|Home | About | Journals | Submit | Contact Us | Français|
This prospective study was designed to determine the safety and efficacy of cyclophosphamide, BCNU and etoposide (CBV) conditioning and autologous peripheral blood stem cell transplant (PBSCT) in children with relapsed or refractory Hodgkin and non-Hodgkin lymphoma (HL and NHL).
Patients achieving CR or PR after 2–4 courses of reinduction underwent a G-CSF mobilized PBSC apheresis with a target collection dose of 5×106 CD34+/kg. Those eligible to proceed received autologous PBSCT after CBV (7200 mg/m2, 450–300 mg/m2, 2400 mg/m2).
Forty-three of 69 patients (30/39 HL, 13/30 NHL) achieved a CR/PR after reinduction. Thirty-eight patients (28 HL, 10 NHL) underwent PBSCT. All initial 6 patients who received BCNU at 450 mg/m2 experienced grade III or IV pulmonary toxicity compared to none of the subsequent 32 receiving 300 mg/m2 (p<0.0001). The probability of OS at 3 years for all patients is 51% and for transplanted patients is 64%. The 3-year EFS is 38% (45% for HL; 30% NHL). The 3-year EFS in transplanted patients is 66% (65% HL; 70% NHL). Initial duration of remission of ≥ 12 vs < 12 months was associated with a significant increase in OS (3 ys OS 70% vs 34%) (p=0.003).
BCNU at 300 mg/m2 in a CBV regimen prior to PBSCT is well tolerated in relapsed or refractory pediatric lymphoma patients. A short duration (< 12 months) of initial remission is associated with a poorer prognosis. Lastly, a high percentage of patients achieving a CR/PR after reinduction therapy can be salvaged with CBV and autologlous PBSCT.
The prognosis for children with newly diagnosed lymphomas has significantly improved over the last 25 years. The survival rate for patients with localized and disseminated non-Hodgkin lymphoma (NHL) is over 95% and over 80%, respectively for most subtypes.(1–11) However, the prognosis for refractory or recurrent NHL in children and adolescents remains poor. Patients enrolled in the Children’s Cancer Group (CCG) 551 that subsequently relapsed had a 12% 5-year overall survival (OS).(12) The 5-year OS for relapsed NHL patients using dexamethasone, etoposide, cisplatin, high-dose cytarabine, and L-asparaginase (DECAL) was 30%.(13) Children diagnosed with Stage I or II HL experience a long-term event-free survival (EFS) greater than 90%.(14, 15) Patients with advanced stage or “B” symptoms at presentation have long-term EFS rates of over 80%.(16) As in NHL, children with relapsed or refractory Hodgkin lymphoma (HL) have a poor prognosis.(17, 18) The 5-year OS is 31% for children with HL reinduced with DECAL.(13) The 8-year OS and EFS are 34% and 23%, respectively with cytosine arabinoside, cisplatin and etoposide (APE).(17)
In adults with lymphoma, autologous stem cell transplant results in 4 to 10 year OS rates of 42–70%.(19–22) Factors associated with a poor prognosis in adults with lymphoma following autologous stem cell transplantation include chemoresistance, large tumor burden, short remission duration, poor performance status, and extranodal relapse.(23, 24) Studies of children with lymphoma treated with high-dose chemotherapy and autologous stem cell rescue are limited by small numbers, a wide variety of pre-transplant chemotherapy and conditioning regimens. Overall, they demonstrate OS rates similar to adults.(18, 25)
Bone marrow transplant (BMT) conditioning regimens including cyclophosphamide, carmustine (BCNU), and etoposide (CBV), either as separate agents or together are effective in adults with recurrent NHL and HL.(21, 26–30) This report describes the results of a prospective study assessing the toxicity and efficacy of CBV and autologous peripheral blood stem cell transplantation in pediatric patients with relapsed or refractory lymphoma who achieve a complete remission/partial remission (CR/PR) after reinduction.
This prospective study enrolled children, initially diagnosed between the ages of 12 months and 21 years, at time of their first relapse or induction failure (defined as failure to achieve a CR with a reinduction chemotherapy for HL, 2 cycles of a reinduction chemotherapy for NHL, or 4 cycles for large cell lymphoma patients). The study excluded patients with low stage HL treated with radiation only or with chemotherapy only, and human immunodeficiency virus (HIV) positive patients. Local Institutional Review Boards approved the study at each institution, and the patient or legal guardians signed an informed consent.
