High histologic and overall response to dose intensification of ifosfamide, carboplatin, and etoposide with cyclophosphamide, doxorubicin, and vincristine in patients with high-risk ewing sarcoma family tumors

The Bambino Gesù Children's Hospital experience

Abstract

BACKGROUND

Ewing sarcoma (ES) and extraosseous ES/primitive neuroectodermal tumors (PNET) share histopathologic features of the ES family of tumors (ESFT). The authors report on their results from a regimen of ifosfamide, carboplatin, and etoposide (ICE) with cyclophosphamide, doxorubicin, and vincristine (CAV) dose intensification in patients with high-risk ESFT.

METHODS

Since 1990, patients with ESFT and with 1 or more of the following risk factors were reviewed: tumor volume > 200 mL, tumor site with a poor prognosis, and pulmonary and/or bone marrow metastases.

RESULTS

Thirty-six patients with ESFT who were involved in the study were divided into 2 arms of 18 patients each. One group received treatment with various regimens, and the other group received treatment with ICE plus CAV. The disease was brought under control more rapidly in the latter patients, for whom surgery was more easily feasible, and up to 90% of patients achieved a major response, with an estimated 3-year overall survival rate of 67% ± 12%.

CONCLUSIONS

The current results showed that ICE plus CAV was tolerated well and was effective in the studied subset of tumors, indicating that dose intensification correlates with better disease control, a high percentage of necrosis, and conservative surgery in patients with high-risk ESFT. Cancer 2006. © 2006 American Cancer Society.

Ewing sarcoma (ES) was described first in 1921 by James Ewing as a bone tumor that, unlike osteosarcoma, responded to radiotherapy.1 Since then, important progress has been made in characterizing this disease, which is generated by neuroepithelial postgangliar and parasympathetic, poorly differentiated, small, round cells. Several neural markers, such as neuron-specific enolase, Leu-7, and neurofilament, are expressed. ES is an undifferentiated bone tumor that also involves surrounding soft tissues. Extraosseous ES, or peripheral primitive neuroepithelial tumor (pPNET), has the same histopathologic features. The 2 combine to form the ES family of tumors (ESFT).2, 3

In Italy, approximately 100 pediatric patients are diagnosed with ES/PNET each year, making it the second most common bone tumor. It mainly affects adolescents, and the risk factors include genetic aberrations, age at diagnosis, tumor volume, site, histologic response, and bilateral and multiple pulmonary or bone marrow metastases.4–10 A multidisciplinary approach (chemotherapy, radiotherapy, and surgery) is taken to the treatment of patients with these tumors.

The most effective drugs for the treatment of ESFT have proven to be ifosfamide, etoposide, cyclophosphamide, vincristine, doxorubicin, carboplatin, and actinomycin D.11–16 These agents in various combinations have led to average 5-year survival rates of 65% in patients with localized tumors and 13% in patients with bone marrow metastases.17–20 Recently, an intensified treatment at induction produced even higher response rates in treatment-naive patients with high-risk ESFT.21, 22 In Phase I and II studies, combined ifosfamide and etoposide with carboplatin (the ICE regimen) in various schedules has proved feasible and effective in the treatment of several pediatric tumors,23–25 and early experience with the ICE regimen plus a hematopoietic growth factor (granulocyte-colony stimulating factor [G-CSF]26) in recurrent pediatric solid tumors has also been promising.

Based on this evidence, we began a pilot study in 1998 using ICE plus cyclophosphamide, doxorubicin, and vincristine (CAV) as dose-intensified, front-line therapy for patients with high-risk ESFT; and, since 1990, all patients who are referred to our institution have been studied to determine the significance of the risk factors described above and the effects of the different chemotherapeutic strategies adopted.

