COMBINATION CHEMOTHERAPY USING
`CYCLOPHOSPHAMIDE, VINCRISTINE, METHOTREXATE,
`5-FLUOROURACIL, AND PREDNISONE IN SOLID
`TUMORS
`JAMES D. BEARDEN, 111, MD*, CHARLES A. COLTMAN, JR., MD*, THOMAS E. MOON, PHD.',
`
`JOHN J. COSTANZI, MD*, JOHN H. SAIKI, MDb, STANLEY P. BALCERZAK, M D " , SAUL E.
`RIVKIN, MD**, FRANCIS S. MORRISON, MDV, MONTAGUE LANE, M D t t , STUART C. SPIGEL,
`MDss
`
`Three hundred and ninety-eight patients with disseminated solid tumors other
`than breast cancer, were treated with a combination chemotherapy protocol
`utilizing cyclophosphamide, vincristine sulfate, methotrexate, 5-fluorouracil,
`and prednisone. Three hundred and eighty were evaluable (95.5%). Partial or
`complete tumor regressions were noted in 73 of 380 (19%) evaluable patients.
`Response to therapy was associated with a prolongation and survival. The
`largest tumor categories were lung, ovary, and gastrointestinal. The proportion
`of complete plus partial responses in evaluable lung cancer patients was 40/236
`(17%), compared to 20/44 (45%) for ovarian cancer patients and 6/39 (15%) for
`gastrointestinal tumors. Of the patients who could be evaluated for toxicity,
`47% had minimal or no toxicity, 51% had moderate to severe toxicity, and 2%
`had life threatening toxicity. Virtually all patients were treated and managed as
`outpatients.
`
`Chwcr 39:21-26, 1977.
`
`T sulfate, methotrexate, 5-fluorouracil, and
`
`HE COMBINATION OF CYTOXAN, VINCRISTINE
`
`prednisone was initially reported by Cooper in
`
`* Wilford Hall USAF Medical Center, Lackland AFB,
`TX.
`' I'niversity of Texas at Houston, fv1.D. Anderson Hospi-
`tal and Tumor Institute.
`' The University of Texas hledical Branch at Galveston.
`5 liniversity of New Mexico School of Medicine, Albu-
`querque.
`' Ohio State University Hospitals, Columbus.
`** Tumor Institute of the Swedish Hospital Medical Cen-
`ter. Seattle, W.4.
`University of Mississippi hledical Center, Jackson, MS.
`Baylor College of Medicine, Houston, TX.
`Keesler USAF Medical Center. h4S.
`Supported in part by Public Health Service Grants CA-
`03906. 12212, 04920, 03392, 12014, 03754, 03392, 04919,
`04915. 03096. 12644, 10379. 16385, 12213. 16957. 10187,
`03389, 13238, 13643, 05587, and 16943 from the National
`Cancer Institute, National Institutes of Health, Department
`of Hralrh, Education, and Welfare.
`Drugs were supplied by Cancer Therapy Evaluation, Di-
`vision of Cancer Treatment, NCI
`This manuscript is a n analysis of SWOG 972 conducted
`from November 1970 to October 1973 by members of the
`Southwest Oncology Croup
`Address for reprints: Southwest Oncology Group, Rose-
`dale State Bank Bldg., Suite 100, 3500 Rainbow Boulevard,
`Kansas City. KS 66103.
`Received for publication April 30, 1976.
`
`the therapy of 60 patients with disseminated
`breast c a r ~ i n o m a . ~ Eighty-eight percent of the
`patients achieved a complete response lasting an
`average of 10 months. The dosage of these drugs
`was such that 45% of the patients had marked
`leukopenia and two drug deaths were reported.
`A number of studies have been published us-
`ing various modifications of the original regimen
`in breast cancer. Ansfield used a modified
`Cooper regimen to test whether the same degree
`of effectiveness could be achieved with less tox-
`icity.' Seventeen of the 18 patients could be eval-
`uated and 11/17 had objective responses (3
`complete), a 65% response rate. The average
`duration of response was 6 months; less toxicity
`was evident with no drug deaths. Spigel el al.,
`using the same modified 5 drug combination,
`treated 23 patients with disseminated breast
`carcinoma." Ten of 23 responses (4 complete)
`were obtained. The median duration of response
`was 8 months.
