SEARCH

SEARCH BY CITATION

Keywords:

  • neuroblastoma;
  • contemporary therapy;
  • MYCN;
  • induction

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

BACKGROUND

The authors exploited a large database to investigate the outcomes of patients with high-risk neuroblastoma in the contemporary era.

METHODS

All patients with high-risk neuroblastoma aged <12 years who were treated during induction at the authors' institution from 2000 through 2011 were studied, including 118 patients with MYCN-amplified [MYCN(+)] disease and 127 patients aged >18 months with MYCN-nonamplified [MYCN(−)] stage 4 disease.

RESULTS

A complete response/very good partial response (CR/VGPR) to induction was correlated with significantly superior event-free survival (EFS) (P < .001) and overall survival (OS) (P < .001) compared with a partial response or less. Patients with MYCN(+) and MYCN(−) disease had similar rates of CR/VGPR to induction (P = .366), and those with MYCN(+) and MYCN(−) disease who attained a CR/VGPR had similar EFS (P = .346) and OS (P = .542). In contrast, only MYCN(+) patients had progressive disease as a response to induction (P < .001), and early death from progressive disease (<366 days after diagnosis) was significantly more common (P < .001) among those with MYCN(+) disease. Overall, among patients who had a partial response or less, MYCN(+) patients had significantly inferior EFS (P < .001) and OS (P < .001) compared with MYCN(−) patients, which accounted for the significantly worse EFS (P = .008) and OS (P = .002) for the entire MYCN(+) cohort versus the MYCN(−) cohort.

CONCLUSIONS

Patients with MYCN(−), high-risk neuroblastoma display a broad, continuous spectrum with regard to response and outcome, whereas MYCN(+) patients either have an excellent response to induction associated with good long-term outcome or develop early progressive disease with a poor outcome. This extreme dichotomy in the clinical course of MYCN(+) patients points to underlying biologic differences with MYCN(+) neuroblastoma, the elucidation of which may have far-reaching implications, including improved risk classification at diagnosis and the identification of targets for treatment. Cancer 2014;120:2050–2059. © 2014 American Cancer Society.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

Population-based data[1-3] and group studies through the 1990s[4-11] demonstrated 5-year event-free survival (EFS) and overall survival (OS) rates of 17% to 37% for patients with high-risk neuroblastoma (HR-NB) and little difference between EFS and OS (see Table 1, Studies completed in the 1990s). In large studies that extended beyond the year 2000, 5-year EFS and OS rates improved to 35% to 64% (see Table 1, Studies completed in the 2000s).[12-16] These results may represent overestimations of survival, because subsequent analyses revealed a more limited definition of HR-NB. Those studies included: 1) patients ages 12 to 18 months who had MYCN-nonamplified [MYCN(−)] stage 4 disease, which is now recognized as intermediate-risk disease[17, 18]; and 2) toddlers who had MYCN(−) stage 3 disease with unfavorable histology, which may be cured by limited therapy.[19] Indeed, a group study conducted from 1999 to 2004 that focused on a well recognized high-risk group—ie, MYCN-amplified [MYCN(+)] NB in infants—yielded 2-year EFS and OS rates of 29% and 30%, respectively,[14] and an international experience that covered the years from 1997 to 2002 for patients aged >18 months who had MYCN(+) stage 4 NB yielded an EFS rate of 18%.[18]

Table 1. Studies of High-Risk Neuroblastoma With Long-Term Follow-Up
Reference (Study Years)No. of PatientsStage and Age of Study PopulationPrognostic Impact of Response/MYCN(+)Staging Studies at DX/FULocal RTCRAAnti-GD2 ImmunotherapyEFSOS
  1. Abbreviations, —, not reported or described; CRA, 13-cis-retinoic acid; DX, diagnosis; FU, follow-up; MYCN(+), MYCN-amplified; RT, radiotherapy; UH, unfavorable histology.

