The intergroup rhabdomyosarcoma study.A preliminary report

THE INTERGROUP RHABDOMYOSARCOMA STUDY A Preliminary Report

HAROLD M. MAURER, MD,* THOMAS MOON, PHD, MILTON DONALDSON, MD, CARLOS FERNANDEZ, MD, EDMUND A. GEHAN, PHD, DENMAN HAMMOND, MD, DANIEL M. HAYS, MD, WALTER LAWRENCE, JR., MD, WILLIAM NEWTON, MD,

ABDELSALAM RACAB, MD, BEVERLY RANEY, MD, EDWARD H. SOULE, MD, W. W. SUTOW, MD, AND MELVIN TEFFT, MD

Four hundred and twenty-three children with newly diagnosed rhabdomyosarcoma have been entered to date in the Intergroup Rhabdomyosarcoma Study (IRS), which began in 1972. Patients were classified into Clinical Disease Groups (Stages) I-IV, based on disease extent and resectability, and treatment regimens were randomly assigned according to group. Two hundred and seventy-eight of 423 patients are evaluable for this analysis. Thus far, for Group I disease (localized/completely resected), disease control achieved by vincristine, dactinomycin, and cyclophosphamide (VAC) given in combination for 2 years has not been enhanced by the administration of postoperative radiation to the tumor bed. To date, 92% of patients in both irradiated and nonirradiated groups exhibit no evidence of disease, and 92-96% are still alive, with the median time of follow-up being 72 weeks. For Group I1 disease (microscopic residual/nodal involvement), VAC given for 2 years has not been found to be more effective than vincristine plus dactinomycin given for 1 year, both groups also having received postoperative irradiation. Thus far, over 85% of patients on either treatment have no evidence of disease and 90% are still alive. Survival and complete remission durations range from 1+ to 143+ weeks and the median duration of follow-up is 45 weeks. Chemotherapy as initial treatment has been studied in Group 111 (localized sarcoma not treated initially by gross total resection) and Group IV (metastases present at diagnosis) patients. They have received either intensive “pulse” VAC or “pulse” VAC plus Adriamycin, and radiation has been administered after 6 weeks. Eighty-one percent of patients in Group 111 and 81 -83% in Group IV have responded favorably, with complete regression of disease having been observed in over one-fourth of patients even before the start of radiation and in approximately one-half of all the patients after receiving radiation therapy. There is no indication as yet that one treatment regimen is superior to the other. Seventy-nine percent of patients in Group 111 are still alive (0+ to 154+ weeks) and 69% remain in continuous response (0+ to 139+ weeks) with the median duration of follow-up being 41-44 weeks. Fifty percent of patients in Group IV are still alive ( O f to 127+ weeks) with a median time of follow-up of 41-44 weeks. Tumors arising either from genitourinary sites or the extremities have had a higher incidence of lymphatic spread than tumors in all other primary sites of origin.

CUWC~ 40~2015-2026, 1977.

HABDOMYOSARCOMA (RMS) REPRESENTS pita1 between 1954 and 1966 was 19.2 months, R between 5 and 15% of all malignant solid with 2- and 5-year survival rates of 47% and tumors and 4 to 8% of all malignant diseases in 35%, respecti~ely.~’ None of the 24 children with children under 15 years of age.al.a’ extensive or metastatic diseases at diagnosis sur-

The median duration of survival for children vived as long as 30 months. with RMS treated at the M. D. Anderson Hos- Recent years have witnessed major advances

Representing Chce r and Acute Leukemia Group B, Chil- dren’s Cancer Study Group, and Southwest Oncology Group.

Supported by U. S. Public Health Service Grant Numbers CA 03735, CA 04646, CA 12014, CA 16943, CA 13539, CA

161 18. * Address for reprints: Harold M. Maurer, MD, Medical

College of Virginia, M. C. V. Station Box 787, Richmond, VA 23298.

Accepted for publication March 29. 1977.

2015

201 6 CANCER November 1977 Vol. 40

in the management of RMS. The contributions which are particularly noteworthy are:

1. Delineation of the histologic features of

2. Development of effective chemotherapy, notably dactinomycin, vincristine, cyclophosphamide and adriamycin, first used singly and then in combination in coordinated programs with surgery and radio-

3. Recognition that RMS could be con- trolled or ablated locally in most patients when doses of radiation were delivered in the range of 5,000 to 6,000 rads to sites of involvement. 9 9 4 ~ 2 4 * 2 6 * a

4 . Recognition that sustained chemotherapy for subclinical micrometastases in- creased disease-free survival. lo

5 . Recognition that intensive multidrug systemic chemotherapy could eradicate established metastatic disease. " Based on these developments, in 1972, Acute

Leukemia Group B, Children's Cancer Study Group, and Southwest Oncology Group combined their resources and initiated a prospec- tive intergroup study of childhood RMS.2421 The primary objectives of the study are:

RMS. 28.12.17

therapy, 2.6.7- 11.19- 16,22,29.27,28,30.32,36,96

1. T o determine if local radiation is a nec- essary adjunct to combined chemotherapy [vincristine (VCR), dactinomycin (AMD), and cyclophosphamide (CYC)] for local disease control in patients who have had (pathologically confirmed) complete ex- cision of a localized tumor. 2. To compare the lengths of remission produced by a regimen of VCR and AMD with that produced by VCR, AMD and CYC in patients with grossly resected regional disease but with microscopic residual tumor at the margin of resection, all of whom have received local radiation. 3. To compare the response rates and duration of complete and partial responses resulting from intensive therapy with VCR, AMD and CYC with the same drug combination plus Adriamycin (ADR) in patients with gross residual disease following surgery or with metastatic disease at diagnosis. 4. To collect a group of cases large enough to permit characterization of the disease more thoroughly than was possible here- tofore, and relate multiple factors including age, histologic cell type, site of tumor origin and extent of disease, to treatment responses and prognosis.