MATERIALS AND METHODS

Patients

From 1990 to April 2005, 38 eligible patients were recruited at our institute. All patients had histologically confirmed ES or extraosseous ES/PNET (EES/PNET). They were considered high-risk patients if they had ≥1 of the following risk factors: tumor volume >200 mL, tumor site with a poor prognosis (pelvis, chest wall, vertebra), or pulmonary or bone marrow involvement at onset. None of the patients had received any prior therapy. Measurable disease also was mandatory for recruitment.

Other eligibility criteria included adequate bone marrow (platelet count >100,000/>L and neutrophil count >1000/μL); renal (serum creatinine level < 1.5 times normal) and hepatic function (serum aspartate aminotransferase and alanine aminotransferase levels < 2.5 times normal, and a serum bilirubin level < 1.5 times normal). Written informed consent was obtained from the patient's parent or legal guardian, and the study protocol was approved by the Institutional Review Board.

Patients

The characteristics of the patient population are given in Table 1.

Table 1. Characteristics of the Study Population

Patient	Age (M)	Histology	Metastasis	Induction Therapy	Conservative Surgery after Induction	Necrosis > 95%*
PNET: primitive neuroepithelial tumor; RMS 88: ifosfamide, doxorubicin, actinomycin D, and vincristine followed by ifosfamide, actinomycin D, and vincristine with or without radiotherapy; ES: Ewing sarcoma; CECAT: cyclophosphamide, etoposide, carboplatin, and thiotepa; RT: radiotherapy; HT: hyperthermia; ICE: ifosfamide, carboplatin, and etoposide; CAV: cyclophosphamide, doxorubicin, and vincristine. * Necrosis: no, < 95%; yes, > 95%. † Disappearance of tumor after induction.
1	113	PNET	No	RMS 88	No	No
2	70	ES	No	CECAT + RT	Yes	Yes
3	135	ES	Yes	CECAT	No	No
4	35	PNET	No	CECAT	No	No
5	149	ES	Yes	CECAT + RT	Yes	Yes
6	145	ES	Yes	CECAT	No	No
7	149	PNET	No	RMS 88	No	No
8	79	PNET	Yes	CECAT	No	No
9	66	ES	Yes	CECAT	No	No
10	148	PNET	Yes	CECAT	No	No
11	102	ES	No	CECAT + HT	No	No
12	101	ES	No	CECAT	No	No
13	97	ES	Yes	CECAT + HT	No	No
14	171	ES	Yes	CECAT + RT	No	No
15	136	PNET	Yes	CECAT	No	No
16	214	ES	No	CECAT	No	No
17	170	ES	Yes	ICE	No	No
18	78	ES	No	ICE	No	No
19	117	ES	No	ICE/CAV	Yes	No
20	189	ES	No	ICE/CAV	Yes	Yes
21	156	ES	No	ICE/CAV	Yes	Yes
22	42	ES	No	ICE/CAV	Yes	Yes
23	83	ES	Yes	ICE/CAV	Yes	Yes
24	54	ES	Yes	ICE/CAV	No	No
25	191	PNET	Yes	ICE/CAV	Yes	No
26	77	ES	No	ICE/CAV	Yes	Yes
27	143	ES	Yes	ICE/CAV	No	No
28	174	PNET	Yes	ICE/CAV	Yes	No
29	20	PNET	No	ICE/CAV	Yes	Yes
30	96	PNET	No	ICE/CAV	No	Yes†
31	149	PNET	No	ICE/CAV	No	Yes†
32	56	PNET	No	ICE/CAV	Yes	Yes
33	156	PNET	No	ICE/CAV	Yes	Yes
34	60	ES	No	ICE/CAV	No	No
35	83	PNET	No	ICE/CAV	Yes	Yes
36	132	ES	Yes	ICE/CAV	No	No