  2. a

    Studies are listed chronologically according to the year each study was completed.

  3. b

    The 5-year EFS and OS rates were 25% and 27%, respectively, for patients with stage 3 MYCN(+) neuroblastoma and 77% and 79%, respectively, for patients with stage 3 MYCN-nonamplified neuroblastoma (Park et al 2009[11]).

Studies completed in the 1990s       
Zage 2008[9] (1993-1995)142Stage 4, >1 y; stage 2 or 3, MYCN(+), >1 y—/——/Postprotocol every 3-6 mo for 2 yYesNoNo23% at 7 y28% at 7 y
Matthay 2009[10] (1991-1996)539Stage 4, >1 y; stage 4, <1 y, MYCN(+); stage 3, MYCN(+); stage 3, UH and/or ferritin ≥143 ng/mLYes/yesMIBG optional/at end of protocolNoRandomizedNo26% at 5 y (stage 4, 21%; stage 3, 55%)36% at 5 y (stage 4, 32%; stage 3, 59%b)
Berthold 2003[7] (1990-1997)335Stage 4, >1 y—/——/—NoNoNo27% at 5 y33% at 5 y
DeBernardi 2003[6] (1992-1997)159Stage 4, >1 yNone/yes—/—NoNoNo17% at 5 y28% at 5 y
Kaneko 2002[5] (1991-1998)221Stage 4, >1 y—/NoneMIBG optional/—NoNoNo34% at 5 y37% at 5 y
Castel 2001[4] (1992-1998)83Stage 4, >1 y; stage 4, <1 y, MYCN(+); stage 3, MYCN(+)None/yes—/—NoNoNo33% at 4 y33% at 4 y, 20% at 5 y
Pearson 2008[8] (1990-1999)262Stage 4, >1 yYes/noneMIBG optional/“local preference”NoNoNo28% at 3 y, 24% at 5 y, 23% at 10 y26% at 5 y, 24% at 10 y
Studies completed in the 2000s       
George 2006[12] (1994-2002)97Stage 4, >1 y; stages 3, 4, 4S, <1 y, MYCN(+); stage 3, >1 y, UH—/NoneMIBG variable/—YesYesNo55% at 3 y, 47% at 5 y, 46% at 7 y72% at 3 y, 60% at 5 y, 53% at 7 y
Berthold 2005[13] (1997-2002)295Stage 4, >1 y; all stages MYCN(+)Yes/yes—/—n = 24n = 39n = 16039% at 3 y, 35% at 5 y58% at 3 y, 45% at 5 y
Canete 2009[14] (1999-2004)35Stages 2, 3, 4, 4S, <1 y, MYCN(+)Yes/—MIBG if available/—YesYesNo29% at 2 y30% at 2 y; stage 4, 20% at 2 y
Sung 2007[15] (1997-2005)52Stage 4, >1 yNone/noneMIBG included/—n = 35YesNo62% at 5 y64% at 5 y
Kreissman 2013[16] (2001-2006)495Stage 4, >1 y; stages 3, 4, 4S, <1 y, MYCN(+); stage 3, >1 y, UH—/——/—YesYesn = 7838% at 5 y50% at 5 y

Better results have been expected with the recent adoption of therapies—many of which were developed in the 1990s—promising more effective induction (dose-dense or dose-intensive chemotherapy),[8, 20, 21] consolidation (13-cis-retinoic acid,[10] local radiotherapy,[22] anti-GD2 monoclonal antibody [MoAb][23, 24]), and salvage (131I-metaiodobenzylguanidine [MIBG] therapy,[25] topoisomerase 1 inhibitors[26-28]). We used the large Memorial Sloan-Kettering Cancer Center (MSKCC) database to investigate the outcome of HR-NB in the contemporary era, ie, since 2000. We took into account MYCN amplification, because large studies differ regarding whether this finding is[4, 6, 10, 13] or is not[5, 8, 12, 15] prognostic in HR-NB (Table 1). We also assessed response to induction vis-a-vis EFS and OS, given differences in past studies regarding whether a good response does[8, 10, 13, 14] or does not[4, 6, 15] have an impact on the survival of patients with HR-NB (Table 1). The data revealed a striking dichotomy in the clinical course of patients with MYCN(+) disease, a scenario not observed in patients with MYCN(−) HR-NB.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

