Abstract
Background
Radiation therapy (RT) remains a critical component of multimodality treatment for medulloblastoma. Traditionally, clinicians strive to start RT within 4–5 weeks of surgery, but the optimal timing after surgery remains unclear.
Methods
Using the National Cancer Database, we identified pediatric and adolescent patients with medulloblastoma treated with curative-intent surgery, RT, and chemotherapy. Factors associated with early or delayed RT were identified using Pearson chi-squared tests. Overall survival (OS) differences based on RT timing were compared using the Kaplan–Meier estimator with log-rank tests. Patient, tumor, and treatment characteristics associated with OS were analyzed with univariate and multivariate Cox proportional hazards models.
Results
Among the 1338 patients analyzed, early RT (defined as initiation ≤3 wk after surgery) was associated with younger age, M1–3 disease, and subtotal resection. Patients who initiated RT early had decreased 5-year OS compared with patients who initiated RT 3.1–4, 4.1–5, or >5 weeks after surgery (72.5% vs. 80.5%, 79.4%, and 77.8%, respectively; P = 0.019), but there was no significant difference in OS among the latter 3 groups (P = 0.788). On multivariate analysis, early RT versus the 3.1- to 4-week interval was significantly associated with poorer OS (adjusted hazard ratio, 1.72; 95% CI: 1.19–2.48; P = 0.004), while time to RT of >5 weeks but within 90 days of surgery did not adversely impact OS (P = 0.563).
Conclusions
In this large national database analysis, delaying RT within 90 days of surgery was not associated with inferior outcomes. Although clinical judgment remains paramount, postoperative RT timing should allow for healing and the development of an optimal treatment plan.
Keywords: medulloblastoma, radiation therapy, treatment timing
Importance of the study
We did not find an adverse survival impact with delay of RT >5 weeks but within 90 days of surgery in pediatric and adolescent medulloblastoma patients. In the absence of evidence that early postoperative RT improves survival, the timing of RT after surgery should focus on adequate postoperative healing and sufficient opportunity for case review and treatment planning. Inferior outcomes were found when radiation was started within 3 weeks of surgery.
Medulloblastoma is the most common malignant pediatric brain tumor, diagnosed in approximately 500 children per year in the United States.1 A combined modality approach utilizing surgery, radiotherapy (RT), and chemotherapy has led to high rates of long-term survival.2 Attempts to minimize radiation exposure by sparing the supratentorial brain3 or reducing radiation dose in the absence of chemotherapy4 have led to higher rates of relapse and decrement in survival.
Previous studies have demonstrated that patient and tumor characteristics, such as young age,5 the presence of metastases,5 histopathologic variants such as large cell or anaplastic,6 and group 3 molecular subgroup,7 are associated with poor outcomes following therapy. Treatment-related factors associated with long-term outcomes include the extent of tumor resection5 and radiation dose to the posterior fossa.8,9 However, the optimal timing of RT after surgery remains unclear.
Three previous randomized trials comparing immediate RT following surgery versus up-front postoperative chemotherapy and delayed RT demonstrated consistently inferior outcomes when RT was delayed longer than 4 to 22 weeks after surgery.5,10,11 It is unclear if differences in the chemotherapy regimen between arms or if myelosuppression secondary to chemotherapy contributed to inferior outcomes, but the interval between surgery and radiation is typically kept to 4 to 5 weeks in an effort to reduce time for potential tumor regrowth, while still allowing time for postsurgical healing. For example, the current Children’s Oncology Group ACNS0332 trial for high-risk medulloblastoma mandates that RT start within 31 days of surgery (NCT00392327), and St Jude’s SJMB12 trial requires that RT start within 36 days of surgery (NCT01878617).
A recent National Cancer Database (NCDB) analysis demonstrated that deferral of postoperative RT more than 90 days after surgery for young children was associated with inferior overall survival (OS).12 In this study, the majority of patients with postoperative RT deferred more than 90 days after surgery ultimately received no RT as part of their initial management. As a result, this study did not address whether the precise interval between surgery and RT impacts patient outcomes. Here, we analyzed data from the NCDB to determine the optimal timing of RT after surgery when RT is delivered as part of the initial treatment regimen.
Materials and Methods
Database and Cohort Selection
We conducted an analysis of the NCDB, a hospital-based registry that includes data from over 1500 hospitals accredited by the Commission on Cancer (CoC). It is estimated to capture 70% of all incident cancer cases in the United States annually. The data used in this study were derived from a de-identified NCDB file. The American College of Surgeons and the CoC have not verified and are not responsible for the analytical or statistical methodology used or for the conclusions drawn from these data by the investigators. This study was performed with the approval of our institutional review board.
We initially queried patients between the ages of 3 and 21 years who had histologically confirmed medulloblastoma diagnosed between 2004 and 2013. Patients with prior malignancies and extraneuraxial metastatic disease (stage M4) were excluded. We further identified patients who underwent treatment with curative intent and with radiation therapy as part of their initial management. As such, we restricted our cohort to those who had a known date of definitive surgery, who initiated adjuvant RT within 90 days of surgery, and who initiated chemotherapy within 14 days prior to radiation and 6 months after surgery. Our time interval for chemotherapy was selected in order to capture both patients who received concurrent chemoradiotherapy (CRT) (ie, chemotherapy started within 14 days before or 14 days after RT start) and those who received sequential RT and chemotherapy (ie, RT followed by chemotherapy). Patients who received chemotherapy >14 days prior to RT were excluded. Patients were required to have known follow-up and to have survived ≥3 months after surgery, as those who died within 3 months after surgery would not have been eligible to complete adjuvant therapy.
Covariates
Relevant patient, tumor, and treatment characteristics were obtained from the database. Age was categorized into approximate quartiles of 3–5, 6–8, 9–13, and 14–21 years, while race was categorized as white, black, or other. Additional covariates included sex, Charlson Deyo comorbidity score (0, 1, or 2), and year of diagnosis. Distance from reporting facility was dichotomized as ≤50 miles or >50 miles based on the “great circle” distance of the patient’s residence from the reporting facility calculated using the Haversine formula.13 Residence environment was categorized as metropolitan, urban, or rural, while median household income was dichotomized into <$35,000 and ≥$35,000 per year based on ZIP code of residence. Patient insurance type was categorized as either private/Medicare or Medicaid/uninsured. Additional tumor and treatment characteristics of interest included histology (classic/not otherwise specified [NOS], desmoplastic, or large cell), Chang-modified M stage (M0 or M1–3), extent of surgical resection (gross total resection or subtotal resection/biopsy), and total RT dose (<54 Gy or ≥54 Gy). Time interval between surgery and initiation of RT was categorized as ≤3, 3.1–4, 4.1–5, or >5 weeks (ie, ≤21, 22–28, 29–35, or 36–90 days). Chemotherapy administration was designated as concurrent CRT if initiated within 14 days before or after initiation of RT. Otherwise, chemotherapy was classified as sequential (ie, if started >14 days after initiation of RT).
Statistical Analysis
Pearson chi-squared tests were used to compare baseline demographic, tumor, and treatment characteristics across time groups. Differences in OS between groups, calculated as time to death after diagnosis, were assessed using the Kaplan–Meier estimator with log-rank testing. Univariate and multivariate Cox proportional hazards models were used to calculate unadjusted and adjusted hazard ratios (HRs), respectively. Covariates in the multivariable model were selected a priori and included the following variables from the database: age, sex, race, year of diagnosis, histology, M stage, extent of resection, total RT dose, and chemotherapy timing. The NCDB suppresses information regarding facility type (ie, academic vs community cancer program) for pediatric patients in order to protect patient confidentiality. However, given the potential confounding impact of facility type and case complexity on survival outcomes, we performed sensitivity analyses with respect to facility volume assuming that higher-volume facilities were more likely to represent academic centers and to treat more complex cases, as has been demonstrated in other disease sites.14–16 We stratified reporting facilities into quartiles based on total number of pediatric and adolescent medulloblastoma cases treated across the study time period and dichotomized at the 75th percentile into high-volume and low-volume facilities. Separate univariate and multivariate Cox proportional hazards models were generated for the high-volume facilities (ie, top quartile) and the low-volume facilities (ie, bottom 3 quartiles). All tests were 2-sided with an alpha value of 0.05. Statistical analyses were performed using SAS software v9.4.
Results
Patient, Tumor, and Treatment Characteristics
We identified 1338 patients who met all inclusion and exclusion criteria. Table 1 shows details of our cohort selection process. Median age of patients in our cohort was 8 years (interquartile range [IQR], 6‒13 y), and the majority of patients (70.3%) initiated adjuvant RT between 3 and 5 weeks after surgery. Baseline demographic, tumor, and treatment characteristics are presented in Table 2. In general, those who initiated adjuvant RT early (within 3 wk of surgery) tended to be younger, have more advanced disease at presentation (M1–3), and be more likely to undergo subtotal versus gross total resection than those who initiated RT later. Conversely, those who initiated adjuvant RT late (>5 wk after surgery) tended to be older, present with M0 disease, be of black race, and reside in areas with lower median household income.
Table 1.
Cohort formation
| Criteria | No. of Patients | Percent |
|---|---|---|
| 1. Medulloblastoma (topographic code C70.x, C71.x, C72.x, morphology code 947x) | 5036 | 100 |
| 2. Diagnosed in 2004–2013 and with known vital status | 4601 | 91 |
| 3. Diagnosis date after facility reference date | 4443 | 88 |
| 4. No prior malignancies | 4332 | 86 |
| 5. Age between 3 and 21 years | 2503 | 50 |
| 6. No distant metastatic disease (M4) | 2439 | 48 |
| 7. Histologic confirmation of diagnosis, excluding those diagnosed at autopsy | 2406 | 48 |
| 8. Known date of definitive surgery | 2293 | 46 |
| 9. Received adjuvant radiation therapy within 90 days of surgery | 1785 | 35 |
| 10. Received adjuvant chemotherapy within 14 days prior to radiation and 6 months after surgery | 1365 | 27 |
| 11. Follow-up period after surgery greater than or equal to 3 months | 1349 | 27 |
| 12. All or part of treatment was provided at reporting facility | 1338 | 27 |
Table 2.
Baseline demographic, tumor, and treatment characteristics
| Patient Characteristics | Total | Time to RT 3.1–4 weeks | Time to RT 4.1–5 weeks | Time to RT ≤3 weeks | Time to RT >5 weeks | P a | |||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| No. | % | No. | % | No. | % | No. | % | No. | % | ||
| Total | 1338 | 100.0 | 471 | 100.0 | 470 | 100.0 | 164 | 100.0 | 233 | 100.0 | |
| Age, y | <0.0001 | ||||||||||
| 3–5 | 309 | 23.1 | 100 | 21.2 | 111 | 23.6 | 52 | 31.7 | 46 | 19.7 | |
| 6–8 | 364 | 27.2 | 140 | 29.7 | 141 | 30.0 | 38 | 23.2 | 45 | 19.3 | |
| 9–13 | 337 | 25.2 | 138 | 29.3 | 115 | 24.5 | 39 | 23.8 | 45 | 19.3 | |
| 14–21 | 328 | 24.5 | 93 | 19.7 | 103 | 21.9 | 35 | 21.3 | 97 | 41.6 | |
| Sex | 0.848 | ||||||||||
| Male | 844 | 63.1 | 297 | 63.1 | 298 | 63.4 | 107 | 65.2 | 142 | 60.9 | |
| Female | 494 | 36.9 | 174 | 36.9 | 172 | 36.6 | 57 | 34.8 | 91 | 39.1 | |
| Race | 0.018 | ||||||||||
| White | 1061 | 79.3 | 389 | 82.6 | 364 | 77.4 | 131 | 79.9 | 177 | 76.0 | |
| Black | 171 | 12.8 | 48 | 10.2 | 61 | 13.0 | 18 | 11.0 | 44 | 18.9 | |
| Other | 106 | 7.9 | 34 | 7.2 | 45 | 9.6 | 15 | 9.1 | 12 | 5.2 | |
| Year of diagnosis | 0.141 | ||||||||||
| 2004–2005 | 240 | 17.9 | 89 | 18.9 | 67 | 14.3 | 38 | 23.2 | 46 | 19.7 | |
| 2006–2007 | 263 | 19.7 | 93 | 19.7 | 96 | 20.4 | 31 | 18.9 | 43 | 18.5 | |
| 2008–2009 | 290 | 21.7 | 115 | 24.4 | 103 | 21.9 | 31 | 18.9 | 41 | 17.6 | |
| 2010–2011 | 263 | 19.7 | 90 | 19.1 | 90 | 19.1 | 34 | 20.7 | 49 | 21.0 | |
| 2012–2013 | 282 | 21.1 | 84 | 17.8 | 114 | 24.3 | 30 | 18.3 | 54 | 23.2 | |
| Histology | 0.577 | ||||||||||
| Classic/NOS | 1178 | 88.0 | 417 | 88.5 | 415 | 88.3 | 148 | 90.2 | 198 | 85.0 | |
| Desmoplastic | 108 | 8.1 | 39 | 8.3 | 38 | 8.1 | 9 | 5.5 | 22 | 9.4 | |
| Large cell | 52 | 3.9 | 15 | 3.2 | 17 | 3.6 | 7 | 4.3 | 13 | 5.6 | |
| M stage | <.0001 | ||||||||||
| M0 | 1149 | 85.9 | 414 | 87.9 | 417 | 88.7 | 110 | 67.1 | 208 | 89.3 | |
| M1–3 | 130 | 9.7 | 42 | 8.9 | 33 | 7.0 | 40 | 24.4 | 15 | 6.4 | |
| Unknown | 59 | 4.4 | 15 | 3.2 | 20 | 4.3 | 14 | 8.5 | 10 | 4.3 | |
| Extent of resection | 0.032 | ||||||||||
| Gross total resection | 290 | 21.7 | 103 | 21.9 | 108 | 23.0 | 25 | 15.2 | 54 | 23.2 | |
| Subtotal resection or biopsy | 203 | 15.2 | 53 | 11.3 | 80 | 17.0 | 31 | 18.9 | 39 | 16.7 | |
| Unknown | 845 | 63.2 | 315 | 66.9 | 282 | 60.0 | 108 | 65.9 | 140 | 60.1 | |
| Total radiotherapy dose | 0.374 | ||||||||||
| 54 Gy or higher | 1026 | 76.7 | 382 | 81.1 | 353 | 75.1 | 121 | 73.8 | 170 | 73.0 | |
| <54 Gy | 207 | 15.5 | 64 | 13.6 | 76 | 16.2 | 28 | 17.1 | 39 | 16.7 | |
| Unknown | 105 | 7.8 | 25 | 5.3 | 41 | 8.7 | 15 | 9.1 | 24 | 10.3 | |
| Chemotherapy timing | 0.008 | ||||||||||
| Concurrent | 1061 | 79.3 | 394 | 83.7 | 371 | 78.9 | 126 | 76.8 | 170 | 73.0 | |
| Sequential | 277 | 20.7 | 77 | 16.3 | 99 | 21.1 | 38 | 23.2 | 63 | 27.0 | |
| Charlson Deyo comorbidity score | 0.296 | ||||||||||
| 0 | 1238 | 92.5 | 440 | 93.4 | 433 | 92.1 | 155 | 94.5 | 210 | 90.1 | |
| 1 | 62 | 4.6 | 18 | 3.8 | 20 | 4.3 | 7 | 4.3 | 17 | 7.3 | |
| 2 or higher | 38 | 2.8 | 13 | 2.8 | 17 | 3.6 | 2 | 1.2 | 6 | 2.6 | |
| Distance from reporting facility | 0.472 | ||||||||||
| ≤50 miles | 977 | 73.0 | 331 | 70.3 | 346 | 73.6 | 123 | 75.0 | 177 | 76.0 | |
| >50 miles | 343 | 25.6 | 130 | 27.6 | 121 | 25.7 | 39 | 23.8 | 53 | 22.7 | |
| Unknown | 18 | 1.3 | 0.0 | 0.0 | 0.0 | 0.0 | |||||
| Residence | 0.208 | ||||||||||
| Metropolitan | 1085 | 81.1 | 379 | 80.5 | 381 | 81.1 | 135 | 82.3 | 190 | 81.5 | |
| Urban | 179 | 13.4 | 61 | 13.0 | 72 | 15.3 | 19 | 11.6 | 27 | 11.6 | |
| Rural | 33 | 2.5 | 17 | 3.6 | 8 | 1.7 | 1 | 0.6 | 7 | 3.0 | |
| Unknown | 41 | 3.1 | 0.0 | 0.0 | 0.0 | 0.0 | |||||
| Income | 0.043 | ||||||||||
| <$35,000 | 405 | 30.3 | 124 | 26.3 | 154 | 32.8 | 46 | 28.0 | 81 | 34.8 | |
| $35,000 or higher | 876 | 65.5 | 321 | 68.2 | 303 | 64.5 | 113 | 68.9 | 139 | 59.7 | |
| Unknown | 57 | 4.3 | 26 | 5.5 | 13 | 2.8 | 5 | 3.0 | 13 | 5.6 | |
| Insurance type | 0.039 | ||||||||||
| Private/Medicare | 824 | 61.6 | 315 | 66.9 | 280 | 59.6 | 101 | 61.6 | 128 | 54.9 | |
| Medicaid/uninsured | 483 | 36.1 | 147 | 31.2 | 175 | 37.2 | 61 | 37.2 | 100 | 42.9 | |
| Unknown | 31 | 2.3 | 9 | 1.9 | 15 | 3.2 | 2 | 1.2 | 5 | 2.1 | |
aPearson’s chi-squared P-value.
Survival Analysis
Median follow-up interval for the overall cohort from date of diagnosis to death or date of last contact was 52 months (IQR, 28‒82 mo). During this time, 291 deaths were recorded (21.7% of patients). Estimated 5-year OS for the overall cohort was 78.7%, though there was a significant difference in OS depending on timing of adjuvant RT. Those who initiated adjuvant RT ≤3 weeks after surgery had shorter estimated 5-year OS compared with those who initiated RT 3–4, 4–5, or >5 weeks after surgery (72.5% vs 80.5%, 79.4%, or 77.8%, respectively; P = 0.019; Fig. 1). Furthermore, there was no significant difference in OS between the latter 3 groups on log-rank testing (P = 0.788).
Fig. 1.
Kaplan–Meier plot of OS stratified by time interval between surgery and initiation of adjuvant radiation therapy.
On univariate Cox regression analysis, time to RT of ≤3 weeks after surgery was significantly associated with increased hazard of death compared with time to RT of 3.1–4 weeks (HR 1.69; 95% CI: 1.20–2.38; P = 0.003). Additional factors associated with poorer OS were M1–3 disease compared with M0 (HR 1.74; 95% CI: 1.25–2.42; P = 0.001) and total RT dose <54 Gy compared with ≥54 Gy (HR 1.57; 95% CI: 1.18–2.09; P = 0.002). Desmoplastic histology was associated with improved OS compared with classic/NOS histology (HR 0.38; 95% CI: 0.20–0.72; P = 0.003). There was a trend toward improved OS on univariate analysis for patients who received sequential RT followed by chemotherapy, and this effect became statistically significant on multivariate regression (adjusted HR 0.70; 95% CI: 0.50–0.98; P = 0.036). All of the other aforementioned variables also remained significant on multivariate Cox regression (Table 3). Specifically, time to RT of ≤3 weeks remained significantly associated with increased hazard of death after controlling for additional demographic and treatment characteristics (adjusted HR 1.72; 95% CI: 1.19–2.48; P = 0.004). In addition, time to RT of 4.1–5 or >5 weeks after surgery was not associated with poorer OS compared with time to RT of 3.1–4 weeks on both univariate and multivariate analyses (adjusted HRs 1.21 and 1.12; 95% CIs: 0.90–1.63 and 0.77–1.63; P = 0.214 and 0.563, respectively).