Treatment

After an initial biopsy was obtained, all patients were treated with 4 courses of chemotherapy followed by evaluation for surgery. The various chemotherapeutic regimens adopted were as recommended by the Italian Soft Tissue Sarcoma Protocol RMS 88 and consisted of combined ifosfamide (9 g/m² daily for 3 days), doxorubicin (60 mg/m² daily for 2 days), actinomycin D (1.5 mg/m² for 1 day), and vincristine (1.5 mg/m² for 1 day) followed by 2 courses of combined ifosfamide (9 g/m² daily for 3 days), actinomycin D (1.5 mg/m² for 1 day), and vincristine (1.5 mg/m² for 1 day) with or without radiotherapy; or 4 courses of CECAT including cyclophosphamide (300 mg/m² daily for 2 days), etoposide (100 mg/m² daily for 3 days), carboplatin (500 mg/m² daily for 2 days), and thiotepa (10 mg/m² daily for 3 days) or the ICE regimen (ifosfamide, 2 g/m² daily for 3 days; carboplatin, 600 mg/m² daily for 1 day; and etoposide, 100 mg/m² daily for 3 days). The different treatments for each patient are shown in Table 2.

Table 2. Event-Free Survival and Overall Survival in the Study Population

Patient	Induction Therapy	EFS in Mos	OS in Mos	Status
EFS: event-free survival; OS: overall survival; RMS 88: ifosfamide, doxorubicin, actinomycin D, and vincristine followed by ifosfamide, actinomycin D, and vincristine with or without radiotherapy; DOD: died of disease; CECAT: cyclophosphamide, etoposide, carboplatin, and thiotepa; RT: radiotherapy; HT: hyperthermia; ICE: ifosfamide, carboplatin, and etoposide; CAV: cyclophosphamide, doxorubicin, and vincristine; Rel: recurrence; CR: complete response.
1	RMS 88	22	25	DOD
2	CECAT + RT	120	144	DOD
3	CECAT	60	68	DOD
4	CECAT	4	5	DOD
5	CECAT + RT	38	42	DOD
6	CECAT	28	35	DOD
7	RMS 88	4	6	DOD
8	CECAT	11	14	DOD
9	CECAT	24	28	DOD
10	CECAT	125	125	CR
11	CECAT + HT	35	39	DOD
12	CECAT	3	4	DOD
13	CECAT + HT	20	24	DOD
14	CECAT + RT	9	11	DOD
15	CECAT	12	14	DOD
16	CECAT	11	14	DOD
17	ICE	20	23	DOD
18	ICE	13	16	DOD
19	ICE/CAV	78	78	CR
20	ICE/CAV	95	95	CR
21	ICE/CAV	41	41	CR
22	ICE/CAV	81	81	CR
23	ICE/CAV	80	80	CR
24	ICE/CAV	13	22	DOD
25	ICE/CAV	8	14	DOD
26	ICE/CAV	11	26	Rel
27	ICE/CAV	20	20	CR
28	ICE/CAV	16	24	DOD
29	ICE/CAV	43	43	CR
30	ICE/CAV	41	41	CR
31	ICE/CAV	14	31	DOD
32	ICE/CAV	50	50	CR
33	ICE/CAV	17	27	CR
34	ICE/CAV	24	24	CR
35	ICE/CAV	19	19	CR
36	ICE/CAV	9	10	DOD

The ICE-CAV regimen is illustrated in Figure 1. It consisted of 2 courses of ICE (ifosfamide, 1.8 g/m² daily for 5 days; carboplatin, 400 mg/m² daily for 2 days; and etoposide, 100 mg/m² daily for 5 days) followed by 2 courses of CAV (cyclophosphamide, 1500 mg/m² daily for 2 days; doxorubicin, 75 mg/m² per day as a 72-hour continuous infusion; and vincristine, 1.5 mg/m² as a 72-hour continuous infusion). Courses were repeated every 21 to 28 days after patients achieved full hematologic recovery. G-CSF was administered at a dose of 5 μg/kg daily until patients achieved neutrophil recovery.