Study subjects were identified from the NB patient registry of MSKCC. This source listed 1185 patients for the period from 2000 to 2011. The current study was limited to the 247 patients who were aged <12 years when diagnosed with HR-NB and who received treatment during induction at MSKCC from the time of diagnosis or after starting induction elsewhere. Excluded were patients who came for consultation alone or only for surgery. The definition of HR-NB conformed with current criteria: MYCN(+) stage 2, 3, 4S, or 4 disease in patients of any age or MYCN(−) stage 4 disease diagnosed at age >18 months.[17, 18] MYCN amplification was defined by international criteria using fluorescence in situ hybridization.[29] In accordance with rules of the MSKCC Institutional Review Board, informed written consents for evaluations and treatments were obtained from guardians; and, for this exploratory study, as a retrospective review, a waiver was obtained for examination and analysis of patient records.

Disease status was defined according to widely accepted international response criteria,[6, 8, 11, 14-16, 30] including 123I-MIBG findings. Complete remission (CR) was defined as no evidence of NB in soft tissue, bones, or bone marrow (BM) and normal urine catecholamine levels. Very good partial remission (VGPR) was defined as a primary mass reduced by ≥90%; no evidence of distant disease in soft tissue, bones, or BM, including a negative 123I-MIBG scan; and normal urine catecholamine levels. A partial response (PR) indicated a decrease >50% in measurable soft tissue disease and in the number of metastatic skeletal lesions on a 123I-MIBG scan and ≤1 positive BM site. A mixed response indicated a decrease >50% in any lesion with a decrease <50% in any other and a 123I-MIBG scan improved by a <50% decrease in the number of positive sites. No response was defined as a decrease <50% but an increase <25% in any existing lesion and unchanged MIBG findings. Finally, progressive disease (PD) was defined as a new lesion or an increase >25% in an existing lesion.

A complete evaluation of disease status comprised computed tomography or magnetic resonance imaging studies, a 123I-MIBG scan, urine catecholamine levels, and BM histology (aspirates and biopsies from bilateral posterior iliac crests and aspirates with or without biopsies from bilateral anterior iliac crests). BM and imaging studies were read by MSKCC specialists outside the Department of Pediatrics who were unaware of treatment or patient status. A complete evaluation was performed at the end of induction. Patients who achieved CR/VGPR underwent a complete evaluation at least every 3 months for an additional 2 years and then had 123I-MIBG scans with or without other staging studies every 3 months for 3 more years. Patients who achieved only a PR or less (≤PR) with induction underwent a complete evaluation every 1 to 3 months while on therapy; if they achieved CR/VGPR, then a complete evaluation was repeated every 3 months for 3 more years.

The software package SPSS (version 11.0; SPSS Inc., Chicago, Ill) was used for the statistical analyses of survival, calculating from the first day of induction chemotherapy. Survival curves were generated according to the Kaplan-Meier method, with point estimates that included ± the standard error, and were compared using a 2-sided log-rank test. EFS continued through the date of PD, toxic death, or secondary cancer. OS was defined through the date of death from any cause. The 2-tailed chi-square test was used for comparisons of response rates to induction therapy.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