Although the study is still in progress, find- ings of considerable clinical interest already have emerged. This report presents these initial results, focusing particularly on the pre- treatment characteristics of the Study Groups, the validity of the clinical grouping (staging) classification, and the response rate, with preliminary results regarding duration of complete and partial responses, and survival produced by each treatment regimen thus far.

METHODS Pathology

The Pathology Subcommittee reviewed the histologic material submitted for each patient registered in the study to determine patient eli- gibility. A patient was eligible if the material could be interpreted as being one or a mixture of the four recognized subtypes of RMS, namely embryonal, alveolar, botryoid and pleomorphic.

Clinical Grouping Classification Since the therapeutic questions posed re-

quired a clinical grouping (staging) classification, a system based on the extent of disease and type of surgery performed was established for all patients, and treatment regimens were randomly assigned to each patient according to Clinical Group. The clinical grouping classification is as follows:

Group I included localized disease, completely resected (regional nodes not involved). a ) confined to muscle or organ of origin and b ) with contiguous involvement-infiltration out- side the muscle or organ of origin, as through fascia1 planes. (This includes both gross in- spection and microscopic confirmation of complete resection.)

Group I1 included a) grossly resected tumor with microscopic residual disease (where the surgeon believes that he has removed all of the tumor, but the pathologist finds tumor at the margin of resection). There should be no evidence of gross residual tumor nor any clinical or microscopic evidence of regional node involvement; b) Regional* disease, completely resected, (all tumor completely resected with no microscopic residual); c ) Regional disease with involved nodes, grossly resected, but with evidence of microscopic residual.

* Regional nodes invol\red and/or extension of tumor into an adjacent organ.

No. 5 INTERGROUP RHABDOMYOSARCOMA STUDY Maurer et al. 2017

Group Regimen Week 0 2 4 6 8 K ) 1 2 ( 4 ( 6 ( 8 p 2 2 2 2 2 6 2 8 3 0 3 2 3 4 3 6 3 8 4 0 4 2 4 4 4 6 4 8 5 0 5 2 5 4 / 5 7 / 2 y r r

00 0000000000

---

8 8 8 8 I

I 1

8 8 8 8 ...x 1 1

000000 0 0 0 0 0 0 000000 000000 000000 000000 8 8 8 8 8

Rpdilion-x

8 8 8 8 1

m, m 0 0 0 000000 00

8 8 8 8 =PO. 1

R d i i - X

900000900000 8 8 8 8

I I A A

mP0. A A

0 Vincristine, 2 mg/m2, I.V. (kloximum single dose, 2 q ) 8 DoctinanyCin,O.O15q/kg/d,LV. (Maximum single jose,0.5mg) x 5 A Adriomycin, 6 0 m g / d , I.V. 0 Cyclophosphamide, 2.5 mg/kg/d, P.O.

Cycbphosphamide, 10mg/kg/d x7

FIG. 1 . Treatment schedules for the Intergroup Khabdomyosarcoma Study. The initial course of cyclophosphamide was reduced from seven to five doses as of May 1 , 1975.

Group 111 included those with incomplete resection or biopsy with gross residual disease. Group IV included patients with metastatic disease present at onset (lung, liver, bones, bone marrow, brain, and distant muscle and nodes).

Study Design The study design and treatment schedules are

illustrated in Fig. 1. Previously untreated patients under 21 years of age are eligible. Patients in Group I are randomized between two treatment regimens: A) VCR ( 2 mg/m2 intra- venously (i.v.) weekly X 12); AMD (0.015 mg/kg/day i.v. X 5; course repeated X 4 in 48 wteks); and CYC (2.5 mg/kg/day p.0. from day 42 up to 24 months); or B) VCR, AMD, and CYC in the same schedule as A, plus postoperative irradiation to the primary tumor site.

Patients in Group I1 all receive radiation to the local tumor site following surgery. In addi- tion, they receive C) VCR (2 mg/m2 weekly X 6 ; six courses over a 48-week period) plus AMD (0.015 mg/kg/day i.v. X 5; with course repeated X 5 in 45 weeks), or D) VCR, AMD, and CYC in the same schedule as in regimen B in Group I above.

Patients in Groups 111 and IV are randomized between E) VCR (2 mg/m2 i.v. weekly X 12); AMD (0.015 mg/kg/day i.v. X 5; course repeated X 4 in 54 weeks); and CYC (10 mg/kg/day i.v. days 1-7 (days 1-5 as of May 1, 1975), course repeated orally days 84-90, then 2.5 mg/kg/day p.0. from 140 to 24 months) (“pulse” VAC); or F) the same drug combination plus Adriamycin (ADR) (60 mg/m’i.v. X 5 in 51 weeks). The first dose of ADR is reduced to 30 mg/m2 if “significant” bone marrow volume is to be irradiated, e.g., the whole retro- peritoneum and whole pelvis. In Group 111 patients, the primary tumor site is radiated after 6 weeks of chemotherapy. In Group IV patients, radiation therapy to the primary and metastatic sites is required only with lung metastases. Other Group IV patients receive radiation therapy at the discretion of the investigator. Surgery may be performed subsequently particularly if nonoperative treatment has been effective in shrinking the primary lesion. The randomiza- tion of patients is based solely on clinical group and not on anatomic site of the primary tumor.

The radiation therapy employed in this study has been standardized as much as possible. Rec-

2018 CANCER November 1977 V O l . 40

ommendations include delivering 5,000 to 6,000 rads in 5 to 6 weeks to the primary site of tumor origin. Radiation to clinically uninvolved regional nodes has not been advised. For children under 3 years of age, it has been advised that they should not receive more than 4,000 rads. Radiation is to be delivered at the rate of 900 to 1,200 rads per week with super-voltage equip- ment. Normal tissue such as lung, liver, kidney, and nervous system requires shielding so as not to exceed tolerance. When metastatic disease is confined to the lung, bilateral pulmonary radiation is given simultaneously with radiation of the primary site. The dose should not exceed 1,800 rads in nine equal fractions. Radiation is to be administered even when there is total regression of all demonstrable lesions from the initial Chemotherapy.