Table 3.
Univariate and multivariate Cox proportional hazards models
| Variable | Univariate Analysis | Multivariate Analysis | ||
|---|---|---|---|---|
| HR (95% CI) | P | HR (95% CI) | P | |
| Time to radiotherapy, wk | ||||
| 3.1–4 | Reference | Reference | ||
| 4.1–5 | 1.10 (0.83–1.46) | 0.523 | 1.21 (0.90–1.63) | 0.214 |
| ≤3 | 1.69 (1.20–2.38) | 0.003 | 1.72 (1.19–2.48) | 0.004 |
| >5 | 1.08 (0.76–1.52) | 0.682 | 1.12 (0.77–1.63) | 0.563 |
| Age, y | ||||
| 3–5 | Reference | Reference | ||
| 6–8 | 1.04 (0.74–1.45) | 0.821 | 1.05 (0.73–1.49) | 0.804 |
| 9–13 | 1.15 (0.82–1.62) | 0.406 | 1.24 (0.86–1.78) | 0.244 |
| 14–21 | 1.21 (0.87–1.69) | 0.257 | 1.32 (0.93–1.89) | 0.124 |
| Sex | ||||
| Male | Reference | Reference | ||
| Female | 0.93 (0.73–1.18) | 0.547 | 0.87 (0.68–1.13) | 0.294 |
| Race | ||||
| White | Reference | Reference | ||
| Black | 1.22 (0.87–1.70) | 0.247 | 1.18 (0.84–1.68) | 0.340 |
| Other | 0.63 (0.37–1.08) | 0.090 | 0.65 (0.37–1.14) | 0.129 |
| Year of diagnosis | ||||
| 2004–2005 | Reference | Reference | ||
| 2006–2007 | 1.01 (0.72–1.43) | 0.950 | 1.05 (0.73–1.51) | 0.812 |
| 2008–2009 | 1.15 (0.82–1.63) | 0.418 | 1.21 (0.84–1.73) | 0.303 |
| 2010–2011 | 1.18 (0.81–1.73) | 0.399 | 1.15 (0.59–2.25) | 0.678 |
| 2012–2013 | 1.15 (0.75–1.77) | 0.530 | 1.13 (0.56–2.30) | 0.728 |
| Histology | ||||
| Classic/NOS | Reference | Reference | ||
| Desmoplastic | 0.38 (0.20–0.72) | 0.003 | 0.43 (0.23–0.81) | 0.009 |
| Large cell | 1.47 (0.89–2.44) | 0.133 | 1.51 (0.89–2.57) | 0.124 |
| M stage | ||||
| M0 | Reference | Reference | ||
| M1–3 | 1.74 (1.25–2.42) | 0.001 | 1.66 (1.16–2.36) | 0.005 |
| Unknown | 1.15 (0.66–2.00) | 0.635 | 1.16 (0.63–2.16) | 0.633 |
| Extent of resection | ||||
| Gross total resection | Reference | Reference | ||
| Subtotal resection or biopsy | 1.17 (0.75–1.82) | 0.488 | 1.12 (0.70–1.79) | 0.646 |
| Unknown | 0.97 (0.70–1.37) | 0.880 | 1.07 (0.55–2.09) | 0.836 |
| Total radiotherapy dose | ||||
| 54 Gy or higher | Reference | Reference | ||
| <54 Gy | 1.57 (1.18–2.09) | 0.002 | 1.50 (1.12–1.99) | 0.006 |
| Chemotherapy timing | ||||
| Concurrent | Reference | Reference | ||
| Sequential | 0.75 (0.55–1.01) | 0.058 | 0.70 (0.50–0.98) | 0.036 |
Our results remained unchanged on sensitivity analyses stratified by facility volume. Among patients treated at the top quartile of facilities by volume (67% of patients in our study), time to RT of ≤3 weeks remained significantly associated with increased hazard of death on both univariate and multivariate models (adjusted HR 1.53; 95% CI: 1.00–2.35; P = 0.049), while the same was found among patients treated at the bottom 3 quartiles of facilities by volume (adjusted HR 2.65; 95% CI: 1.24–5.67; P = 0.012). Time to RT of 4.1–5 or >5 weeks was not associated with decreased survival in either group of patients.
Discussion
Using the NCDB, we examined the impact of RT timing after surgery on outcomes among medulloblastoma patients who received RT as part of their initial management. Current clinical trials enrolling patients with newly diagnosed medulloblastoma stipulate that adjuvant RT must be initiated within 31 or 36 days after surgery (NCT00392327 and NCT01878617), in an effort to reduce time for tumor regrowth and to maximize the potential benefit of adjuvant therapy. However, potential benefits to a longer delay in starting postoperative irradiation include better wound healing, time to allow for pathology review and case discussion, potential reoxygenation of the tumor bed to increase radiosensitivity, and shrinkage of the surgical cavity, which allows a smaller boost volume. In this study, we found no detrimental survival impact of delaying radiation more than 5 weeks after surgery in this nationwide hospital-based cohort. Interestingly, we observed that initiation of RT within 3 weeks of surgery was associated with inferior OS.
Five retrospective reviews have reported on the impact of the time interval between surgery and radiation on medulloblastoma patient outcomes (Table 4). Similar to our study, none of the previous studies observed a detriment for delayed radiation after surgery. These reviews were small and single-institutional, and to our knowledge, no large study has previously examined the effect of timing of postoperative RT in medulloblastoma. These results, along with our data, argue for reconsideration of a strict requirement to start radiation therapy within a set time period after surgical resection, as is mandated on the current medulloblastoma protocols.
Table 4.
Previous retrospective studies analyzing time to RT after surgery for medulloblastoma
| Citation | # of Patients | Cutoff* | Endpoints | Finding |
|---|---|---|---|---|
| del Charco et al, 199825 | 53 | 45 days | OS, FFP, PFC | NS |
| Jenkin et al, 200026 | 150 | 25 days | OS | NS |
| Abacioglu et al, 200217 | 30** | 21 days | OS, DFS, PFC | Worse DFS for <21 days |
| 42 days | NS | |||
| Paulino et al, 200327 | 63 | 31 days | FFP, PFC | NS |
| Christopherson et al, 201428 | 50 | 28 days | PFC | NS |
*Interval between surgery and radiation therapy used as the cutoff for statistical analysis.
**All patients were ≥16 years old.
Abbreviations: FFP, freedom from progression; PFC, posterior fossa control; DFS, disease-free survival; NS, no significant difference observed for any endpoint.
Similar to our findings, Abacioglu et al also previously observed inferior outcomes for patients who initiated RT within 3 weeks of surgery, reporting a 5-year disease-free survival of 0% compared with 85% for patients who started radiation 3–6 weeks after surgery.17 In our study, patients who received radiation within 3 weeks of surgery tended to be younger, have more advanced disease at presentation (M1–3), and be more likely to undergo subtotal versus gross total resection than those who initiated RT later. These observations suggest that there may be a selection bias, as more aggressive tumors may be treated with RT sooner. Even after adjusting for these factors in multivariable analysis, early RT was still associated with an elevated hazard of death. While factors such as poor surgical healing and hypoxia leading to radiation resistance or an immunosuppressive microenvironment could be playing a role in this finding, it is possible that other unmeasured confounders are contributing to this observation. For example, extent of resection was unknown for 63% of the patients in our cohort, and this information could potentially help to explain the inferior outcomes among the patients who received early RT.
Recent reports of radiation timing in the adjuvant treatment of other central nervous system malignancies have generally shown a similar lack of survival benefit, and in some cases, a detriment, with earlier initiation of postoperative radiation. Among 2855 patients with glioblastoma treated on 16 Radiation Therapy Oncology Group studies, patients who initiated RT 4–6 weeks after surgery experienced improved OS compared with patients who initiated RT within 4 weeks of surgery.18 In a recent report of 161 patients with glioblastoma, there were no OS or progression-free survival (PFS) differences between patients who started RT within 28 days of surgery and those who started RT greater than 28 days after surgery.19 For those who underwent biopsy or subtotal resection, initiation of RT within 28 days was associated with decreased OS. Others have also shown decreased OS and PFS with initiation of RT within 24 days of surgery.20 The authors speculate that this may be related to hypoxia in the operative bed/residual tumor causing decreased radiosensitivity, and potentially decreased patient performance status at the time of initiation of RT.19 In a review of 198 patients with glioblastoma enrolled in 4 clinical trials receiving adjuvant RT plus temozolomide plus an experimental agent from 2004–2010, patients who began adjuvant chemoradiation at 30–34 days after surgery experienced longer OS and PFS than those who began treatment within 30 days of surgery.21 Data from the MD Anderson Cancer Center and Texas Children’s Hospital suggest that earlier initiation of postoperative RT (31.0 vs 46.0 days) is associated with greater risk of developing imaging changes after treatment for pediatric ependymoma, with average grade of imaging changes higher among patients with posttreatment symptoms.22 In addition, rat models have demonstrated increased damage to normal brain tissue with earlier initiation of RT postoperatively.23 Despite this accumulating data, clinical judgment will remain critical to identify patients with high-risk features who may be at risk of neurologic deterioration with a prolonged delay to the initiation of RT.
Consistent with previous studies,5,8,9 we observed an increased hazard of death for patients with M1–3 disease and patients who received less than a total dose of 54 Gy on both univariate and multivariate analyses. In addition, desmoplastic histology was associated with a favorable prognosis, which has also been previously reported.6 Although large cell histology was not a significant negative prognostic factor in our analysis, our findings were limited by the small proportion (3.9%) of patients with large cell tumors in our sample. Similarly, the extent of resection was reported in only 36.9% of the patients analyzed, and we did not observe a significant difference in survival based on this factor. Notably, a recent multi-institutional analysis found that the extent of resection was less prognostic when the molecular subgroup was taken into account.24 Interestingly, we found that sequential RT followed by chemotherapy was associated with improved survival compared with concurrent CRT on multivariate analysis. Etiology of this finding is unclear but may lend further evidence for allowing sufficient time for postoperative healing. We observed no differences in survival based on sex, race, or year of diagnosis.
Limitations of our study include the retrospective nature of our analysis, which may be subject to selection bias and information bias. Our analysis depended on the accuracy of the data encoded in the NCDB, and our results could have been affected by misclassification or missing information. For example, the type and doses of chemotherapy as well as molecular subtype information are not included in the database, and additional patient information including performance status is not recorded. Furthermore, while we attempted to control for known prognostic factors, factors such as M stage and extent of resection were not available for every patient. Other potential confounding factors include case complexity and facility type (ie, academic vs community hospital), which is suppressed in the NCDB for pediatric patients. As such, we performed sensitivity analyses using facility volume as a surrogate for facility type, assuming that academic centers treat higher volumes of patients and more complex cases. Our results remained unchanged on these analyses.
Conclusions
Using the data from a large national hospital-based database, we observed no detrimental impact on survival with delay of radiation therapy within 90 days of surgery. Given the relatively small number of medulloblastoma cases diagnosed each year, it would be difficult to carry out a prospective randomized trial to determine the optimal timing of RT after surgery. Although certain high-risk clinical factors may warrant early initiation of RT, our analysis suggests that the timing of RT after surgery should focus on adequate postoperative healing and sufficient time for case review, evaluation for treatment at a childhood cancer center16 with the potential for proton therapy if deemed appropriate, and complex treatment planning. Potential theoretical benefits include decreased hypoxia, increased radiosensitization, and a smaller boost cavity size, with no current evidence for poorer outcomes if initiation of RT is delayed.
Funding
This research was supported in part by the KL2 Mentored Career Development Award of the Stanford Clinical and Translational Science Award to Spectrum (NIH KL2 TR 001083) (to E.L.P.).
Conflict of interest statement
Dr Soltys serves as a consultant for Inovio Pharmaceuticals, Inc. The other authors declare that there are no conflicts of interest.
References
- 1. Smoll NR, Drummond KJ. The incidence of medulloblastomas and primitive neurectodermal tumours in adults and children. J Clin Neurosci Off J Neurosurg Soc Australas. 2012;19(11):1541–1544. [DOI] [PubMed] [Google Scholar]
- 2. Packer RJ, Gajjar A, Vezina G et al. Phase III study of craniospinal radiation therapy followed by adjuvant chemotherapy for newly diagnosed average-risk medulloblastoma. J Clin Oncol Off J Am Soc Clin Oncol. 2006;24(25):4202–4208. [DOI] [PubMed] [Google Scholar]
- 3. Bouffet E, Bernard JL, Frappaz D et al. M4 protocol for cerebellar medulloblastoma: supratentorial radiotherapy may not be avoided. Int J Radiat Oncol Biol Phys. 1992;24(1):79–85. [DOI] [PubMed] [Google Scholar]
- 4. Thomas PR, Deutsch M, Kepner JL et al. Low-stage medulloblastoma: final analysis of trial comparing standard-dose with reduced-dose neuraxis irradiation. J Clin Oncol Off J Am Soc Clin Oncol. 2000;18(16):3004–3011. [DOI] [PubMed] [Google Scholar]
- 5. Zeltzer PM, Boyett JM, Finlay JL et al. Metastasis stage, adjuvant treatment, and residual tumor are prognostic factors for medulloblastoma in children: conclusions from the Children’s Cancer Group 921 randomized phase III study. J Clin Oncol Off J Am Soc Clin Oncol. 1999;17(3):832–845. [DOI] [PubMed] [Google Scholar]
- 6. Rutkowski S, von Hoff K, Emser A et al. Survival and prognostic factors of early childhood medulloblastoma: an international meta-analysis. J Clin Oncol Off J Am Soc Clin Oncol. 2010;28(33):4961–4968. doi:10.1200/JCO.2010.30.2299. [DOI] [PubMed] [Google Scholar]
- 7. Northcott PA, Jones DTW, Kool M et al. Medulloblastomics: the end of the beginning. Nat Rev Cancer. 2012;12(12):818–834. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Berry MP, Jenkin RD, Keen CW, Nair BD, Simpson WJ. Radiation treatment for medulloblastoma. A 21-year review. J Neurosurg. 1981;55(1):43–51. [DOI] [PubMed] [Google Scholar]
- 9. Hughes EN, Shillito J, Sallan SE, Loeffler JS, Cassady JR, Tarbell NJ. Medulloblastoma at the joint center for radiation therapy between 1968 and 1984. The influence of radiation dose on the patterns of failure and survival. Cancer. 1988;61(10):1992–1998. [DOI] [PubMed] [Google Scholar]
- 10. Bailey CC, Gnekow A, Wellek S et al. Prospective randomised trial of chemotherapy given before radiotherapy in childhood medulloblastoma. International Society of Paediatric Oncology (SIOP) and the (German) Society of Paediatric Oncology (GPO): SIOP II. Med Pediatr Oncol. 1995;25(3):166–178. [DOI] [PubMed] [Google Scholar]
- 11. Kortmann RD, Kühl J, Timmermann B et al. Postoperative neoadjuvant chemotherapy before radiotherapy as compared to immediate radiotherapy followed by maintenance chemotherapy in the treatment of medulloblastoma in childhood: results of the German prospective randomized trial HIT ‘91. Int J Radiat Oncol Biol Phys. 2000;46(2):269–279. [DOI] [PubMed] [Google Scholar]
- 12. Kann BH, Park HS, Lester-Coll NH et al. Postoperative radiotherapy patterns of care and survival implications for medulloblastoma in young children. JAMA Oncol. 2016;2(12):1574–1581. [DOI] [PubMed] [Google Scholar]
- 13. Sinnott RW. Virtues of the Haversine. Sky Telesc. 1984;68(2):158. [Google Scholar]
- 14. Chen Y-W, Mahal BA, Muralidhar V et al. Association between treatment at a high-volume facility and improved survival for radiation-treated men with high-risk prostate cancer. Int J Radiat Oncol. 2016;94(4):683–690. [DOI] [PubMed] [Google Scholar]
- 15. Go RS, Bartley AC, Crowson CS et al. Association between treatment facility volume and mortality of patients with multiple myeloma. J Clin Oncol. 2016;35(6):598–604. [DOI] [PubMed] [Google Scholar]
- 16. Wolfson J, Sun CL, Wyatt L, Stock W, Bhatia S. Adolescents and young adults with acute lymphoblastic leukemia and acute myeloid leukemia: impact of care at specialized cancer centers on survival outcome. Cancer Epidemiol Biomarkers Prev. 2017;26(3):312–320. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Abacioglu U, Uzel O, Sengoz M, Turkan S, Ober A. Medulloblastoma in adults: treatment results and prognostic factors. Int J Radiat Oncol Biol Phys. 2002;54(3):855–860. [DOI] [PubMed] [Google Scholar]
- 18. Blumenthal DT, Won M, Mehta MP et al. Short delay in initiation of radiotherapy may not affect outcome of patients with glioblastoma: a secondary analysis from the radiation therapy oncology group database. J Clin Oncol. 2009;27(5):733–739. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Randolph DM 2nd, McTyre ER, Paulsson AK et al. Impact of timing of radiotherapy in patients with newly diagnosed glioblastoma. Clin Neurol Neurosurg. 2016;151:73–78. [DOI] [PubMed] [Google Scholar]
- 20. Adeberg S, Bostel T, Harrabi S et al. Impact of delays in initiating postoperative chemoradiation while determining the MGMT promoter-methylation statuses of patients with primary glioblastoma. BMC Cancer. 2015;15:558. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Han SJ, Rutledge WC, Molinaro AM et al. The effect of timing of concurrent chemoradiation in patients with newly diagnosed glioblastoma. Neurosurgery. 2015;77(2):248–253; discussion 253. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Gunther JR, Sato M, Chintagumpala M et al. Imaging changes in pediatric intracranial ependymoma patients treated with proton beam radiation therapy compared to intensity modulated radiation therapy. Int J Radiat Oncol Biol Phys. 2015;93(1):54–63. [DOI] [PubMed] [Google Scholar]
- 23. Peker S, Abacioglu U, Sun I, Yuksel M, Pamir MN. Irradiation after surgically induced brain injury in the rat: timing in relation to severity of radiation damage. J Neurooncol. 2004;70(1):17–21. [DOI] [PubMed] [Google Scholar]
- 24. Thompson JC, Yee SS, Troxel AB et al. Detection of therapeutically targetable driver and resistance mutations in lung cancer patients by next-generation sequencing of cell-free circulating tumor DNA. Clin Cancer Res. 2016;22(23):5772–5782. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. del Charco JO, Bolek TW, McCollough WM et al. Medulloblastoma: time-dose relationship based on a 30-year review. Int J Radiat Oncol Biol Phys. 1998;42(1):147–154. [DOI] [PubMed] [Google Scholar]
- 26. Jenkin D, Shabanah MA, Shail EA et al. Prognostic factors for medulloblastoma. Int J Radiat Oncol Biol Phys. 2000;47(3):573–584. [DOI] [PubMed] [Google Scholar]
- 27. Paulino AC, Wen BC, Mayr NA et al. Protracted radiotherapy treatment duration in medulloblastoma. Am J Clin Oncol. 2003;26(1):55–59. [DOI] [PubMed] [Google Scholar]
- 28. Christopherson KM, Bradley JA, Rotondo RL et al. Local control in non-metastatic medulloblastoma. Acta Oncol. 2014;53(9):1151–1157. [DOI] [PubMed] [Google Scholar]