Patient Characteristics

The total series of patients included 118 MYCN(+) patients (1 patient with stage 2 disease, 13 patients with stage 3 disease, 1 patient with stage 4S disease, and 103 patients with stage 4 disease), 127 MYCN(−) patients with stage 4 disease diagnosed at age >18 months, and 2 patients with unknown MYCN status. The latter 2 patients were excluded from further analysis, leaving a final total of 245 patients for the current study (Table 2). Thirty-one patients were infants (ie, aged <18 months), and all had MYCN(+) disease. MYCN(+) patients and MYCN(−) patients received the same upfront and salvage therapies. Thus, the initial (first-line) induction regimens were all for HR-NB: 223 patients (91%) received Children's Oncology Group[16, 21] (or similar MSKCC[20]) regimens, and 22 patients (9%) received other group-wide[8, 9] or single-institutional programs. Consolidation of CR/VGPR included anti-GD2 MoAb with or without autologous stem cell transplantation (ASCT).[24] Initial second-line treatments for refractory or progressive disease with induction included high-dose conventional chemotherapy[31-34] or moderate-dose regimens[28, 35] using agents with known anti-NB activity. Relapse also was treated uniformly, including intrathecal radioimmunotherapy for central nervous system relapse,[36] high-dose conventional chemotherapy[31-34] for disseminated or soft-tissue relapse, and moderate-dose chemotherapy[28, 35] plus radiotherapy for focal skeletal relapse. Patients who achieved second CR/VGPR received MoAb.

Table 2. Patient Characteristics
 No. of Patients (%)
CharacteristicMYCN(+), n = 118MYCN(−), n = 127Entire cohort, n = 245
  1. Abbreviations: ASCT, autologous stem-cell transplantation; CR/VGPR, complete remission/very good partial remission; MoAb, immunotherapy with anti-GD2 monoclonal antibody; MR/NR, minor response/no response; MYCN(+), MYCN-amplified; MYCN(−), MYCN-nonamplified; PD, progressive disease; ≤PR, partial remission or less.

  2. a

    These included 1 patient with stage 4S disease and 3 patients with stage 3 disease.

  3. b

    These included patients who had PD as a response to induction and patients who relapsed <366 days from diagnosis.

  4. c

    These included 2 deaths during induction (tumor lysis syndrome, pulmonary failure), 1 transplantation-related toxic death, and 1 secondary myelodysplastic syndrome (detected at 42 months and successfully treated).

  5. d

    These included 2 transplantation-related toxic deaths, 2 secondary leukemias (diagnosed at 28 months and 58 months and successfully treated), 1 death from hemolytic-uremic syndrome during induction, and 1 late death in an accident without prior relapse.

  6. e

    These were previously reported alkylator-based conventional chemotherapy regimens (see the 4 reports by Kushner and colleagues[31-34]).

  7. f

    These did not include 2 patients who died of toxicity during first-line induction.

No. of males:females [ratio]61:57 [1.07]47:53 [0.89]108:110 [0.98]
Age at diagnosis, y   
Range0.12-10.971.54-11.820.12-11.82
Median2.003.653.02
Induction regimens   
A3973/N889 (75)95 (75)184 (75)
ANBL053220 (17)19 (15)39 (16)
Other9 (8)13 (10)22 (9)
Responses to induction   
CR/VGPR68 (58)59 (46)127 (52)
PR13 (11)22 (17)35 (14)
MR/NR13 (11)46 (36)59 (24)
PD24 (20)a0 (0)24 (10)
Early PD: <366 db50 (42)17 (13)67 (27)
Early death from PD: <366 d26 (22)6 (5)32 (13)
Non-neuroblastoma events4 (3)c6 (5)d10 (4)
Consolidation of first CR/VGPR   
MoAb40/68 (59)34/59 (58)74/127 (58)
ASCT and MoAb26/68 (38)24/59 (41)50/127 (39)
No ASCT or MoAb2/68 (3)1/59 (2)3/127 (2)
Initial treatment for ≤PR   
High-dose chemotherapye37/48 (77)f52/68 (76)89/116 (77)
ASCT and MoAb5/48 (10)f5/68 (7)10/116 (9)

Responses to Induction (Table 2)