Study pathologists centrally reviewed patient materials from initial diagnosis and relapse to confirm diagnosis utilizing the Revised European American Lymphoma (REAL) Classification. Six cases of large cell lymphoma had additional immunoperoxidase staining using an automated immunostainer (Ventana, Tucson, AZ) and heat-induced epitope retrieval with a microwave. Lineage specific stains included: anti-CD-20 (DAKO, Carpenterial, CA) for B-cell lineage, CD3 or CD45RO (DAKO) for T-cell lineage, and CD30 (DAKO) and ALK-1 (DAKO) for anaplastic large cell lymphoma (ALCL).
No evidence of disease by physical exam and imaging studies (CT scan), including negative bone marrow and cerebrospinal fluid (CSF), constituted a CR. A reduction in the total volume of all measured lesions by at least 50% with no single lesion increasing by more than 25% and no new lesions constituted a PR. Stable disease (SD) was defined as a reduction in the total volume of all measured lesions of less than 50% with no single measured tumor lesion increasing in volume by >25% and no new lesions. Failure to achieve a CR with growth in any measured lesion by >25% in volume, or development of new lesions or new sites of tumor constituted progressive disease (PD). Recurrence was defined as redevelopment of tumor at any site after achievement of a CR.
The protocol did not prescribe the reinduction chemotherapy regimen. Among the HL (N=39) patients, 13 received ifosfamide, carboplatin, etoposide (ICE) and 11 received vinorelbine and ifosfamide (VI).. Among the NHL patients (N=30), 17 received ICE.
Patients completing two courses of reinduction chemotherapy with at least stable disease and no evidence of marrow involvement proceeded to peripheral blood stem cell (PBSC) collection. Patients were mobilized following chemotherapy and granulocyte-colony stimulating factor (G-CSF) therapy and collected after hematological recovery. Patients not meeting response criteria after two courses received two additional courses of reinduction therapy. Patients with persistent marrow involvement after four cycles of reinduction therapy came off study.
Mobilization consisted of G-CSF 10 micrograms/kg daily for at least 3 days prior to and on the days of apheresis. The minimum goal of collection was 2 × 106 CD34+ cells/kg and the target was 5 × 106 CD34+ cells/kg. PBSC harvesting used standard procedures on either a COBE Spectra or Fenwall CS-3000 Plus apheresis machine.
Patients in PR or CR after no more than 4 cycles of reinduction chemotherapy and with adequate PBSC stored, proceeded to CBV conditioning and autologous PBSC transplantation (PBSCT). Patients with PD or SD were ineligible to proceed. Patients going to transplant met the following criteria of organ function: liver transaminsase <2.5× normal, total bilirubin <1.5 mg/dl, a glomerular filtration rate (GFR) of >60 ml/min/1.73 m2 and a serum creatinine ≤1.5 mg/dl, a shortening fraction ≥28% on heart echocardiogram or an ejection fraction >50% of normal, and on pulmonary function testing a vital capacity (VC ≥65% and a total lung capacity (TLC) ≥50% of normal.
All patients received cyclophosphamide 1800 mg/m2/D over 1 hour on days -5, -4, -3, -2 and etoposide 800 mg/m2/D over 24 hours on days -8, -7, and -6. The first 6 patients received BCNU at 150 mg/m2/day for 3 days and the remaining 32 patients received 100 mg/m2/day for 3 days on days -8, -7, and -6, following the amendment in February 2000 reducing the BCNU dose. Protocol treatment included Mesna to prevent hemorrhagic cystitis. Patients received methylprednisolone as a protective agent to reduce the risk of pulmonary toxicity as follows: 1 mg/kg days -9 to -2, then tapered off by day +6. All other supportive care was per local institutional guidelines.
Patients received 5 micrograms/kg G-CSF over 2 hours after the infusion of the stem cells daily until the absolute neutrophil count (ANC) was >2000/mm3 for 3 days.
The first of three days of an ANC of >500/mm3 and a platelet count of >50,000/mm3 without platelet transfusion defined neutrophil and platelet engraftment.
Imaging at sites of disease occurred on days +28, +100, 6 months, and 1, 2, 3, and 5 years after transplant. Patients underwent pulmonary function tests at 3, 6, and 12 months post-transplant. If pulmonary function tests were abnormal, patients repeated them at 6 month intervals until normalization or stabilization.