Figure 1. Illustration of the ifosfamide, carboplatin, and etoposide-cyclophosphamide, doxorubicin, and vincristine (ICE-CAV) treatment plan. RT: radiotherapy; PBSCT: peripheral blood stem cells; CE: cyclophosphamide and etoposide; ETC: etoposide, thiotepa, and cyclophosphamide.

Response Criteria

Disease was assessed at the baseline and after every 2 courses of treatment by diagnostic imaging in 2 perpendicular dimensions using computed tomography scans or magnetic resonance imaging; ultrasound tomography and X-rays were used to study distant sites. Bone marrow infiltration was evaluated by biopsies and aspirates, including immunocytology with a monoclonal antibody panel at onset. Bone marrow analysis was performed every 2 courses in patients with known involvement. Tc-99m bone scans were obtained at baseline and at the end of the treatment. Response Evaluation Criteria in Solid Tumor (RECIST) categories were used to assess response. A complete response (CR) was defined as the disappearance of all known disease for at least 4 weeks, a partial response (PR) was defined as a reduction ≥ 50% in the sum of the products of the greatest dimensions of measurable lesions for at least 4 weeks, a minor response (MR) was defined as a reduction < 50% but > 25% in the sum of the products of the greatest dimensions of measurable lesions for at least 4 weeks, stable disease (SD) was defined as a reduction < 25% or no increase in the sum of the products of the greatest dimensions of measurable lesions, and progressive disease (PD) was defined as any increase in known disease or the appearance of new lesions.

Statistical Analysis

Both overall survival (OS) and event-free survival (EFS) were considered. OS was calculated from the date of histologic diagnosis to the date of last follow-up or death from any cause. EFS was calculated from the date of histologic diagnosis to the date of disease progression (i.e., an increase in the volume of any preexisting lesion and/or the appearance of any new lesion), recurrence, or death from any cause. Patients who experienced no events of interest were censored at the time of last follow-up.

Clinical variables were classified as follows: age at diagnosis (< 120 mos or > 120 mos), surgery group (Group I after induction and Group II later), histology (ES vs. PNET), treatment (others vs. ICE-CAV), necrosis (< 95% or > 95%), and metastasis (yes or no). Correlations between these variables and the treatments were analyzed using the chi-square test and the Fisher exact test.27 The impact of the same variables on OS and EFS was assessed using Kaplan–Meier survival curves28 and log-rank significance tests with 95% confidence intervals. SPSS 11.0 software for Windows (SPSS Inc., Chicago, IL) was used for the analyses. The treatment was defined as effective if it improved ≥ 1 of the following parameters: major responses, percentage of necrosis, time of surgery, and survival.

RESULTS

Enrolment began in 1990, and 38 patients had been accrued by April 2005. Two patients were ruled out because they had undergone complete tumor resection at onset. Therefore, the current study involved 22 patients with ES and 14 patients with EES/PNET. The median patient age was 114 months (range, 20-214 mos). The primary disease site was the pelvis (5 bone and 4 soft tissue tumors), femur (2 tumors), scapula (2 tumors), hip bone (2 tumors), clavicle (1 tumor), vertebra (4 tumors), humerus (3 tumors), distal tibia (3 tumors), abdomen (2 tumors), rib (1 tumor), fibula (2 tumors), chest wall (4 tumors), pretibial soft tissue (1 tumor), foot soft tissue (1 tumor), and radius (1 tumor); 9 of 36 evaluable patients had pulmonary metastases, and 7 of 36 patients had bone marrow disease at baseline. The characteristics of the study population are shown in Table 1, and patient survival is shown in Table 2. Statistical analyses according to age, histology, treatment, timing of surgery, necrosis, and metastases were conducted on the entire population. The log-rank test showed that both EFS and OS were dependent on all variables except for age older than 120 months and histotype (Table 3). The ICE-CAV regimen raised the percentage of necrosis to > 90% (P = .04) and enabled more conservative surgery (P = .01) (Table 4).