CR/VGPR was achieved with first-line induction treatment in 127 of 245 patients (52%; 95% confidence interval [CI], 45%-58%), including 68 of 118 (58%; 95% CI, 48%-67%) MYCN(+) patients and 59 of 127 (46%; 95% CI, 38%-56%) MYCN(−) patients (P = .366). PR rates also were similar between the MYCN(+) cohort (11%; 95% CI, 6%-18%) and the MYCN(−) cohort (17%; 95% CI, 11%-25%; P = .2). In contrast, a PD response to induction was limited to those with MYCN(+) disease, occurring in 24 of 118 (20%; 95% CI, 13%-29%) MYCN(+) patients compared with 0 of 127 MYCN(−) patients (P < .001).

EFS and OS Correlated With MYCN

For the entire cohort of 245 patients, the EFS rates at 3 years, 5 years, and 7 years were 44%, 37%, and 35%, respectively; and the OS rates were 67%, 57%, and 53%, respectively. MYCN(+) patients (n = 118) had significantly inferior EFS and OS compared with MYCN(−) patients (n = 127) (Table 3, Fig. 1). Thus, at 3 years, 5 years, and 7 years, the EFS rates were 37%, 32%, and 32%, respectively, in MYCN(+) patients versus 49%, 41%, and 38%, respectively, in MYCN(−) patients (P = .008); and the OS rates were 56%, 48%, and 46% versus 77%, 66%, and 60%, respectively (P = .002).

Table 3. Event-Free Survival and Overall Survival
 EFS Rate, % OS Rate, % 
Patient Subset3 Years5 Years7 YearsP3 Years5 Years7 YearsP
  1. Abbreviations CR/VGPR, complete remission/very good partial remission; EFS, event-free survival; MYCN(+), MYCN-amplified; MYCN(−), MYCN-nonamplified; OS, overall survival; ≤PR, partial remission or less.

All patients, n = 245443735 675753 
MYCN(+), n = 118373232.008564846.002
MYCN(−), n = 127494138 776660 
CR/VGPR, n = 127585150< .001807470> .001
≤PR, n = 118282119 533835 
All MYCN(+) patients, n = 118        
Infants, n = 31343434.824555555.671
Children, n = 87393232 564543 
Patients in CR/VGPR, n = 127        
MYCN(+), n = 68544848.346756868.542
MYCN(−), n = 59635452 868072 
Patients in ≤PR, n = 118        
MYCN(+), n = 501499< .001291814< .001
MYCN(−), n = 68373026 705350 
image

Figure 1. (A) Event-free survival and (B) overall survival are illustrated for all patients who had MYCN-amplified and MYCN-nonamplified high-risk neuroblastoma.

Download figure to PowerPoint

Early treatment failure was significantly more common among patients with MYCN(+) disease. PD <366 days after diagnosis (including PD response to induction) emerged in 50 of 118 (42%; 95% CI, 33%-52%) MYCN(+) patients versus 17 of 127 (13%; 95% CI, 8%-21%) MYCN(−) patients (P < .001); and death from PD <366 days after diagnosis occurred in 26 of 118 (22%; 95% CI, 15%-31%) MYCN(+) patients versus 6 of 127 (5%; 95% CI, 2%-10%) MYCN(−) patients (P < .001) (Table 2). Early death among patients who achieved CR/VGPR with first-line induction was rare: this occurred in 1 of 26 MYCN(+) patients who received consolidation with ASCT plus MoAb, in 1 of 40 MYCN(+) patients who received consolidation with MoAb, in 0 of 24 MYCN(−) patients who received consolidation with ASCT plus MoAb, and 1 of 34 MYCN(−) patients who received consolidation with MoAb.

