Breast carcinomas, however, were present only as
`undifferentiated tumors.
`
`Discussion
`
`The first question that we have to answer is whether
`the tumors composing the panels are truly representative
`of the populations on which the clinical research is con—
`ducted. From the standpoint of a pathologist, the answer
`is that both the colorectal carcinomas and the mela-
`
`nomas used in the panels are good representatives of
`such tumors found in the human population in the
`United States. Where the discrepancy between the type
`
`

`

`No. 7
`
`CHEMOTHER RESPONSE OF TUMORs IN PATIENTS AND NUDE MICE
`
`Giovanella et a1.
`
`1151
`
`of tumors prevailing in the panel and in the population
`at large is quite striking is in" the breast carcinomas. The
`tumors in our panel are, without exception, very un-
`differentiated, whereas in the population at large, only
`40% of the total breast carcinomas would be so classi-
`fied.7 This discrepancy is caused by the biological be-
`havior of human breast carcinomas heterotransplanted
`in nude mice. Under optimal conditions from 60% to
`70% of human colorectal carcinomas and melanomas
`
`can be successfully heterotransplanted in nude mice.
`However, only about 20% of breast carcinomas can be
`so transplanted (our unpublished data). The tumors that
`do give positive takes all belong to the less differentiated
`category and, with few exceptions, do not possess estro-
`gen receptors.
`The vast majority of the tumors in the panels had not
`received previous chemotherapy of any type. In partic-
`ular, 10/14 melanomas, 10/14 colorectal carcinomas,
`and 8/14 breast carcinomas had not previously been
`treated with any of the drugs used in our investigations.
`A difficult choice has been the selection of the clinical
`
`results with which we would compare our experimental
`
`results. Not all the clinical trials in the literature agree
`among themselves as to the usefulness of a given drug
`against a certain type of cancer, nor do they agree on
`the percentages of positive results, largely because of the
`different criteria used to assess the effectiveness of che-
`
`motherapeutic agents. We are also limited by the fact
`that we required well defined quantitative results. We
`have chosen the most recent clinical compilations in
`which precise percentages of responders were given and
`in which most of the patients were residents of the US
`(the latter stipulation to minimize possible population
`variations)."'l6
`Because the various clinical series do not always agree
`among themselves, we have reported the figures ob-
`tained from different series separately. When all the clin-
`ical series agree about results, we consider the chemo-
`therapeutic agent effective or not effective against the
`given type of cancer. The clinical series being studied
`agreed 21 times and disagreed 6 times. Taking into ac-
`count the 21 cases of agreement, our results in the nude
`mice agreed with the clinical consensus in 18 instances
`(9 active drugs and 9 inactive drugs). There were two
`false-positives (Adriamycin [doxorubicin] and 5-fluo—
`rouracil [S-FU] in melanomas) and one false-negative
`(Adriamycin in breast carcinomas).
`Given the way the tests were conducted, it is not im-
`possible that the false positives represent tumors that
`really are sensitive to the drug tested. To clarify this. one
`must remember that the concept of responsiveness is a
`relative and necessarily arbitrary one. We have arbi-
`
`TABLE 10. Results of Panel Studies (Nude Mice/Clinical Series)
`of Colorectal Carcinomas
`
`Nude
`Carter,
`Cline,
`Smith
`
`mice
`Friedman'O
`Haskell "
`er a1. '6
`
`—
`—
`—
`—
`Adriamycin
`+
`+
`+
`+
`5-FU
`—
`—
`—
`—
`Methotrexate
`—
`—
`+
`+
`Cytoxan
`-
`—
`—
`—
`Alkeran
`—
`—
`—
`—
`Vincristine
`-
`—
`—
`—
`Vinblastine
`-
`—
`—
`—
`Methyl-CCNU
`
`
`
`
`+ +" —BCNU —
`
`*15% responders.
`A positive (+) result on this table is the equivalent of a ++ result
`on Tables 6—8 and indicates 99% or more inhibition.
`
`trarily decided that 20% of the tumors must respond in
`order to classify the tumor type as responsive to a given
`drug. Even if we keep our limits reasonably elastic, 15%
`to 18% might still be considered positive, but 5% to 10%
`would certainly not be. However, 5% to 10% still gives
`one responding tumor in every 10 to 20, an infrequent
`but not impossible occurrence, especially when a group
`of 14 tumors is being considered.
`
`TABLE 1 1. Results of Panel Studies (Nude Mice/Clinical Series)
`of Breast Carcinomas
`
`Nude
`mice
`
`Carter9
`
`Davis, Carbone”
`mod. by Rubensls
`
`Adriamycin
`5-FU
`Methotrexate
`Cytoxan
`Alkeran
`Vincristine
`Vinblastine
`Methyl-CCNU
`BCNU
`
`|++++t+l
`
`l+++++++
`
`+l+++++++
`
`* 96% and 92% inhibition.
`A positive result (+) on this table is the equivalent of a ++ result
`on Tables 6—8 and. with the exceptions noted, indicates 99% or more
`inhibition.
`
`TABLE 12. Correlations
`
`Breast carcinomas
`
`Colorectal carcinomas
`
`Melanomas
`
`6/7 Positives“
`0 False-positives
`1/1 Positive
`0 False-Positive
`2/2 Positivesr
`2 False-
`Positives:
`
`‘ Methotrexate is borderline in nude mice test.
`1‘ Vinblastine is borderline in Comis & Carter series.
`1 BCNU is borderline in Luce series.
`
`