Table 3. Association between Clinical Variables (Including Treatments) and 3-Year Survival Probability

Clinical Variable	No. of Patients	EFS			OS
		No. of Events	Survival, %	Log-rank*	No. of Events	Survival, %	Log-rank*
EFS: event-free survival; OS: overall survival; NS: not significant; ES: Ewing sarcoma; PNET: primitive neuroectodermal tumor; CT: chemotherapy; ICE: ifosfamide, carboplatin, and etoposide; CAV: cyclophosphamide, doxorubicin, and vincristine. * Log-rank P value (95% confidence interval).
Age
< 120 mos	18	10	80.2	NS	10	83.3	NS
> 120 mos	18	12	76.0		12	78.9
Histology
ES	22	14	51	NS	13	62	NS
PNET	14	8	48		8	51
Surgery after induction CT
Yes	13	4	61	.01	3	64	.061
No	23	18	21		18	23
Necrosis
≤ 95%	23	19	18	.0007	19	22	.0005
> 95%	13	3	85		2	87
Metastasis
Yes	16	13	61	.015	13	62	.043
No	20	9	12		9	24
Treatment
Other	18	17	22	.0035	17	27	.023
ICE-CAV	18	6	67		5	74

Table 4. Correlations between Conventional Clinical Variables and Treatments

Clinical Variable	No. of Patients			P
	Total	ICE-CAV
		Yes	No
ICE: ifosfamide, carboplatin, and etoposide; CAV: cyclophosphamide, doxorubicin, and vincristine; NS: not significant; CT: chemotherapy.
Age
< 120 mos	18	9	9	NS
> 120 mos	18	9	9
Histology
Ewing sarcoma	22	10	12	NS
Primitive neuroepithelial tumor	14	8	6
Surgery after induction CT
Yes	13	11	2	.01
No	23	7	16
Necrosis
≤ 95%	23	7	16	.04
> 95%	13	11	2
Metastasis
Yes	16	6	10	NS
No	20	12	8

Among the patients who received treatments other than ICE-CAV, 7 patients achieved a PR, 7 patients had a MR, and 4 patients had PD; conservative surgery was possible after induction in 2 of 18 patients with 95% tumor necrosis. Two of 18 patients underwent amputation for inoperable disease after therapy, 1 patient underwent a complete resection after receiving a myeloablative regimen, 1 patient refused amputation of the leg, and another patient underwent emergency surgery because of cardiac compression by the mass and died 1 month later. Eleven of 18 patients in this group could not undergo complete tumor resection at any time during treatment. At the time of last follow-up, 1 patient remained alive in CR 125 months after diagnosis, all remaining patients had died of their disease. The median EFS was 31 months (range, 3-125 mos) and the median OS was 35 months (range, 4-144 mos).

Among the patients who received treatment with ICE-CAV, 2 patients achieved a CR, 15 patients achieved a PR, and 1 patient had PD (after the first course of CAV). The overall response (CR + PR) to ICE-CAV was 94.4% (11% CR and 83% PR). After ICE-CAV, 7 patients with bone disease and 5 patients without bone disease underwent surgery. Surgery was not attempted after ICE-CAV in 6 patients, because the lesions had disappeared in 2 patients with pelvic EES/PNET, or it had to be delayed in 3 patients with ES, while 1 patient with ES (of the rib) had PD. In 9 of 12 patients who underwent surgery, histology demonstrated rates of necrosis ranging from 95% to 100%. Two patients with pelvic disease achieved a CR after induction. None of the patients suffered severe consequences of the treatment, and no amputation was required.

Among 18 patients, 12 patients achieved a CR, 1 patient was in first recurrence (local lymph node involvement and a pulmonary metastasis) at 10 months and was receiving second-line therapy, and 5 died of disease (1 patient had ES of the humerus, and 4 patients had EES) 22 months, 14 months, 24 months, 31 months, and 9 months, respectively, after their diagnoses. Recurrences were local in 2 patients and distant in 3 patients. The median OS was 40.3 months (range, 10-95 mos), and the EFS was 36.6 months (range, 9-95 mos).