Among the MYCN(+) patients, infants (n = 31) and children (n = 87) had overlapping outcomes (Table 3). Thus, at 3 years, 5 years, and 7 years, the EFS rates were 34%, 34%, and 34%, respectively, in infants compared with 38%, 31%, and 31%, respectively, in children (P = .828); and the corresponding OS rates were 55%, 55%, and 55%, respectively, in infants compared with 56%, 44%, and 42%, respectively, in children (P = .664).

Correlation of EFS and OS With Response to Induction

EFS and OS rates for the patients who achieved CR/VGPR with initial induction therapy (n = 127) were significantly better than those for the patients who had ≤PR (n = 118) (Table 3; Fig. 2). Thus, at 3 years, 5 years, and 7 years, the EFS rates were 58%, 51%, and 50%, respectively, for patients who achieved CR/VGPR versus 28%, 21%, and 19%, respectively, for those who had <PR (P < .001); and the corresponding OS rates were 80%, 74%, and 70%, respectively, for patients who achieved CR/VGPR versus 53%, 38%, and 35%, respectively, for those who had <PR (P < .001).

image

Figure 2. (A) Event-free survival and (B) overall survival are illustrated for all patients with high-risk neuroblastoma who achieved complete remission/very good partial remission (CR/VGPR) compared with patients who achieved a partial response or less (≤PR).

Download figure to PowerPoint

Among the patients who achieved CR/VGPR, the EFS and OS rates were similar for the MYCN(+) patients (n = 68) and the MYCN(−) patients (n = 59) (Table 2, Fig. 3). Thus, at 3 years, 5 years, and 7 years, the EFS rates were 54%, 48%, and 48%, respectively, for MYCN(+) patients versus 63%, 54%, and 52%, respectively, for MYCN(−) patients (P = .346); and the OS rates were 75%, 68%, and 68%, respectively, for MYCN(+) patients versus 86%, 80%, and 72%, respectively, for MYCN(−) patients (P = .542). These results indicate that the long-term OS rates for MYCN(+) patients and MYCN(−) patients who attained CR/VGPR after induction were approximately 20% greater than their respective long-term EFS rates.

image

Figure 3. (A) Event-free survival and (B) overall survival are illustrated for patients with MYCN-amplified and MYCN-nonamplified high-risk neuroblastoma who achieved complete remission/very good partial remission (CR/VGPR).

Download figure to PowerPoint

Among the patients who had ≤PR to induction, EFS and OS rates for MYCN(+) patients (n = 50) were significantly inferior to those for MYCN(−) patients (n = 68) (Table 3, Fig. 4). Thus, at 3 years, 5 years, and 7 years, the EFS rates were 14%, 9%, and 9%, respectively, for MYCN(+) patients versus 37%, 30%, and 26%, respectively, for MYCN(−) patients (P < .001); and the OS rates were 29%, 18%, and 14%, respectively, for MYCN(+) patients versus 70%, 53%, and 50%, respectively, for MYCN(−) patients (P < .001). These results demonstrate that the long-term EFS and OS rates for the MYCN(+) patients were similar, whereas the long-term EFS and OS rates for the MYCN(−) patients differed by approximately 20% to 30%.

image

Figure 4. (A) Event-free survival and (B) and overall survival are illustrated for patients with MYCN-amplified and MYCN-nonamplified high-risk MYCN-neuroblastoma who achieved a partial response or less (≤PR).

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

Patients with MYCN(−) HR-NB display a broad, continuous spectrum with regard to response and survival, whereas patients with MYCN(+) HR-NB either have an excellent response to induction associated with a good long-term outcome or develop early PD with a poor outcome. This extreme dichotomy in the clinical scenarios of MYCN(+) patients may result from underlying, distinctive genetic features. The elucidation of these divergent biologic profiles may have far-reaching implications, including improved risk classification at diagnosis and the identification of targets for treatment. Unfortunately, there are currently no known biologic findings at diagnosis that can predict the outcome of MYCN(+) patients, because virtually all MYCN(+) NBs are associated with certain biomarkers (eg chromosome 17q gain and chromosome 1p deletion) and not with others (eg chromosome 11q deletion and ATRX gene mutations).