SAS software, Version 9.1 (SAS Institute, Cary,NC) was used for statistical analysis. OS is the time on study or transplant to death or last follow-up. Relapse, progression, or death defined the events in the event-free survival determination. EFS is the time from on study or transplant to an event or last follow-up for those who did not have an event. Toxicity was graded according to the NCI CTCAE v. 2.0 common toxicity scale. Descriptive statistics summarized demographic and clinical variables Chi-square tests were used to compare response rates for HL and NHL. Fisher’s exact test compared toxicity rates by BCNU dose. The days to myeloid and platelet recovery, EFS, OS, and time to relapse were determined by the method of Kaplan-Meier. Patients who did not obtain recovery were censored at first event or last follow-up following transplant. The log-rank test compared time to event distributions between groups. Statistically significant p-values were less than 0.05. The above outcomes measures were tested in the following subgroups: NHL vs, HL, nodular sclerosing HL vs. other HL patients, Burkitts/Burkitt like vs. lymphoblastic NHL, CR vs. PR just prior to transplant, length of CR1 >1 year vs. <1yr, HL age >15yrs vs. <15 yrs, relapse on or off therapy, failed induction vs. relapse, and the dose of BCNU. The effect of length of CR1 was also separately tested in HL and NHL patients.
Sixty-nine patients enrolled, 39 with HL and 30 with NHL. Thirty-eight patients received CBV conditioning and autologous PBSCT; 28 with HL and 10 with NHL. Study investigators amended the protocol on February 11, 2000, to reduce the dose of the BCNU from 450 mg/m2 to 300 mg/m2.
The study entry patient characteristics for the HL patients and the NHL patients are in Table 1. Four HL patients entered for failed induction and 35 at time of first relapse. Four NHL patients entered for failed induction and 26 at time of first relapse.
Forty-one patients (15 NHL, 26 HL) had pathology material available for central classification of histological subtypes. The histologic subtypes of the 39 cases of HL and the 30 NHL are shown in Table 1.
Of the 39 HL patients, 30 achieved a CR/PR. Four failed due to PD, four achieved only a SD status, and one was inevaluable. Among the 30 NHL patients, 13 achieved a CR/PR among the 26 patients evaluable (4/8 with lymphoblastic lymphoma [LBL], 2/5 with diffuse large B-cell lymphoma [DLCL], 4/5 with ALCL, and 3/8 with Burkitt’s lymphoma [BL]). Ten NHL patients failed due to PD, three achieved only a SD status, and four were inevaluable. The NHL patients were less likely to achieve CR/PR than the HL patients (45% vs. 77%, p=0.007). Overall, 43 of the 69 patients (62%) achieved a CR/PR, and 38 (28 HL, 10 NHL) proceeded to transplant. Four patients refused to proceed to CBV conditioning and autologous stem cell transplantation (Figure 1).
Fifty-one patients underwent PBSC apheresis; forty successfully obtained adequate stem cells for transplant (2 × 106 CD34+/kg). The median CD34+ dose/kg stored was 6.2 × 106/kg (range 0.2 × 106 to 94.7 × 106/kg). The median cell dose infused was 6 × 106CD34+/kg. The myeloid engraftment rates for an ANC of 500/mm3 or 1000/mm3 were not significantly different based on CD34+/kg cell dose. However, the median time to recover the platelet count of 50,000/mm3 was significantly improved with a higher CD34+/kg cell dose ([>4×106/kg vs ≤4×106CD34+/kg] [20 (4–48d) vs 41 (39–480d)]). There was no significant difference in the engraftment kinetics between the HL vs. the NHL patients.
All 6 patients receiving 450 mg/m2 of BCNU developed grade III or IV pulmonary toxicity with onset between 30 to 60 days post-transplant. Four of the six required supplemental oxygen (duration 5 days – 9 months) and five received steroids. All but two came off supplemental oxygen therapy after a few days. One patient developed persistent abnormal pulmonary function tests (PFTs), but came off oxygen after 9 months. The other patient relapsed 6 months after transplant while still on steroids and oxygen. None of the subsequent 32 patients treated with 300 mg/m2 of BCNU developed grade III or IV pulmonary toxicity (p<0.0001).
The median follow-up from date of study entry of the survivors is 3.2 years (maximum 6.3 years). The 3-year OS in all 69 patients from study entry is 51% (95% CI, 38%–62%) (Figure 2A). The 3-year OS from date of study entry for HL patients is 63% (95% confidence interval (CI), 46%–76%) and for NHL patients is 34% (95%CI, 17%–52%) (Figure 2B) (HL vs. NHL: P=0.0087). The 3-year EFS of all 69 patients from study entry is 38% (95% CI, 27%–50%) (Figure 3A). The 3-year EFS from study entry for HL patients is 45% (95% CI, 29%–60%) and for NHL patients is 30% (95% CI, 15%–46%) (HL vs. NHL: P=0.015) (Figure 3B).