The Kaplan–Meier probability curves for the 2 treatments are shown in Figures 2 and 3. No multivariate analysis of the variables independently associated with survival could be attempted because of the small number of patients involved.

Figure 2. Overall survival (OS) according to the treatment received. Solid line: ICE/CAV; broken line: other treatments.

Figure 3. Event-free survival (EFS) according to the treatment received. Solid line: ICE/CAV; broken line: other treatments.

Toxicity of ICE-CAV

In all, 7 courses of ICE-CAV were administered. All patients were evaluated for toxicity after ICE-CAV, and toxicity was graded according to World Health Organization (WHO) criteria. Grade 3 and 4 hematologic toxicity occurred in all ICE courses administered, with a median of 2 red cell transfusions required per patient (range, 1-10 transfusions; total, 44 transfusions) and 4 platelet transfusions per patient per course (range, 1-14 transfusions; total, 87 transfusions). Eight episodes of documented sepsis occurred.

The CAV regimen produced the same Grade 3 and 4 WHO hematologic toxicities that were produced by ICE but prompted the need for fewer transfusions (red blood cells, 42 transfusions; platelets, 61 transfusions); 2 episodes of sepsis were reported. No other toxicity or organ failure was reported.

DISCUSSION

To our knowledge, this is the first report on the successful use of ICE as a front-line therapy in children with high-risk ESFT, because no other studies have investigated the activity and feasibility of this combination in these patients. Our combination of ICE plus CAV proved to be just as active and tolerable as other regimens that have been tested in recent clinical trials.

Despite advances in multimodal treatment, the chance for survival among patients with high-risk ESFT is < 50% at 5 years. The prognosis and risk of recurrence are correlated with the histologic response to induction therapy, as documented elsewhere.4, 5, 8, 9, 29–34 For the current report, the combination of ICE with CAV was investigated with the objective of improving the histologic response, revealing impressive results and a high percentage of necrosis thus enabling the last 18 patients who were referred to our institution to undergo conservative surgery.

The ICE combination was tested first in pediatric tumors by Fields et al.35 as a 2-day schedule in an extensive study that involved both adults and children, and produced an interesting response rate, 25% of which concerned heavily pretreated patients with soft tissue sarcomas. The Pediatric Oncology Group reported an overall response rate of 43% (23% in ES) in 92 pediatric patients after administering ICE in a 3-day schedule.36 Marina et al.37 tested carboplatin in association with ifosfamide at a dose of 2 g/m² for 2 days and etoposide at a dose of 100 mg/m² for 2 days and reported an area under the curve for carboplatin of 8 mg/mL per minute: Although their regimen produced a low CR rate, they reported approximately 33% major responses among patients with neuroblastoma, rhabdomyosarcoma, and ESFT. Our decision to use the ICE regimen in a 5-day schedule at maximum dosage derived from the report from Cairo et al.,25 who treated 123 patients with refractory or recurrent pediatric solid tumors and obtained a 63% overall response rate and 50% major response rate among patients with soft tissue sarcomas and ES. ICE supported with G-CSF was used as reinduction therapy in 96 selected patients with sarcoma (including 24 ESFTs), and a 48% overall response rate was obtained.26

In conclusion, the results of the current study demonstrate that ICE-CAV is a good choice for treating this subset of tumors; a 95% CR rate after ICE-CAV plus surgery and an estimated 3-year OS rate of 67% ± 12% show that dose intensification enables good control of the disease, a high percentage of necrosis, and a better chance of conservative surgery. These findings are important; however, a larger patient population and longer follow-up will be needed to evaluate the real effectiveness of the entire protocol in this family of tumors.

Acknowledgements

The authors thank M. Mercuri, M.D., and M. Manfrini, M.D. (Istituti Orthopedici Rizzoli, Bologna, Italy, for performing surgery on patients with skeletal involvement.