The MYCN(+) and MYCN(−) cohorts in the current study received the same induction, consolidative, and salvage treatments (Table 2); this uniformity lends validity to our comparison of outcomes. A key finding was PD response to first-line induction with MYCN(+) disease, before consolidation could start. With regard to widely used postinduction treatments, among patients who achieved CR/VGPR with first-line induction, early death was rare, occurring in 2 of 66 MYCN(+) patients consolidated with ASCT plus MoAb (n = 26) or with MoAb (n = 40), and in 1 of 58 MYCN(−) patients consolidated with ASCT plus MoAb (n = 24) or with MoAb (n = 34). Initial treatment of refractory disease or PD involved chemotherapy with well recognized anti-NB activity rather than investigative agents.

Patients with MYCN(+) and MYCN(−) NB have equivalent CR/VGPR rates with induction (P = .366), and MYCN is not prognostic for patients in CR/VGPR at completion of induction. Thus, MYCN(+) and MYCN(−) patients who achieve CR/VGPR with induction have similar long-term EFS and OS (Table 3, Fig. 3). In distinct contrast, PD response to induction is limited to MYCN(+) patients (P < .001), and early PD is significantly associated with MYCN(+) disease (P < .001). Early treatment failure not only adversely affects EFS but contributes to a dismal OS, because PD of HR-NB during, or soon after completion of, induction responds poorly to salvage therapy,[31] and short time from diagnosis to PD is a significant adverse factor for OS.[37-41] Indeed, MYCN(+) patients are significantly more likely than MYCN(−) patients to die early from PD (P < .001), ie, <366 days from diagnosis. Overall, among patients with ≤PR to induction, MYCN(+) disease is associated with significantly inferior EFS (P < .001) and OS (P < .001); these results account for the significantly worse long-term EFS (P = .008) and OS (P = .002) of the entire MYCN(+) cohort compared with the entire MYCN(−) cohort (Table 3, Fig. 1). Age had no prognostic impact on MYCN(+) disease: infants and children with this chromosomal aberration had similar long-term EFS and OS (Table 3).

The PD findings, combined with the significantly worse EFS and OS of MYCN(+) patients compared with MYCN(−) patients who attain ≤PR postinduction, suggest that a major response to upfront therapy is crucial for a good outcome in MYCN(+) patients and is less crucial in MYCN(−) patients. This point is supported by the absence of continued decline in EFS for MYCN(+) patients over the long term, whereas EFS rates for MYCN(−) patients decrease steadily beyond 3 years (Table 3, Fig. 1). The findings reflect a paucity of late relapses among patients who have MYCN(+) HR-NB compared with those who have MYCN(−) HR-NB.

The paramount prognostic importance of initial response for MYCN(+) patients, but not for MYCN(−) patients, is also illustrated by the relation between EFS and OS rates within subsets of patients. Thus, the MYCN(+) patients who had ≤PR to induction had long-term EFS and OS rates that approximated each other (Table 3, Fig. 4). This similarity between EFS and OS is consistent with a rapid demise of these patients after PD/relapse and is attributable to the persistence of underlying chemoresistance as well as ineffective second-line therapy (see below). In contrast, MYCN(+) patients who achieved CR/VGPR with induction had a long-term OS that was approximately 20% higher than their EFS: evidence of prolonged survival postrelapse attributable to continued chemosensitivity (Table 3, Fig. 3). This difference between the outcomes of MYCN(+) patients in ≤PR versus those in CR/VGPR was not seen among the patients with MYCN(−) HR-NB; indeed, all patients with MYCN(−) HR-NB—those in CR/VGPR and those with ≤PR after induction—had long-term OS rates that, like the MYCN(+) patients in CR/VGPR postinduction, were approximately 20% higher than their EFS rates (Table 3, Figs. 3 and 4).