`

`1152
`
`CANCER October I
`
`1983
`
`Vol. 52
`
`More disturbing is the existence of the false-negative,
`particularly of a drug such as Adriamycin in breast car-
`cinoma, for Adriamycin is undoubtedly active against
`at least 30% of these neoplasms. As we have stated,
`breast carcinoma has limited percentage of positive takes
`(approximately 20%) in heterotransplanting in nude
`mice. Having selected a defined subclass of breast car-
`cinomas that comprises about 20% of such tumors, it
`
`is not inconceivable that we would find such a class
`resistant to a drug that is ineffective against 60% or more
`of breast carcinomas.
`
`Considering the large number of human tumors stud-
`ied, the close correlation between results observed clin-
`ically and those obtained experimentally makes it dif-
`ficult to avoid the conclusion that the human tumor
`
`heterotransplanted in the nude mouse is a good predic-
`tor of the results to be expected in the clinical setting.
`Such parallelism is further strengthened by the episodic
`cases in which the same human tumor has been sub-
`jected to the same treatment in the mice and in the
`patient with the same results.
`
`REFERENCES
`
`l. Bogden AE, Kelton DE, Cobb WR, Esber HJ. A rapid screening
`method for testing chemotherapeutic agents against human tumor xen-
`ografts. In: Houchens DP, Ovejera AA, eds. Proceedings of the Sym-
`posium on the Use of Athymic (Nude) Mice in Cancer Research. New
`York: Gustav Fischer, 1978; 231—250.
`2. Giovanella BC, Stehlin JS, Williams LI, Lee SS, Shepard RC.
`Heterotransplantation of human cancers into nude mice: A model
`system for human cancer chemotherapy. Cancer I978; 42:2269—228 l.
`3. Giovanella BC, Fogh J. Present and future trends in investiga-
`tions with the nude mouse as a recipient of human tumor transplants.
`In: Fogh J, Giovanella BC, eds. The Nude Mouse in Experimental
`and Clinical Research. New York: Academic Press, 1978; 281-312.
`4. BeIIet RE, Danna V, Mastrangelo MJ, Berd D. Evaluation of a
`“nude" mouse human tumor panel as a predictive secondary screen
`for cancer chemotherapeutic agents. J Natl Cancer Inst 1979; 63: I 185—
`1 188.
`
`5. Gehan EA. The determination of the number ofpatients required
`in a preliminary and a follow-up trial ofa new chemotherapeutic agent.
`J Chmn Dis 1961; 13:346-353.
`6. Giovanella BC, Stehlin JS. Heterotransplantation of human
`malignant tumors in “nude“ thymusless mice: 1. Breeding and mainte-
`nance of “nude" mice. J Natl Cancer Inst 1973; 51:615—619.
`7. Hultborn KA, Tornberg B. Mammary carcinoma: The biologic
`character of mammary carcinoma studied in 517 cases by a new form
`of malignancy grading. Acta Radiol [Suppl] (Stockh) 1960; 19621—143.
`8. Giovanella BC, Yim SO, Stehlin JS, Williams LI. Development
`of invasive tumors in the “nude” m0use after injection of cultured
`human melanoma cells. J Natl Cancer Inst 1972; 48:1531—1533.
`9. Carter SK. Chemotherapy in advanced breast cancer. Int J Ra-
`dial Oncol Biol Phys 1978; 4309—31 I.
`10. Carter SK, Friedman M. Integration ofchemotherapy into com-
`bined modality treatment of solid tumors: I1. Large bowel carcinoma.
`Cancer Treat Rev 1974; 1:111—129.
`11. Cline MJ, Haskell CM. Cancer Chemotherapy. Philadelphia:
`WB Saunders 1980;160.
`12. Comis RL, Carter SK. Integration of chemotherapy into com-
`bined modality therapy of solid tumors: IV. Malignant melanoma.
`Cancer Treat Rev 1974; 1:285—304.
`13. Davis TE, Carbone PP. Drug treatment of breast cancer. Drugs
`1978; 16:441—464.
`14. Luce JK. Chemotherapy of melanoma. Semin Oncol 1975;
`2:179—185.
`15. Rubens RD. Breast cancer. In: Pinedo HM, ed. Cancer Che—
`motherapy. New York: Elsevier Press, 1979; 376-41 1.
`1
`16. Smith FP, Bynre PJ, Cambareri RC, Schein PS. Gastrointestinal
`cancer. In: Pinedo HM, ed. Cancer Chemotherapy. New York: Elsevier
`Press, 1979; 292—316.
`17. Giovanella BC, Stehlin JS, Santamaria C el al. Human neo-
`plastic and normal cells in tissue culture: 1. Cell lines derived from
`malignant melanomas and normal melanocytes. J Natl Cancer Inst
`1976; 56:1131—1142.
`l8. Leibovitz A. Development of media for isolation and cultiva-
`tion of human cancer cells. In: Fogh J, ed. Human Tumor Cells In
`Vitm. New York: Plenum Press, 1975;23—50.
`l9. beibovitz A, Stinson JC, McCombs WB 111, McCoy LE, Mazur
`KC, Mabry ND. Classification of human colorectal adenoearcinoma
`cell lines. Cancer Res 1976; 36:4562-4569.
`20. Fogh J, Trempe G. New human cell lines. In: Fogh J, ed. Hu-
`man Tumor Cells In Vitro. New York: Plenum Press, 1975; 1 15-159.
`21. Cailleau R, Young R, Olive M, Reeves W] Jr. Breast tumor
`cell lines from pleural efi‘usions. J Natl Cancer Inst 1974; 53:661—674.
`22. Brinkley BR, BeaII PT, Wible LJ, Mace ML, Turner DS, Cail-
`Ieau RM. Variations in cell form and cytoskeleton in human breast
`carcinoma cells in vitro. Cancer Res 1980; 40:31 18—3129.
`
`