`Kaufman et al., using another modified
`Cooper regimen, reported on 42 patients with
`disseminated breast carcinoma with 23 respond-
`ers (55%).7 There were 3 deaths attributed to
`drugs. The median survival of the patients who
`responded was 13 months. Lee et al. evaluated a
`21
`
`WCK1044
`Page 1
`
`

`
`22
`
`Hemoglobin (m%)
`White blood cells
`(granulocytes/L)
`
`Platelets/L
`
`TABLE I. Hematopoietic Toxicity Criteria
`None
`Mild
`Moderate
`9.0-9.9
`7.0-8.9
`2 10.0
`> 4,000 (1500)
`t 1,000,000
`
`2,000-2,999
`(< 1,000)
`50,000-74,999
`
`3,000-3,999
`(< 1500)
`75,000-99,999
`
`VOl. 39
`
`Severe
`5.0-6.9
`
`1,000-1,999
`(< 500)
`25,000-49,979
`
`Life threatening
`< 5.0
`< 1,000 (25)
`< 25,000
`
`similar regimen in 32 breast cancer patients
`with 6 complete and 9 partial responses (42%).'
`The most recent modification used four drugs
`in combination (vincristine was exdbded). Can-
`ellos et al. reported that 17 of 25 patients with
`advanced breast cancer achieved a response.'
`Seven of 25(28%) were complete responders.
`The median duration of response was 9 months.
`At the time of publication, the median survival
`of the responders had not been reached but was
`calculated in excess of 13 months.
`In view of the extensive positive experience
`with these various combinations in breast can-
`cer, the present study was designed to examine
`the influence of the modified five-drug program
`in disseminated tumors other than breast can-
`
`MATERIAIS AND METHODS
`
`Between hovember 1970 and October 1973,
`398 patients with disseminated tumors other
`than breast cancer were treated with cytoxan,
`vincristine, methotrexate, 5-FU, and prednisone
`by the Southwest Oncology Group.
`The drug dosages were: cyclophosphamide
`i bo mg, daily, orally; vincristine sulfate, 0.025
`mg/kg, or a maximum of mg, intravenously,
`weekly; methdtrexate, 0.5 mg/kg, intravenously,
`weekly; 5-fluorouracil, 10 mg/kg, br a max-
`imum of 500 mg, intravenously, week1 ; and
`prednisone 45 mg/dAy
`mg/day for two weeks, and 15 mg/day for the
`remainder of an 8-week period with a reduction
`of 5 mg/day each week until the dosage was 0.
`
`for two wee E s, 30
`
`SWlVAL CUM
`
`lVl'AL FAIL
`236 139
`
`0 LLKj
`
`144
`
`64
`
`A WAUirOIHIIl
`
`FIG. 1. Cumulative survival curves
`comparing lung vs ovary and other
`tumor types.
`
`WCK1044
`Page 2
`
`

`
`No. 1
`
`COMF-P I N SOLID TUMORS Bearden et al.
`S!RVIV& CUM - LUG, O V M + GIER
`
`23
`
`FIG. 2. Cumulative
`survival
`curves by response comparing lung
`cancer vs ovary and other tumor
`types.
`
`b- = 0
`0 I p.
`
`Following eight weeks of therapy, vincristine
`was given every other week while methotrexate
`and 5-fluorouracil were administered on a
`weekly basis. Cyclophosphamide therapy was
`also continued at 100 mg/day orally.
`Tumor measurements; complete blood cell
`and platelet counts were performed once weekly.
`Renal and hepatic function, metastatic bone
`survey, and liver scan were monitored every two
`months. Progressive disease or death were end-
`points for chemotherapy.
`Toxicity criteria are those established by the
`Southwest Oncology Group. Hematopoietic tox-
`icity criteria are shown in Table 1. If the white
`blpod cell (WBC) feli below 3,00O/cu mm or
`platelet count below lOO,OOO/cu mm, therapy
`with the cytotoxic drugs was discontinued for
`one OF more weeks until recovery of counts above
`these levels occurred. Vincristine therapy was
`discontinued when patellar reflex was lost. No
`formal antacid regimen was specified.
`Criteria for response are defined as follows:
`increasing disease-50%
`increase in the size of
`measured lesions; no response-no
`change in
`
`size of measured lesions after three courses;
`least a 50% decrease in size
`mixed response-at
`of one or more lesions with increase in size of
`others; partial response-at
`least a 50% de-
`crease in the sum of the products of two per-
`pendicular diameters of all measured lesions;
`complete response-disappearance
`of all le-
`sions A new lesion was regarded as progressive
`disease.
`Chi-square tests were used in testing for the
`difference in response rates. Survival curves were
`calculated using the method of Kaplan and
`Meier.' The statistical significance of differences
`between survival curves was tested by the gener-
`alized Wilcoxon test.'
`
`RESULTS
`Three hundred and eighty of the 398 patients
`are considered in this report. Eighteen were not
`evaluable; 12 due to insufficient data, five be-
`cause they were not treated according to the
`protocol, and one patient committed suicide af-
`ter being on study two weeks.
`
`WCK1044
`Page 3
`
`

`
`24
`
`C A N C E R JQ7lUQYy 1977
`
`Vol. 39
`
`-
`
`1
`
`-
`
`1
`
`
`
`I
`
`1
`
` I
`
`I
`
`I
`
`I
`
`I
`
`1
`
` 1 - m
`
`0 c, P 0 0 (1 W (1- v 0. (1 r- - 0
`
`m o e m - c, N d - -
`N
`
`N
`
`TABLE 3. Overall Toxicity
`Severity
`Number
`80
`68
`133
`28
`6
`65
`
`None
`hlild
`hloderate
`Severe
`Life threatening
`Unknown
`
`?h
`(25)
`(22)
`(42)
`( 9)
`( 2)
`
`Table 2 shows the response rates of the differ-
`ent tumor types. There is not a statistically sig-
`nificant difference in the proportions of com-
`plete response or complete plus partial
`responses among the different lung cell types.
`However, the proportion of complete plus par-
`tial responses in lung patients is 40/236 (16.8%)
`compared to 20/44 (45.4%) for ovary patients.
`This is a statistically significant difference (P =
`.01). There were six complete and I t partial
`responses in the 25 (68%) ovarian carcinoma
`patients who had no prior alkylator therapy
`while only 3/19 (16%) with prior alkylator ther-
`apy showed response.
`The median time from onset of treatment to
`maximum response is 45 days for complete re-
`sponders and 29 days for partial responders.
`The median duration of complete response is
`194 days compared with 153 days for partial
`responders. The difference between these curves
`is not statistically significant (P = .46). Figure 1
`shows the survival of all lung patients (236)
`compared to the survival of all patients without
`lung cancer (144 patients with ovary and other
`disease sites). The median survival of all lung
`cancer patients was 124 days or 4.13 months
`and 221 days or 7.37 months for all other
`patients. There was a statistically significant
`difference between these two survival curves (P
`= .02). The survival by response for all patients
`without lung cancer is shown in Fig. 2. While no
`median survival has been reached for the
`patients achieving a complete response (CR),
`median survivals of 367 days and 146 days were
`obtained for partial responders and patients ob-
`taining less than a partial response (other re-
`
`Ovary + Other
`Complete response
`Complete response
`Complete response
`Partial response
`Partial response
`Partial response
`
`TABLE 4
`Lung
`C R S
`PARS
`OTHKS
`C R S
`P A R S
`OTH RS
`
`P-Value
`
`.17
`.02
`.01
`.37
`.10
`.01
`
`I
`
`WCK1044
`Page 4
`
`

`
`No. 1
`
`COMF-P I N SOLID TUMORS Bearden et al.
`
`25
`
`FIG. 3. Cumulative survival curve
`for all lung cancer patients calcu-
`lated from date of diagnosis.
`
`1
`
`03.03
`
`I30 C Q
`
`W.03
`
`3YJ.03
`
`52003
`
`650.K
`
`780.03
`
`910.03
`
`1WO.M
`
`sponse). There is a significant difference be-
`tween the survival for complete responders and
`other responders ( P = .02) as well as the differ-
`ence between partial responders and other re-
`sponders (P = .04). The survival of complete
`responders and partial responders in Fig. 2 is
`not significantly different (P = .20).
`The survival by response for lung cancer
`patients and ovary plus other patients is shown
`in Fig. 2. The median survival of 167, 145 and
`112 days was computed for lung cancer patients
`who obtained a complete response, partial re-
`sponse, or other response, respectively. The fol-
`lowing pair-wise comparison of survival curves
`was performed, and the associated P = values
`are depicted in Table 4. Although the survival
`curves for complete responders from the ovary
`plus other cancer group appear to be sub-
`stantially better than that of the complete re-
`sponders from the lung cancer group, the asso-
`ciated P = value is only .17. This P value (as all
`P values) is directly related to the small number
`of patients in each group.
`The survival from diagnosis to last follow-up or
`death of all lung patients is shown in Fig. 3. The
`
`median survival from diagnosis is 251 days or
`8.37 months.
`Survival curves for each of the different lung
`cell types are shown in Fig. 4. Note that survival
`is slightly better for patients with oat cell carci-
`noma compared with the other lung cell types.
`There is no evidence of a difference between
`these curves except for the difference between
`oat cell carcinoma and squamous cell carcinoma
`(P = .09).
`Table 3 outlines the degree of toxicity in the
`subjects in which data is available. Two hun-
`dred and eighty-one (89%) of the patients had
`than moderate toxicity, 28 (9%)
`no worse
`patients had severe toxicity, and 6 (2%) patients
`experienced life-threatening toxicity. Of interest
`is the fact that in the 13 patients who had carci-
`noma of the pancreh, six of them (46%) had
`severe toxicity.
`
`DISCUSSION
`In the present study, the median survival from
`diagnosis to last follow-up or death of all lung
`patients is 251 days or 8.37 months. The average
`
`WCK1044
`Page 5
`
`

`
`26
`
`c
`$ :
`
`CANCERJanUUTy 1977
`
`Vol. 39
`
`treated patients noted above. There was a differ-
`ence in median survival of oat cell carcinoma
`and squamous cell carcinoma that approached
`statistical significance.
`Ovarian carcinoma is highly responsive to
`single alkylating agents with significant pro-
`longation of survival. Fifty percent responses
`with a median survival of 12 months are seen
`with melphalal alone (9). In the present report
`there were 17 of 25 responses (68%) in patients
`who had received no prior alkylating agents. In
`the 19 patients with prior alkylators, there were
`only three partial responses (16%). The overall
`survival was 12 months in this small series. The
`five-drug combination is superior to melphalan
`used alone for remission induction but is not an
`exciting prospect to salvage patients who have
`failed on prior alkylators.
`The response rate in gastrointestinal malig-
`nancies is not well characterized because of
`small numbers (Table 2). A 26% response rate
`for colon cancer falls into the range of response
`rates reported for single agents. The overall tox-
`icity is evaluated and is not excessive. One hun-
`dred and forty-eight patients had either none or
`mild toxicity (45%). Only 15% could be classi-
`fied as having severe or worse toxicity (Table 3).
`The fact that six of 13 patients with carci-
`noma of the pancreas had severe toxicity was
`surprising and unexplained in three of the six
`patients. The other three of the six had liver
`dysfunction and obstructive jaundice that could
`have magnified the severe toxicity seen with
`met hot rexate.
`
`FIG. 4. Cumulative survival curves for lung by cell type.
`
`"A"&
`
`survival from diagnosis to last follow-up or death
`of lung patients is 239 days.
`The median survival from initiation of ther-
`apy with COMF-P of all 236 patients with lung
`carcinoma in the current study was 124 days
`(4.13 months) which is comparable to the un-
`
`REFERENCES
`1. Ansfield, F. J.. Kamirez, G., Korbitz, B. C., Davis, H.
`L. Jr.: Five-drug therapy for advanced breast cancer: A
`phase 1 study. CCR 55:183-187, 1971.
`2. Canellos, G. P., DeVita, V. T., et al. Cyclical combina-
`tion chemotherapy for advanced breast carcinoma. Br. Med.
`I., 1:218-220, 1974.
`3. Cooper, R. C . : Combination chemotherapy in hor-
`mone resistant breast cancer. Roc. ,4m. Ass. Cancer Res. 10:15,
`1969.
`4. Garland, L. H., Sissin, hi. A , : Results of radiotherapy
`of bronchial cancer. Radiofopy 67 :48-62. 1956.
`i. Gehan. E. A , : .4 generalized wilcoxon test for com-
`paring arbitrarily singly censored samples. Bzometrzka
`52:203-223, 1955.
`6. Kaplan, E. L. and Lleier, P.: Non-parametric estima-
`incomplete observations. J . Am. Stat. Assoc.
`tion from
`
`53:457-481, 1958
`7. Kaufman, S., Goldstein, hl. : Combination chemother-
`apy in disseminated carcinoma of the breast. Surg. Gynecol.
`Obsted. I34:83-6, 1973.
`8. Lee. J. R.I., Abeloff, M. D., el a/.: An evaluation of five-
`drug combination chemotherapy in the managemenr of re-
`the breast. Surg. Gjnecol. Obstel.
`current carcinoma of
`138:77-80, 1974.
`9. Rutledge. R., Burns, B. C.: Chemotherapy for ad-
`vanced ovarian cancer. Am. 3. Obstet. Gynecol. 96:761, 1966.
`10. Spigel, S. C. Coltman. C. A.. Costanzi, J. J.: Dissemi-
`nated breast carcinoma: Treatment with combination
`chemotherapy. Arch. Intern. Med. 132:575-578, 1973.
`1 1. Tinney, W. L). : Clinical features of bronchogenic car-
`cinoma. .Ifuyo Clzn. h o c . 19:354-357. 1944.
`
`WCK1044
`Page 6