Major deviations in chemotherapy are judged to exist if any of the following apply: failure to begin therapy by the 21st day following the de- finitive operative procedure or within 42 days following the initial operative procedure; use of chemotherapy not specified by the assigned regimen; and failure to administer appropriate dose or schedule of drugs as specified by protocol unless omission or reduction in dosage is neces- sitated by toxicity or acceptable clinical reason.

A major radiotherapy deviation exists if any of the following apply: less than 85% of the prescribed dose is administered; the volume of radiation is too limited for the known extent of the disease; the rate of delivery of radiation is pro- longed excessively; or radiation is commenced prior to the prescribed time without evidence of local tumor progression during the initial chemotherapy.

Evaluation Process Each patient is thoroughly evaluated from the

standpoints of pathology, staging, and treatment received by members of the multidisciplinary Steering Committee. Patients are classified as evaluable or nonevaluable, depend- ing on the degree of variance from the recommendations prescribed in the protocol. Eval- uability requires review of pathology, surgery, radiotherapy, and chemotherapy data by members of the committee for documentation of com- pliance with the recommendations prescribed in the protocol. Nonevaluability status is assigned to patients who have major protocol deviations in chemotherapy or radiotherapy.

Portals and dosage of radiation therapy are evaluated by the two radiotherapists on the Steering Committee. A percentage of or varia-

tion from recommended doses and volumes is made for each patient by reviewing the radiotherapy data including the check list, isodose curves and portal films. Portal films are not routinely requested but are requested when the additional data are required in questionable cases. The details of radiotherapy will be the subject of a separate report.

Response to all treatments received is de- scribed as complete (CR) or as partial (PR). In the latter category, at least 50% reduction of gross disease is required. Other designations of response include : no measurable disease (considered a partial response), where the site of involvement precludes definite measurement of disease extent but where gross residual disease is known to exist at the start of chemotherapy but regresses either partially or completely with treatment; mixed response (MR), where regression and progression of disease are noted simultaneously at different sites; no response (NR) and progressive disease (PD). Additional terms used are as follows: complete remission, absence of any demonstrable tumor; and continuous response, maintenance of the CR, PR status or continued regression of disease in patients in PR status.

Delivery of at least 6 weeks of treatment according to protocol is considered an adequate trial. Deaths occurring within 6 weeks of regis- tration a re considered “early deaths.” Reappearance or progression of disease at any site or the appearance of disease at new sites is considered a relapse and the patient is taken off the study.

Statistical analyses of survival and disease- free periods have been performed using the Kaplan-Meier method for calculating and pro- jecting curves. This method incorporates all of the follow-up information for each patient who has died or relapsed as well as for those patients who are still alive or in remission. Each patient’s follow-up information influences the curve only up to the point of his last contact date or death or relapse. The statistical significance of the differences among curves has been evaluated by the modified Wilcoxin statistic. ‘

RESULTS

A total of 423 patients have been randomized on study between November 21, 1972 and March 8, 1976. Thirty-three (8%) are not eligible, 82 (19%) are considered too soon to eval- uate because of insufficient data, and 308 (73%) are eligible.

No. 5 1 NTERGROUP RHABDOMYOSARCOMA STUDY Maurei et a/ . 2019

Clinical Group Of the 308 eligible patients, 50 (16%) had

localized disease which was completely resected (Group I). Eighty-five patients (28%) had regional disease which could be grossly resected but they had microscopic residual tumor at the margin of resection and/or nodal involvement (Group 11). The group with the largest number of patients (1 11 = 36%) consists of those with gross residual disease following surgery (Group 111); together Groups 111 and IV (metastatic disease, 62 patients = 20%) comprise 56% of the total number of eligible patients.

Age Distribution The frequency of RMS by age and the rela-

tionship between age and extent of disease are shown in Fig. 2. The largest entry rate has been between ages 0 and 6 years and the entry rate decreases thereafter. We have observed a significant relationship between clinical group and age. Patients under age 2 years tend to present in Group I and 11, whereas patients between ages 11 and 15 tend to aggregate in Clinical Groups 111 and 1V. These relationships are highly significant (p = ,005).

Sex and Race Distribution Males are predominantly affected, com-

prising 59% of all eligible cases (M : F ratio, 1 : 41 ). Blacks account for 12% of cases with the remaining cases being classified as nonblack. There is no significant correlation of sex or race with clinical group.

Primary Site of Tumor Origin The distribution of tumors by primary site

and Clinical Group is shown in Table 1. As in most series, lesions in the head and neck consti- tute the largest group (36%) and are mostly unresectable. The other frequent sites of origin are the extremities (23%) and genitourinary region (18%). Tumors arising in these sites are more amenable to local or regional resection than those in the head and neck.

Lymph Node Involvement Two hundred and seven Group 1-111 patients

with sufficient information have been evaluated for lymph node involvement by tumor. Lym- phatic spread has not occurred in patients in this series with nonmetastatic RMS of the orbit (0/17). Primary sites elsewhere in the head and neck region rarely are associated with lymphatic spread (2/62, 3%), including those patients in

9 4

AGE (YEARS)

A11 Eligible Potiants Grwpr I EL P Groupm anz FIG. 2. Relationship between age at diagnosis and clinical

group.

whom radical excisions have been performed (1/19, 5%). Lymphatic spread is also unusual in lesions of truncal origin (including retro- peritoneal space, 3/30, 10%). Although biopsy of regional lymph nodes has not been a routine practice in this study, the limited information available seems to show that grossly resected lesions of the genitourinary tract or extremities have a higher incidence of lymphatic spread than was appreciated previously (10/52, 19%; 8/46, 17%, respectively). The difference in incidence of lymphatic spread from primary lesions of the genitourinary tract including sper- matic cord and extremities (18/98, 18%) compared with other. primary sites of origin (5/109, 5%) is statistically significant (p = .02) when all regionally localized lesions are considered.

Histologic Types As in most series, embryonal and alveolar

variants have been the most prevalent cell types

TABLE 1 . lntergroup Rhabdomyosarcoma Study-Patient Distribution by Tumor Site and Clinical Group

Group I I 1 111 IV Total (%)

Orbit Other head and

neck Trunk Extremities Genitourinary lntrathoracic GI and hepatic Perineum-anus Ketroperitoneum Other

TOTAL

2 5 13 1 21 (7) 3 18 50 17 88 (29)

5 6 7 8 26 (8) 18 25 11 18 72 (23) 21 18 9 8 56 (18)

1 1 1 2 5 (2) 0 2 3 3 8 ( 3 ) 0 4 2 1 7 (2) 0 5 12 2 19 (6) 0 1 3 2 6 ( 2 )

50 85 111 62 308 (100)

2020 CANCER November 1977 Vol. 40

TABLE 2. Intergroup Khabdomyosarcoma Study-Tumor Histology

Percent

Embryonal 57

Botryoid 7 Pleornorphic 2 Special undifferentiated, Type I 5 Special undifferentiated, Type I 1 5

Alveolar 18

Undifferentiated, Type ? 6

(Table 2). Of interest has been the association of the histologic pattern with primary disease site. For patients with trunk lesions, 48% have been classified as alveolar cell type; 38% of the patients with lesions of the extremity have been similarly classified. Also, 78% of the patients with head and neck (including orbit) lesions and 67% of patients with genitourinary lesions have been classified as embryonal cell type. Of the patients classified as botryoid cell type, 75% have had genitourinary lesions.

As tumors were accrued, two consistently re- curring cell types were recognized and coded as “special undifferentiated cells, Types I and II.’’l**E Type I exhibits the morphologic features of extra skeletal Ewing’s sarcoma, while Type I1 more closely resembles large cell Ewing’s sarcoma of bone.

Small undifferentiated mesenchymal cell types that could not be satisfactorily classified into well-identified entities were coded as “Type, Indeterminate” and assigned to one group for the purpose of separate statistical analysis.

Special undifferentiated cells Types I and I1 each account for 5% of the sarcomas. These lesions ,tend to occur predominantly in the extremities. Six percent of the sarcomas could not be classified into well-identified entities and fell into the category of small undifferentiated mesenchymal cell, Type Indeterminate. Whether these subdivisions are clinically important is yet to be determined. The clinical course and response to treatment of these histologic subtypes will be the subject of a separate report.

Treatment Results Of the 308 eligible patients, 278 (90%) are

evaluable for this analysis and 30 (10%) are not evaluable because of major protocol violations in radiotherapy (7 patients), chemotherapy (10 patients), early death (8 patients) and incorrect clinical grouping ( 5 patients). Of the eight early deaths, six were due to toxicity, one was due to a

probable brain hemorrhage and one occurred because of advanced metastatic disease. Thir- teen patients are not included in the tabulation of evaluable patients because they were in- correctly grouped at the time of the randomiza- tion, based on later review of all the data.

Clinical Group I (Disease Localized and Completely Resected)

By definition, these patients are considered to be in complete remission from the date entered on study since the only known disease was resected completely. The median duration of follow-up for these patients is 72 weeks. Two of the 24 patients (8%) who received no postoperative radiation of the primary site have relapsed, at the local site; and one of them has died. One of the 13 patients (8%) who received postoperative radiation developed metastatic disease and died. Duration of complete remission ranges from 1 to 144+ weeks with 19 patients continuing for more than 60 weeks. At least 80% of the patients in each treatment arm are projected to remain disease-free at 60 weeks by actuarial analysis. Local recurrence rates (8%) and duration of disease-free interval in the irradiated and nonirradiated groups are not statistically different at this time (p > .49). There is also no significant difference in survival from the start of chemotherapy (p = .38), with 19 patients alive beyond 90 weeks.

The disparity between the number of patients (24) who received no postoperative radiation compared with the number of those patients ( 13) who received postoperative radiation is related to specific clinical reasons why radiation could not be administered in several of the patients (e.g., radiation therapy refused by patient).

Five patients in Group I (5/42) had amputa- tions performed for primary lesions involving an extremity or had hemiscrotectomies for lesions in the paratesticular area and were assigned Regimen A, omitting postoperative irradiation to the tumor bed, as there was no site to irra- diate. From the standpoint of the study, these patients are evaluated separately and are not included with these patients. All five patients in this group continue in complete re‘mission for a range of 49 to 136 weeks.

Clinical Group I1 (Regional Disease, Re- sected)

The duration of disease-free interval and survival for Group I1 patients ranges from 1+ to 143+ weeks with 27 patients continuing in com-


TABLE 3. Intergroup Rhabdomyosarcoma Study-Response Duration by Clinical Group and Treatment For Evaluable Group 111 Patients (Clinical Group Determined by Institution)

Median Weeks Clinical No. still *Duration time of No. still Duration to C R Group Number No. i n C R ofCR follow-up No. PR in PR of PR median

treatment evaluated C R (%) (%) (wk) range (wk) (% ) (%) (wk) (range)

111-E 48 25(52) 20(80) 0+-139+ 44 14(29) 9(64) 3-71 8(0-34) 111-F 47 22(47) 15(68) 20-125+ 41 16(34) 9(56) 0-56 7(0-22)

p = .33 p- = .18 p = .28 p = .38 p = .78 TOTAL 95 47 35 30 18

* Median has not yet been reached

plete remission for more than 60 weeks. Group I1 patients are also considered to be in complete remission from the date they begin chemotherapy. The median follow-up for these patients is 45 weeks. Four of 33 (12%) who received VCR and AMD for 1 year have relapsed, (one local, three metastatic), while 6 of 39 (15%) who received VCR, AMD and CYC for 2 years have relapsed (four local, two metastatic). The difference in relapse rates is not significant (p = .48), nor is there a significant difference in disease- free interval (p = .91) or survival (p = .55). At least 75% of patients in each treatment arm are projected to remain disease-free at 60 weeks, and the projected survival is greater than 80% for the same time period.

Although the number in each subgroup is still small, no significant differences in relapse rate among the Clinical Group I1 subgroups (IIa, Ilb, and IIc, i.e., regional resections with or without microscopic residual or nodal involvement) have emerged as yet.

Clinical Group I11 (Localized Disease, Not Completely Resected)

The response to Regimens E and F in Group 111 patients is shown in Table 3. Twenty-five of 48 patients (53%) receiving Regimen E (“pulse” VAC) have achieved a complete remission and 14/48 (29%) have achieved a PR. For those receiving Regimen F (“pulse” VAC plus ADR), 22/47 (47%) have achieved a complete remission and 16/47 (34%) have had a PR. Thus, 81% (77/95) of the patients on Regimens E and F have responded favorably. The median time to complete remission is almost equal for both regimens (8 weeks for Regimen E and 7 weeks for Regimen F). The median duration of follow-up is 41 to 44 weeks with 15 patients continuing in complete remission for at least 60 weeks. Twenty-eight and 32% of complete remissions have been achieved at 6 weeks, prior to the start of radiation in Regimens E and F, respectively.

After obtaining a complete remission, 5/25 (20%) patients in Regimen E and 7/22 (32%) in Regimen F have relapsed. This difference is not statistically significant (p = .33). Five of 14 (36%) and 7/16 (44%) patients in partial response in Regimens E and F, respectively, have relapsed. This difference is not statistically significant (p = .28).

Nine of 10 relapses in Regimen E involved local recurrences or extension into adjacent sites; one was due to distant metastases. In con- trast, relapses in Regimen F have been essen- tially evenly divided, with metastases in 8/14 patients and local recurrence or adjacent extension in 6/14 patients.

As yet, there is no significant difference in either length of complete remission (p = . l B ) , ranging from O+ to 139+ weeks, or survival, ranging up to 154+ weeks, between Regimen E and F (Fig. 3). Although 33 patients have survived for at least 60 weeks, 7 of 48 patients (1 5%) in Regimen E (plus two early deaths) and 13/47 (28%) in Regimen F (plus three early deaths) have died (p = .65).

Clinical Group IV (Metastatic Disease Pres- ent at Diagnosis

Clinical Group IV patients’ response are shown in Table 4. Thirteen of 26 (50%) patients who received Regimen E (“pulse” VAC) have achieved a complete remission and 8/26 (31%) have obtained a PR. Thirteen of 30 patients (43%) obtained a complete remission in Regi- men F (“pulse” VAC plus ADR) and 12/30 (40%) have obtained a PR. Thus, 81% and 83% of patients with metastatic disease at diagnosis have responded favorably to Regimens E and F, respectively. These differences are not statistically significant (p = .25). The median duration of complete response is 37.4 weeks (33.7 weeks for Regimen E and 42.8 weeks for Regimen F), and eight of these patients continue in complete response for more than 60 weeks. Although pa-


.40

I .oo .90

.80

.70

CR-RELAPSE / LFU

0.00 20 40 60 80

TRT START-DEATH / LFU

1 TO$ fATHS oVAC* A VAC-

1 I I 1 I

XRT ADR. XRT

P = .65

TOTAL FAIL 25 5 0 VAC*XRT (E) 22 7 h VAC-ADRiXRT (F)

P8.18 FIG. 3. Duration of complete response (upper graph) and survival (lower graph) by regimen for Clin- ical Group 111 patients.

QOO 2 0 40 60 80 WEEKS

tients who achieve a complete remission on “pulse” VAC plus ADR have done so earlier than those achieving complete remission on “pulse” VAC alone (medians: 5 versus 11 weeks), this difference is not statistically significant (p = .33). As among Group 111 patients, 23% and 31% of complete remissions, respectively, have occurred prior to the start of radiation to the primary site.

There is no statistically significant difference in relapse rates between regimens: 9/13 (69%) on Regimen E and 7/13 (54%) on Regimen F have relapsed after obtaining a complete remission (p = .39). All eight patients who obtained a PR in Regimen E have relapsed, whereas only half (6/12) of the patients with PR in Regimen F have relapsed.

As in Group 111, better local disease control is

obtained with Regimen F than Regimen E. In Regimen E, 8/17 patients (47%) have recurred locally, 8/17 (47%) have recurred distantly, and 1/17 (6%) have recurred both locally and distantly. Four of 13 patients (31%) in Regimen F have had local recurrences, 81 13 (62%) have had metastases, and 1/13 (8%) have had both local recurrence and metastases.

So far, there is no significant difference in either length of complete remission, ranging from 1+ to 127+ weeks, or survival, ranging from O+ to 142+ weeks, between Regimens E and F (Fig. 4). The median length of follow-up is 41 to 44 weeks. Sixteen of 26 patients (62%) who received Regimen E (plus one early death), and 12/30 (40%) who received Regimen F (plus two early deaths) have died (p = .54). Nineteen patients have survived for more than 60 weeks.

TABLE 4. Intergroup Rhabdomyosarcoma Study-Response Duration by Clinical Group and Treatment For Evaluable Group IV Patients (Clinical Group Determined by Institution)

Duration of Median Weeks Clinical No. still CR (wk) time of No. still Duration toCR Group Number No. in CR median follow-up follow-up No. inPR ofPR median

treatment evaluated CR(%) (%) (range) (wk) PR(%) PR (70) (wk) (range)

IV-E 26 13(50) 4(31) 33.7 41 8(31) 4-40 1 1 (1-26)

IV-F 30 13(43) 6(46) 42.8 44 12(40) 6(50) 6-74 5 (0-100)

TOTAL 56 26 10 44 20 6

(4- 127 + ) (1+-113+)

p = .34 p = .75 p = .25 p = .21 p = . 3 3


U I W z 4 u v)

u,

.30 TOTAL FAIL

13 9 o VACAXRT ( € 1 13 7 A VAC-ADRZXRT ( F )

Pa.75 0.00 20 40 60 00

TRT START- DEATH/LFU 1.00

(3 .90

1 > .?O a

0

z .BO

3 .60 v)

.50

I- .40 27 17 0

32 14 A

3 O J

VAC'XRT (€1 VAC-ADR'XRT ( F )

P8.54

a L 1

0.00 20 40 60 80

WEEKS FIG. 4. Duration of complete response (upper graph) and

survival (lower graph) by regimen for Clinical Group IV patients.

Lengths of survival from the time of disease recurrence for each clinical group is of interest. The median length of survival after disease recurrence is surprisingly similar for Clinical Groups I1 through IV. For Group I1 it is 12 weeks, 17 weeks for Group 111, and 16 weeks for Group IV. Even though the number of patients in Group I and I1 who have relapsed is still small, the data currently suggest that recurrent disease is generally resistant to treatment.

The validity of the clinical grouping classification for RMS employed in this study is sub- stantiated by the curves shown in Fig. 5 . The projected survival curves for all patients of each clinical group are shown and compared. There is a statistically significant difference in survival between each pair of curves (all p < .05) except that between Groups I1 and 111. The median duration of follow-up ranges from 41 to 72 weeks. These data clearly indicate that survival decreases substantially as the clinical group progresses from I through IV.

Toxicities Table 5 summarizes the toxicities of each

treatment regimen for all eligible patients. Only

I .oo

90

.00

(3 .70

> 5 .60

v)

z

3

8 40

8

3 .so 52 I-

n

30

20

.I0

0.00

nTAL DEBT- \ 42 2 o GROUP I \

0.00 22 44 66 88 110 132 154

WEEKS FIG. 5. T h e survival durations for each Clinical Group of

patients are compared, (Survival duration equals the interval between treatment start and death or last follow-up date.)

the worst toxicity experienced by each patient is recorded. Severe toxicity (primarily transient leukopenia) has been reported in six Regimen A and four Regimen B patients. Of the 50 eligible Group I patients, none has had a life-threatening or fatal toxicity. Of the 85 eligible Group I 1 patients, two Regimen C and two Regimen D patients have had life-threatening toxicity, pri-

TABLE 5. Intergroup Rhabdomyosarcoma Study-Most Severe Toxicity By Treatment (Eligible Patients)

Degree of toxicity Kegi- Moder- men None Mild ate Severe L.T.* Fatal Total

A 3 2 11 8 0 0 2 4 R 3 1 9 6 0 0 1 9 c 8 8 12 12 2 0 42 D 9 2 12 17 2 1 43 E-3 8 2 8 30 6 2 56 F-3 5 3 5 35 5 2 55 E-4 4 2 6 14 4 0 30 F-4 2 0 8 16 5 1 32 G' 1 5 0 1 0 0 7

* Life-threatening. + Amputees or hemiscrotectomized patients assigned

Regimen A.

CANCER November 1977 Vol. 40 2024

marily infection; one patient died of bleeding presumably due to thrombocytopenia. In Group 111 and IV patients there have been 20 life- threatening complications and 5 deaths. The five (2 Regimen E, 3 Regimen F) fatalities were early deaths related to leukopenia and sepsis. None of the Regimen F patients had received ADR, since death took place prior to week 5. There was no obvious difference in frequency or severity of toxicity in Group 111 or IV patients between Regimens E and F. Overall, 66% of patients treated with Regimen E and 80% treated with Regimen F have had toxicity of at least a sevefe degree. No toxic deaths have occurred on Regimens E and F since the initial course of CYC was reduced by protocol modifi- cation from 7 to 5 doses (May 1, 1975). Adria- mycin cardiac toxicity has not been a significant problem; only two cases have been reported. One child had a transient arrhythmia after the first dose and the other exhibited EKG changes after receiving 180 mg (two doses) of the drug. Neither had received prior radiation therapy to the lungs or mediastinum.

R habdomyosarcoma is an extraordinarily dif- ficult tumor for which to study therapeutic programs. It can arise virtually anywhere and each site of origin has its own special management problems. The biologic characteristics of the disease are still incompletely understood because of the relative infrequency of this tumor. Knowl- edge of these characteristics will undoubtedly improve treatment results.

In review of the large collection of histologic material obtained in this study, two additional subtypes of RMS were identified and coded as special undifferentiated Type I and Type 11. The morphologic characteristics of each are suf- ficiently well defined and they can be identified independently by more than one reviewer. There is some clinical difference in primary site predilection of these two subtypes when compared with site predilection in children with RMS of other cell types. Although it is pre- mature to generalize, the limited data available suggest that these tumors are as sensitive to therapy used in this study as the classic forms of RMS.

Earlier reviews of childhood RMS yield little information regarding lymphatic spread of the disease, probably due to the limited size of each series being reported. 1'31 Although the data from the IRS are still incomplete, it seems that

the incidence of lymphatic spread from genitourinary and extremity sites is high enough to indicate the need for an aggressive staging approach to RMS in these sites." At this time we would suggest regional lymph node biopsy for all children with RMS arising in the extremities when the regional lymph nodes are not part of the primary resection. We would also encourage routine regional lymph node dissection for RMS in all genitourinary primary sites especially when the primary treatment is surgical resection. The therapeutic advantage of lymph node dissection in patients with RMS remains to be established; recognition of such spread for accu- rate staging and therapy deserves attention and will be the subject of a separate report.

That the clinical grouping classification employed in the IRS has prognostic validity is confirmed by the significant fall in survival as the clinical group progresses from I through IV. The method of clinical grouping also appears to be well understood by investigators, as there is agreement between the review committee and the investigator grouping in over 90% of cases. It is hoped that this grouping scheme will be widely accepted by others treating RMS, to allow for future comparisons of therapeutic results among institutions using different therapeutic programs. The lack of uniformity in staging among the previously reported studies has pre- cluded this kind of comparison.

The preservation of maximum function is an important aim of therapy. Each modality has its own undesirable sequelae. The IRS addressed itself to refining therapy for Group I patients. We assessed the need for postoperative local irradiation as an adjunct to chemotherapy in patients free of any known residual disease. Thus far, 92% of patients in both irradiated and nonirradiated groups remain disease-free at 60+ weeks and local disease recurrence rates are not significantly different as yet. Thus, radiation therapy does not seem to influence local disease control when patients received VCR, AMD and CYC. However, more time is needed for accrual and follow-up of Group I patients before a statistically valid conclusion can be reached and we can recommend that local irradiation be deleted from the treatment plan for this category of pa- t ients.

For Group I1 patients, there is no indication as yet that the three-drug regimen that includes CYC administered over a 2-year period is superior to the 1-year intensive two-drug regimen consisting of VCR and AMD. Eighty-five percent or more of patients remain disease-free in

both groups and over 87% are still alive. It is our belief that further patient accrual and follow-up is required to give confidence to these early find- ings.

Group 111 and IV patients are randomized to receive either the more intensive “pulse” VAC or “pulse” VAC plus ADR. Both combinations are remarkably effective in inducing remission of gross disease, producing favorable responses in over 80% of patients. Complete regression of disease has been observed in approximately half of the patients of these two groups and in over one quarter even before the start of regional radiation therapy. s8 This gives some encourage- ment to the future use of chemotherapy as the initial treatment method for a broader range of lesions. There is still no indication that within Clinical Groups I11 and IV, one chemotherapy treatment program is superior to the other.

Seventy-five of 85 (79%) patients in Group 111 are alive and 53/95 (56%) remain in continuous response (CR, PR). These 53 patients in continuous response represent 69% (53/77) of the total number of responders in this group.

Twenty-eight of 56 (50%) patients in Group IV are alive and 16/56 (29%) continue to main- tain their response status (CR, PR). Of the total number of Group IV patients who responded (46/56), 16/46 (35%) remain in response.

Significant therapeutic problems continue for

Groups 111 and IV patients. Although “cure” is now possible for patients with localized gross residual or metastatic disease, a substantial pro- portion escape disease control and die. Delinea- tiop of high risk factors may allow for identification of these individuals and permit the early use of other therapeutic programs of greater in- tensity.

When all clinical groups are considered together, 208/265 (79%) evaluable patients are alive and 170/237 (72%) remain in continuous response with a median duration of follow-up of from 41 to 72 weeks. Although no significant difference between treatment options for each of the four clinical groups has emerged as yet, the overall follow-up period is still short. Differences may emerge later as the follow-up period in- creases.

In conclusion, the progress made in the treatment of childhood RMS has been substantial over the past decade. With each advance, reas- sessment of prevailing therapy is essential. The value of an intergroup study for investigating therapeutic alternatives for, and the natural his- tory of, less common cancers has been demon- strated. Through this approach, an important step has been taken toward a better under- standing of childhood RMS, and further refine- ments in treatment programs are now being studied.


REFERENCES

1 . Angervall, L., and Enzinger, F. M.: Extraskeletal neo- plasm resembling Ewing’s sarcoma. Cancer 36:240-251, 1975.

2. Donaldson, S. S., Castro, J. R., Wilbur, J. R., and Jesse, R. H.: Rhabdomyosarcoma of head and neck in children-Combination treatment by surgery, irradiation, and Chemotherapy. Cancer 31 :26-35, 1973.

3. Edland, R. W.: Embryonal rhabdomyosarcoma. Am. J. Roentgenol. Radium Ther. Nucl. Med. 93:671-685, 1965.

4. Edland, R. W.: Embryonal rhabdomyosarcoma. Five- year survival of a patient treated by radiation and chemotherapy. Am. J . Romtgenol. Radium Ther. Nucl. Med. 99:400-403, 1967.

5. Fernandez, C. H., Sutow, W. W., Merino, 0. R., and George, S. L. : Childhood rhabdomyosarcoma. Analysis of coordinated therapy and results. Am. J . Roentgenol. Radium Ther. Nucl. Med. 123:588-597, 1975.

6. Gehan, E.: A generalized Wilcoxin Test for comparing arbitrarily singly-censored samples. Biomclrika 52:203-224, 1965.

7. Ghavimi, F., Exelby, P. R., D’Angio, G. J., Cham, W., Lieberman, P. H., Tan, C., Mike, V., and Murphy, M. L.: Multidisciplinary treatment of embryonal rhabdomyosarcoma in children. Cancer 35:677-686, 1975.

8. Ghavimi, F., Exelby, P. R., D’Angio, G. J., Whitmore, E. F.,Jr., Lieberman, P. H., Lewis, J. L., Jr., Mike, V., and Murphy, J. L. Combination therapy of urogenital embryo-

nal rhabdomyosarcoma in children. Cancer 32: 1178-1 185, 1973.

9. Grosfeld, ,J. L., Clatworthy, H. W., Jr., and Newton, W. A., ,Jr.: Surgery 4:637-647, 1969. 10. Heyn, R. M., Holland, R., Newton, W. A,, Jr., TeKt, M., Breslow, N. and Hartmann, J. R. The role of combined chemotherapy in the treatment of rhabdomyosarcoma in children. Cancer 34:2128-2142, 1974. 1 1 . Holton, C. P., Chapman, K. F., Lakcey, R. W., Hatch, E. I . , Baum, E. S., Favara, B. E.: Extended combination therapy of childhood rhabdomyosarcoma. Cancer 32:1310-1316, 1973. 12. Horn, R. C., Jr., and Enterline, H. T.: Rhabdomyosar- coma-A clinicopathological study and classification of 39 cases. Cancer 11:181-199, 1958.

13. .Jaffe, N., Filler, R. M., Farber, S., T’raggis, D. G.. Vawter, G: F., TefTt, M., and Murray, J. E.: Rahbdomyo- sarcoma in children. Improved outlook with a multidisciplinary approach. Am. 3. Surg. 125:482-487, 1973.

14. ,James, D. H.,J r . , Husto, 0.. Wren, E. L., and,John- son, W. W.: Childhood malignant tumors. Concurrent chemotherapy with dactinomycin and vincristine sulfate. J A M A 197:1043-1045, 1966.

15. Jereb, B., Cham, W., Lattin, P., Exelby, P., Ghavimi, F., D’Angio, G. ,J., and Tefft, M.: Local control of embryonal rhabdomyosarcoma in children by radiation therapy when combined with concomitant chemotherapy. Int. J .


Radial. Oncol. Biol-Phys. 1 :217-225, 1976. 16. Kilman, J. W., Clatworthy, H. W., Jr., Newton,

W. A,, Jr., and Grossfeld, J. L.: Reasonable surgery for rhabdomyosarcoma. A study of 67 cases. Ann. Surg. 178:346-351, 1973.

17. Lawrence, W., ~Jr . , Jegge, G., and Foote, F. W.: Em- bryonal rhabdomyosarcoma-A chinicopathological study. Cancer 17:361-376, 1964.

18. Lawrence, W., ~ J r . , Hays, D. M., and Moon, T. E.,: Lymphatic metastasis with childhood rhabdomyosarcoma. Cancer 39: 5 56-5 59, 1 977.

19. Maurer, H. M., for the I.R.S. Committee: Intergroup Rhabdomyosarcoma Study (1.R.S.)-Progress report. Proc. Am. Soc. Clin. Oncol. 24, 1976 (Abstr.).

20. Maurer, H . M.: Rhabdomyosarcoma in children- Progress and problems. I n Conflicts in Childhood Cancer. L. F. Sinks a n d J . 0. Godden, Eds. New York, Alan R. Liss, 1975; pp. 345-357.

21. Maurer, H. M.: The Intergroup Rhabdomyosarcoma Study-(N.1.H.)-Objectives and clinical staging classification. J . Pediatr. Surg. 10:977-978, 1975.

22. Pinkel, D., and Pinkren, ,J: Rhabdomyosarcoma in children. J A M A 175:293-298, 1961.

23. Pratt, C . B., Hustu, H. D., Fleming, I. D., and Pinkel, D. : Coordinated treatment of childhood rhabdomyosarcoma with surgery radiotherapy, and combination chemotherapy. Cancer Res. 32:606-610. 1972.

24. Sagerman, R. H.: The role of diagnostic radiology and radiation therapy in the management of soft tissue sarcomas in children. Cancer 35:946-949, 1975.

25. Sagerman, R. H. : Tretter, P., and Ellsworth, R. M.: The treatment of orbital rhabdomyosarcoma of children with primary radiation therapy. Am. J . Roentgenoi. Radium Ther. Nucl. Med. 114:31-34, 1972.

26. Soule, E. H., and Newton, W. A., Jr.: Intergroup Rhabdomyosarcoma Study (1.R.S.)-Identification of a histologic subgroup,? Ewing’s tumor of soft tissue. Proc. Am. Soc. Clin. Oncol. 301, 1976 (Abstr.). nant tumors-in U.S. children. J . Pediat. 86:254-258, 19751

27. Steinberg,,J., Haddy, T. B., Porter, F. S., and Thur- man, W. G. : Clinical trials with cyclophosphamide in children with soft tissue sarcoma. Cancer Chnother. Rep. 28:39-41, 1963.

28. Stobbe, G. C., and Dargeon, H. W.: Embryonal rhabdomyosarcoma of the head and neck in children and adoles- cents. Cancer 3:826-836, 1950.

29. Sutow, W. W. : Chemotherapy in childhood cancer (except leukemias). Cancer 18:1585-1589, 1965.

30. Sutow, W. W., Berry, D. H., Haddy, T. B., Sullivan, M. P., Watkins, W. L., and Windmiller, J . : Vincristine sulfate therapy in children with metastatic soft tissue sarcoma. Pediatrics 38:465-472, 1966.

31. Sutow, W. W., Sullivan, M . P., Ried, H. L., Taylor, H. G., and Griffith, K. M.: Prognosis in childhood rhabdomyosarcoma. Cancer 25:1385-1390, 1970.

32. Tan, C. T. C., Golbey, R. G., Yap, C. L., Wollner, N., Hackethal, C. A,, Murphy, L. M., Dargean, H. W., and Durchenal, J. H. : Clinical experiences with actinomycin D, KS 2 and K 1 (KS 4). Ann. N. T. Acud. Sci. 89:426-444, 1960.

33. Tefft, M. , Fernandez, C. H. , and Moon, T. E.: Rhab- domyosarcoma-Response with chemotherapy prior to radiation in patients with gross residual disease. Cancer, 39:665-670, 1977.

34. Tefft, M., and Jaffe, N.: Sarcoma of the bladder and prostate in children. Rationate for the role of radiation therapy based on a review of the literature and a report of fourteen additional patients. Cancer 32: 1161-1 177, 1973.

35. Wilbur, ~ J . R.: Combination chemotherapy of embryonal rhabdomyosarcoma. Cancer Chemother. Rep. 58:281-284, 1974.

36. Wolner, N., Tan, C., Ghavimi, F., Rosen, G., Tefft, M., and Murphy, M. L.: Adriamycin in childhood leukemia and solid tumors. Proc. Am. Assoc. Cancer Res. 12:75, 1971 (Abstr.).

37. Youne. I . L., and Miller, R. W.: Incidence of malie-

The intergroup rhabdomyosarcoma study.A preliminary report

Documents

Transcript of The intergroup rhabdomyosarcoma study.A preliminary report