Patients achieving a CR/PR vs. SD/PD had significantly better OS and EFS at 3 years: OS 73% vs. 6%, p<0.001 (Figure 4A); EFS 54% vs. 10%, p<0.001. (Figure 4B) This analysis excludes five patients inevaluable for response due to early death from toxicity or infection. EFS and OS were similar in those patients who achieved a CR vs. PR following reinduction chemotherapy.
Overall, 26 of the 38 patients who underwent CBV conditioning and autologous PBSCT survived: 20 of 26 are progression free (14 HL, 6 NHL). The 3-year OS from transplant is 64% (95% CI, 45%–78%) (HL 64%, 95%CI, 41%–80%; NHL 70%, 95% CI, 33%–89%). The 3-year EFS from transplant is 66% (95% CI, 48%–79%) (HL 65%, 95% CI, 43%–80%; NHL 70%, 95% CI, 33%–80%). (HL vs. NHL: P=NS)
The median time to relapse after CBV conditioning and autologous PBSCT for HL patients was 0.6 years (range 0.2 to 2.0 years), and for NHL patients was 0.1 year (0.1–0.2 years). Twelve patients died after PBSCT: 7 from PD, 4 from infection, and 1 from toxicity. There was no difference in efficacy between the 450 mg/m2 BCNU dose and the 300 mg/m2 BCNU dose.
The median length of first remission was 11.7 months. EFS and OS were determined for all patients from study entry and from the date of PBSCT for those with a short (<12 months) vs. a longer first remission (≥12 months). From study entry, there was a significant increase in OS in patients with an initial CR ≥ 12 vs. < 12 months (3 yr OS 70% vs. 34%, p =0.003) (Figure 5). EFS was only marginally better with length of CR ≥ 12 months vs. shorter (p=0.05). This difference was only evident in HL patients (OS: p<0.001, EFS: p=0.007) (NHL patients (OS: p=0.37, EFS: p=0.71).
The major finding of this study is that 300 mg/m2 of BCNU within a CBV conditioning regimen prior to autologous PBSCT is well-tolerated by children with relapsed/refractory lymphoma. A dose of 450 mg/m2 of BCNU was associated with unacceptable pulmonary toxicity. Five of the six patients who developed pulmonary toxicity in this study also received mantle irradiation around the time of CBV conditioning and autologous PBSCT. Reece and others first described pulmonary toxicity in patients receiving 600 mg/m2 of BCNU and irradiation.(28, 30–32) Subsequent studies in adults using either 450 mg/m2 or 300 mg/m2 of BCNU reported less pulmonary toxicity. (20, 31, 32) The BCNU dose of 300 mg/m2 was not associated with any reduction in efficacy. The efficacy of CBV conditioning in this pediatric study is comparable to that demonstrated in studies of adults. Reece et al. reported 5-year OS and EFS of 53% and 47% using a BCNU dose of 600 mg/m2 in CBV and autologous PBSCT in 56 relapsed HL patients.(28) Using 300 mg/m2 BCNU in a CBV conditioning regimen resulted in a five-year OS and failure-free of 51% and 40%, respectively.(28) Brice et al. reported an OS and progression-free survival at 4 years of 66% and 60%, respectively, in 280 patients undergoing myeloablative conditioning and autologous BMT for relapsed HL.(33) The largest series of autologous stem cell transplantation in adult NHL patients with aggressive histologies demonstrated a 43% 5-year OS and relapse free survival (RFS). (19) Poor prognostic factors in adults include “B” symptoms, extranodal disease, duration of first remission, and chemosensitivity.(28, 30, 33) We confirmed that chemosensitivity in children and adolescents with relapsed NHL and HL following reinduction therapy was associated with a better outcome for children with both HL and NHL, but a duration of first remission was only significant in predicting outcome in children with recurrent HL.
Studies in pediatric patients are confounded by small numbers and a mixture of patients receiving autologous transplant and allogeneic transplant.(25) However, this is the first pediatric prospective autologous SCT study reported that follows the outcome of patients from the time of relapse, induction failure or progression and not just at the time of conditioning and autologous SCT. This is of critical importance because most AutoSCT results only report outcomes from the time conditioning starts and not the time from disease relapse/progression. Our study notes that only 62% of patients achieved a CR/PR after reinduction therapy and were eligible for CBV conditioning and autologous PBSCT and 90% of those patients proceeded to AutoSCT. For recurrent HL in pediatric patients, a report by Verdeguer et al. from the Spanish Pediatric BMT Group reported on 20 children with relapsed or refractory HL who underwent autologous BMT; 18 with CBV (cyclophosphamide 6000 mg/m2, BCNU 300 mg/m2 and etoposide of 1000 mg/m2).(34) Most were in CR2 or greater and 14 of 18 had prior irradiation. Among all 20 children, the 5 year OS and EFS was 95 and 62%, respectively.(34) Four of five children who relapsed post-transplant were successfully salvaged. Only one death occurred from transplant related causes. The only pulmonary toxicity was pneumonia and the transplant regimens were well tolerated.(34) For recurrent NHL in pediatric patients, Bureo reported on 46 children with relapsed NHL who underwent a transplant (14 allogeneic, 32 autologous) including 21 children with LBL, 19 with BL, and 6 with large cell lymphoma (LCL). In this study where all patients were at least a CR2, 13 of the children received transplant in 1st CR, 9 due to prolonged time to achieve CR or failure of frontline therapy. Overall EFS was 58% with a median follow-up of 33 months. EFS was similar for autologous and allogeneic recipients. The only factor which predicted outcome was disease status at BMT (CR1 vs. CR2 vs. CR3 vs. more advanced disease).(35) Brugieres reported that myeloablative therapy and autologous SCT for children with relapsed ALCL is dismal in patients with early relapse (28% 3 year EFS).(36)
Of patients achieving a CR/PR, autologous PBSC apheresis was successful in 89% of the children and adolescents on this trial in achieving an adequate CD34+/kg cell yield. Engraftment results in the current study is consistent with other reported studies involving myeloablaitve conditioning and autologous PBSCT.(34)
The weaknesses of this study include a lack of consistent reinduction regimen and small numbers of patients with NHL who were eligible to proceed to CBV conditioning and autologous PBSCT either from lack of response to reinduction therapy or refusal. Most HL patients received either ICE or a combination of VI. Most NHL patients received ICE. Surprisingly, our reinduction efficacy in patients with LBL is higher than expected but the numbers are quite small.(37)
The high rate of post-transplant relapse in this study is potentially due to two mechanisms: tumor cells in the PBSC product, and/or inadequate disease control using cytotoxic agents. It is generally accepted that autologous PBSC are less likely to be contaminated with tumor cells than bone marrow. PBSC were not collected until patients achieved a CR/PR/SD with negative marrow disease which potentially reduced the risk of tumor cell contamination.
New approaches are needed to improve the EFS reported in our study. Reinduction therapy, especially for NHL patients, is not adequate. Salvage therapy might be improved by the addition of rituximab (anti-CD20 monoclonal antibody) in patients with B-cell disease (COG ANHL0121).(38) The addition of anti-T-cell or anti CD-30 monoclonal antibodies or other targeted therapy in patients with T-lymphoblastic lymphoma, ALCL, or HL may improve the reinduction rate. The use of monoclonal antibodies or other therapies post-BMT might also reduce the relapse rate.(39, 40) Improved conditioning regimens using conventional chemotherapy agents prior to autologous transplant are unlikely to significantly impact outcome.(34) However, there is considerable emerging evidence of a graft vs. lymphoma effect following allogeneic SCT.(41–46) There may be benefit from combining autologous PBSCT with reduced intensity allogeneic transplantation to add graft vs. lymphoma effects after cytoreduction.(47) Pilot studies using CBV and autologous PBSCT followed by reduced intensity allogeneic BMT show promise in children with high-risk lymphoma.(25, 48) A larger cohort with a longer follow-up is required to determine the feasibility of these approaches.
The authors would like to thank Amber Quiltan from the COG publications office for her assistance in the preparation of this manuscript.
Grant Support: Supported by grants Chair’s Grant - U10 CA98543, Statistics and Data Center Grant - U10 CA98413 from the Division of Cancer Treatment, National Cancer Institute, National Institutes of Health, Department of Health and Human Services (COG) as well as from CureSearch National Childhood Cancer Foundation. A complete listing of grant support for research conducted by CCG and POG before initiation of the COG grant in 2003 is available online at: http://www.childrensoncologygroup.org/admin/grantinfo.htm
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Presented in part at the Second International Childhood, Adolescent and Young Adult Non Hodgkin's Lymphoma Symposium New York, NY May, 2006 and at the American Society of Hematology Annual Meeting, San Diego, CA, December, 2004.
Conflict of interest: The authors have no conflicts of interest to disclose.