These large differences between long-term EFS and OS during the period from 2000 to 2011 among all patients with MYCN(−) HR-NB and among the MYCN(+) patients in CR/VGPR, but not among those in ≤PR postinduction, were not seen in the 1990s (Table 1).[4-11] In that decade, EFS and OS were closely related, which supported the view that PD/relapse of HR-NB was synonymous with a subsequent rapid demise from PD or from the toxicity of salvage therapy. Exceptions to this scenario led to the initial reports of long-term survival in children with HR-NB despite persistence or relapse of disease.[38, 42] This phenomenon of chronic NB was much less common in MYCN(+) patients than in MYCN(−) patients.[37-42]An underlying biologic factor for chronicity may be mutations of the ATRX gene, which recently were found to be significantly associated with MYCN(−) HR-NB in older patients,[43] in whom NB is often characterized by an indolent and prolonged, although ultimately lethal, course.[44]

Consistent with a low likelihood of prolonged survival after MYCN(+) PD/relapse, EFS and OS were poor and virtually identical in the few reports that offered details about MYCN(+) patients. Thus, in a multicenter study of MYCN(+) infants from 1999 to 2004, the EFS and OS rates at only 2 years were 29% and 30%, respectively, for all stages (the 2-year OS rate was 20% for patients with stage 4 disease)[14]; reviews of an international experience from 1990 to 2002 found that MYCN(+) stage 4 and 4S disease in patients aged <18 months had 5-year EFS and OS rates of 28% and 34%, respectively,[45] and that MYCN(+) stage 4 disease in patients aged >18 months had 5-year EFS and OS rates <25% each[17]; and a group-wide study from 1991 to 1996 of patients with MYCN(+) stage 3 disease (all ages) yielded 5-year EFS and OS rates of 25% and 27%, respectively.[11]

One reason why the above-described clinical scenario, which was typical in the 1990s (ie, death soon after PD/relapse), may no longer hold true is the availability of novel, relatively nontoxic salvage (ie, second-line) therapies that lack cross-resistance with induction. Examples include chemotherapy regimens (eg, irinotecan-temozolomide[28]), investigative agents (eg, fenretinide, ABT-751, crizotinib),[46] and targeted radiotherapy (131I-MIBG[25] and 131I-monoclonal antibodies[36]). Another factor contributing to prolonged survival may be improved disease surveillance. In the 1990s, MIBG scintigraphy often used 131I and was not regularly performed (Table 2)[5, 8, 10]; subsequently, 123I-MIBG became widely adopted and proved to be superior in detecting NB, especially relapse that is asymptomatic (and presumably has lower tumor burden), which may be more amenable to control than bulky metastatic relapse.[47] Unfortunately, the welcome prospect that the recent advances in therapy and surveillance might, after relapse, lead to a chronic course with long-term survival plus good quality of life, or even a cure, is not relevant to patients who develop early PD—a subset of patients associated significantly with MYCN(+) NB (P < .001) (Table 3).

In conclusion, results from the 1970s through the contemporary era indicate that better upfront, consolidative, and salvage therapies have improved the survival of patients with HR-NB.[1-3, 12-15, 18, 23, 24] Yet MYCN(+) NB differs significantly from MYCN(−) HR-NB with regard to 1) PD response to induction, and 2) extreme differences in outcome, ie, early death from disease or excellent PFS and OS. The 2 sharply divergent clinical scenarios of MYCN(+) patients merit investigation at the molecular/genetic level, analogous to other investigations into the underlying biology of NB,[43, 46] to identify markers predictive at diagnosis of good response/good outcome versus poor response/early demise and to expand the availability of targets for therapy.

FUNDING SUPPORT

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

This study was supported in part by grant CA10450 from the National Institutes of Health (Bethesda, Md), the Robert Steel Foundation (New York, NY), and Katie's Find-A-Cure Fund New York, New York).

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES