Upfront observation versus radiation for adult pilocytic astrocytoma

Authors

  • Adrian Ishkanian MSc, MD,

    1. Department of Radiation Oncology, Princess Margaret Hospital, University of Toronto, Toronto, Ontario, Canada
    Search for more papers by this author
  • Normand J. Laperriere MD,

    1. Department of Radiation Oncology, Princess Margaret Hospital, University of Toronto, Toronto, Ontario, Canada
    Search for more papers by this author
  • Wei Xu PhD,

    1. Department of Biostatistics, Princess Margaret Hospital, University of Toronto, Toronto, Ontario, Canada
    Search for more papers by this author
  • Barbara-Ann Millar MD,

    1. Department of Radiation Oncology, Princess Margaret Hospital, University of Toronto, Toronto, Ontario, Canada
    Search for more papers by this author
  • David Payne MD,

    1. Department of Radiation Oncology, Princess Margaret Hospital, University of Toronto, Toronto, Ontario, Canada
    Search for more papers by this author
  • Warren Mason MD,

    1. Department of Medicine, Princess Margaret Hospital, University of Toronto, Toronto, Ontario, Canada
    Search for more papers by this author
  • Arjun Sahgal MD

    Corresponding author
    1. Department of Radiation Oncology, Princess Margaret Hospital, University of Toronto, Toronto, Ontario, Canada
    2. Department of Radiation Oncology, Princess Margaret Hospital and the Sunnybrook Health Sciences Center, University of Toronto, Toronto, Ontario, Canada
    • Princess Margaret Hospital, 610 University Avenue, Toronto, ON, M5G 2M9, Canada
    Search for more papers by this author
    • Fax: (416) 946-2111


Abstract

BACKGROUND:

Although pilocytic astrocytoma accounts for up to 40% of all childhood brain tumors, it is a rare disease in adults. Consequently, there are few mature data on the impact of up-front treatment options after surgery that include observation or adjuvant radiotherapy.

METHODS:

Ten women and 20 men were identified who were diagnosed with pilocytic astrocytoma from 1971 to 2007 and were retrospectively reviewed. The median patient age was 30 years (range, 18-64 years), and the median follow-up was 87 months (range, 16-420 months). Initial surgery included biopsy (10% of patients), subtotal resection (57% of patients), or gross-total resection (33% of patients). Nineteen patients were observed postoperatively, whereas 11 patients received up-front postoperative adjuvant radiotherapy (50 grays in 25 fractions). No patient received adjuvant or concurrent chemotherapy. Progression-free survival (PFS) and overall survival (OS) were calculated using the Kaplan-Meier method. Differences between survival curves were analyzed with the log-rank test.

RESULTS:

For the entire cohort, the 5-year and 10-year OS rates were 95% and 85%, respectively, and the 5-year and 10-year PFS rates were 63% and 35%, respectively. The median PFS was 8.4 years. Initial radiation, compared with observation, did not have an impact on OS but significantly improved PFS. The 5-year PFS rate for patients who were observed versus those who received radiation was 42% versus 91%, respectively; and, at 10 years, the PFS rate was 17% versus 60%, respectively (P = .005). Patients who progressed after observation (11 of 19 patients) received various salvage therapies, resulting in a 2-year PFS rate of 68% compared with 33% for patients who progressed after initial radiation (3 of 11 patients) and were salvaged with either chemotherapy or surgery (P = .1).

CONCLUSIONS:

Adjuvant radiotherapy for pilocytic astrocytoma significantly prolonged PFS at both 5 years and 10 years compared with observation. However, equivalent OS was observed, which reflected the efficacy of salvage therapies. Cancer 2011;. © 2011 American Cancer Society.

Pilocytic astrocytoma (World Health Organization grade 1) accounts for up to 40% of all childhood brain tumors.1-4 However, this is a rare disease in adults; consequently, there are few mature data to guide treatment policy with respect to initial postsurgical management. Typically, patients are either observed or offered up-front adjuvant radiation. There also is a poor understanding regarding the outcomes of patients who receive treatment at first and second progression according to initial management. In this, we have documented the long-term outcomes of patients with adult pilocytic astrocytoma who were treated at the Princess Margaret Hospital (PMH), Toronto, Ontario, Canada, between 1971 and 2007. For this study, we specifically analyzed the impact of postoperative observation as opposed to adjuvant radiation for both progression-free survival (PFS) and overall survival (OS), and we report subsequent outcomes for those patients who developed progressive disease.

MATERIALS AND METHODS

Patients

Thirty adult patients, including 10 women and 20 men, were identified from the PMH registry who had pilocytic astrocytoma diagnosed between ages of 18 and 64 years (median age, 30 years) and were retrospectively reviewed. The dates of surgery ranged from 1971 to 2007. The primary disease location included 13 supratentorial tumors, 13 infratentorial/brainstem tumors, and 4 spinal cord tumors. All patients were diagnosed with pilocytic astrocytoma, and pathology was reviewed at the time of consultation at the University Health Network, University of Toronto, Toronto, Canada. Initial surgery was biopsy alone in 3 patients, subtotal resection (STR) in 17 patients, and gross total resection (GTR) in 10 patients. Nineteen of 30 patients were observed postoperatively and received no adjuvant therapy, and 11 of 30 patients received adjuvant treatment with postoperative radiotherapy. No patient received chemotherapy either concurrently or in an adjuvant setting. Adjuvant postoperative radiotherapy was defined as radiation given within 6 months of the primary operation. All patients received with 5000 centigrays in 25 fractions. The radiotherapy technique consisted of a mixture of computed tomography (CT)-based 3-dimensional conformal technology, and 2-dimensional conventional technology. Essentially, a limited number of fields was used to encompass at least a 2-cm margin beyond the tumor according to the contrast-enhanced CT information and/or magnetic resonance imaging (MRI) studies. The median patient follow-up was 87 months (range, 16-420 months).

Statistical Analysis

PFS and OS were calculated using the Kaplan-Meier method. Differences between survival curves were analyzed using the log-rank test. All P values were 2-sided. Results were considered significant if P ≤ .05. Statistical analyses were performed using version 9.2 of the SAS system and user's guide (SAS Institute, Inc., Cary, NC).

RESULTS

The presenting signs and symptoms (before any surgical intervention) in all patients are summarized in Table 1. Headache was the most common, followed by visual deficits and nausea/vomiting. After surgery, 19 patients (63%) were observed, and 11 patients (37%) received adjuvant postoperative radiotherapy (Table 2). Of the 27 patients who underwent STR or GTR, 11 received postoperative radiation, and 16 were observed. All 3 patients who underwent biopsy-only had been observed.

Table 1. Presenting Signs and Symptoms
Signs and SymptomsBiopsy Group, n=3STR Group, n=17GTR Group, n=10
  1. STR indicates subtotal resection; GTR, gross total resection.

Headache3118
Visual changes190
Nausea/vomiting044
Incoordination034
Gait disturbance/ataxia052
Sensory changes131
Weakness022
Back pain020
Slurred speech011
Seizure110
Vertigo001
Decreased cognition100
Hemorrhage010
Table 2. A Treatment Summary of the Entire Cohort
Patient No.Primary SiteType of SurgeryPostsurgical TherapyTime to First Progression, moTreatment of First ProgressionTime to Second Progression, moTreatment of Second Progression
  • STR indicates subtotal resection; OBS, observation; Sx, surgery; XRT, radiotherapy; GTR, gross total resection; CT, chemotherapy and temozolomide unless otherwise noted; Bx, biopsy; CCNU, carmustine.

  • a

    These patients received a total XRT dose of 5000 centigrays in 25 fractions.

  • b

    This patient received a total XRT dose of 5400 centigrays in 30 fractions.

1Spinal cordSTROBS36Sx48Sx
2InfratentorialSTROBS101Sx+XRT
3InfratentorialSTRXRTa
4SupratentorialSTRXRTa
5InfratentorialGTROBS
6InfratentorialGTROBS
7SupratentorialGTROBS
8InfratentorialGTROBS36Sx132Sx+XRT
9InfratentorialSTRXRTa120Sx+CT12CT
10Spinal cordSTRXRTa76CT
11InfratentorialBxOBS88XRT6CT
12InfratentorialGTRXRTa
13SupratentorialSTRXRTa60CT10Sx
14SupratentorialSTRXRTa
15InfratentorialSTROBS31XRT
16InfratentorialSTROBS
17Spinal cordGTRXRTa
18SupratentorialSTRXRTa
19SupratentorialBxOBS50CT4XRT
20InfratentorialSTROBS33XRT
21InfratentorialGTROBS
22SupratentorialSTRXRTa
23InfratentorialSTROBS27XRT
24SupratentorialGTROBS
25Spinal cordGTRXRTa
26SupratentorialGTROBS48Sx
27SupratentorialBxOBS12XRT
28SupratentorialSTROBS
29SupratentorialSTROBS10XRT+CTb72 (CCNU)
30SupratentorialSTROBS

For the entire cohort, the 5-year and 10-year OS rate was 95% and 85%, respectively, and the 5-year and 10-year PFS rate was 63% and 35%, respectively (Fig. 1). The median PFS was 8.4 years. There was no significant difference in OS or PFS according to tumor location (data not shown). We also observed no significant difference in PFS or OS for the patients who underwent biopsy, STR, or GTR (data not shown). For those who died in this cohort, the cause of death in all patients was secondary to progression of their known intracranial disease.

Figure 1.

These Kaplan-Meier survival curves illustrate (Top) overall survival and (Bottom) progression-free survival for the entire cohort.

Upfront Observation Versus Adjuvant Radiation

Outcomes were assessed according to up-front management, which was observation after initial surgery in 19 of 30 patients and postoperative adjuvant radiation in 11 of 30 patients. The 5-year OS probability for the observation group versus the postoperative radiation group was 91% versus 100%, respectively, and the 10-year OS probability was 91% versus 80%, respectively (P = .94) (Fig. 2). The 5-year PFS probability for the observation group versus the postoperative radiation group was 42% versus 91%, respectively, and the 10-year PFS probability was 17% versus 60%, respectively (Fig. 2). The median time to progression in the observation group was 4 years and had not been reached in the postoperative radiation group. The improved PFS favoring adjuvant postoperative radiotherapy was statistically significant (P = .005) at both 5 years and 10 years (Fig. 2).

Figure 2.

These Kaplan-Meier survival curves illustrate (Top) overall survival and (Bottom) progression-free survival for the patients who initially underwent observation compared with the patients who received postoperative radiation.

Outcomes of Patients Who Progressed: First Progressors

Fourteen of 30 patients (47%) developed disease progression (Table 2). All recurrences were confirmed radiographically using either CT and/or MRI studies. Thirteen of 14 patients developed progression at the primary site of disease, and 1 of 14 patients progressed beyond the volume encompassed within the initial radiation fields. Despite progression, 80% of all progressors were alive at 10 years, and the median OS had not been reached. The 5-year PFS probability after progression was 39%, and the median PFS probability after progression was 3.7 years.

Figure 3 summarizes the patients who progressed according to the initial treatment approach and describes subsequent treatments and outcomes. Three of the 14 progressors had received previous radiation, and salvage therapy consisted of either further surgery plus chemotherapy (1 of 3 patients) or chemotherapy alone (2 of 3 patients). Eleven of the 14 patients had progressed locally after observation, and salvage therapy consisted of further surgery (3 of 11 patients), radiotherapy alone (5 of 11 patients), chemotherapy alone (1 of 11 patients), surgery plus postoperative radiotherapy (1 of 11 patients), and surgery followed by chemotherapy (1 of 11 patients). Figure 4 illustrates the OS and PFS after first progression separated according to initial management approach (observation vs adjuvant radiation). No significant differences were observed for OS (P = .43) or PFS (P = .17). Among the patients who received treatment for disease progression, we also compared those who received radiotherapy alone after observation (n = 5) with those who underwent further surgery and received any additional adjuvant therapy (n = 5), and we observed no significant difference in OS (P = 1.0) or PFS (P = .70). In addition, among those who received treatment for disease progression, we compared the patients who underwent surgery alone (n = 3) versus the patients who received radiotherapy and any additional adjuvant treatment (n = 6), and we observed no significant difference in OS (P = 1.0) or PFS (P = .55).

Figure 3.

This chart is a summary of treatment and outcomes for the entire cohort. OS indicates overall survival; PFS, progression-free survival.

Figure 4.

These Kaplan-Meier survival curves illustrate (Top) overall survival and (Bottom) progression-free survival after first progression with patients separated according to the initial management approach (patients who initially underwent observation compared with patients who received postoperative radiation).

Outcomes of Patients Who Progressed: Second Progressors

Seven patients had a second disease progression documented. Two of these patients underwent salvage surgery alone, 1 received radiotherapy alone, 3 received chemotherapy alone, and 1 underwent surgery followed by radiation. The 5-year OS probability after second progression for all 7 patients was 63%, and the median survival had not been reached. There was no statistically significant difference in OS or PFS according to the type of salvage therapy (P = .72).

DISCUSSION

Pilocytic astrocytoma is a rare disease in adults. To our knowledge, there are no randomized studies (and few mature data overall) comparing up-front postoperative radiation with observation to guide optimal management. Because of the long natural history of the disease as a grade 1 tumor, it has been suggested that observation after a GTR is safe, and radiation is reserved for salvage5; however, there is controversy regarding the use of adjuvant radiation after STR or biopsy.

Here, we report on a series of 30 patients with adult pilocytic astrocytoma who had a median follow-up of 7.3 years. In our cohort, 63% of patients initially were observed, and 37% initially received postoperative radiation. The published literature, as summarized in Table 3, reveals that outcomes typically are reported without necessarily stratifying for these 2 groups. In comparing outcomes from the current series with outcomes reported in the literature, our OS and PFS rates at 5 years are consistent. However, we specifically report OS and PFS for patients who received initial adjuvant radiation versus initial observation. At 5 years, the OS rate was 100% for the radiation group versus 91% for the observation group, and the 5-year PFS rate was 91% versus 42%, respectively. Given the long follow-up in our series, we also report 10-year OS and PFS rates for the radiation group versus the observation group. At 10 years, the OS rate was 80% versus 91%, respectively, and the 10-year PFS rate was 60% versus 17%, respectively. When comparing these 2 groups, we conclude that postoperative radiation results in a statistically significant improvement in PFS at both 5 years and 10 years (Fig. 2). The absolute improvement in PFS was approximately 50% at 5 years and approximately 40% at 10 years and favored up-front radiotherapy (P = .005). No effect was observed for OS with the receipt of postoperative radiotherapy at either 5 years or 10 years, which likely reflects the indolent course of this disease, even with progression, and the success of salvage therapies. It is noteworthy that all deaths were secondary to progression of the known pilocytic astrocytoma; and, as such, cause-specific survival paralleled the OS analysis.

Table 3. Literature Summary of Outcomes in Patients With Adult Pilocytic Astrocytomaa
VariableBrown 20045Minehan 199513Stuer 20078Robinson 200516Ellis 200914Bell 200417Minehan 200912Kim 200111Current Series
  • Bx indicates biopsy; STR, subtotal resection; GTR, gross total resection; XRT, radiotherapy; CT, chemotherapy; Sx, surgery; NR, not reported; OS, overall survival; FFR, freedom from recurrence; PFS, progression-free survival.

  • a

    Values shown are the number of patients unless noted otherwise.

No. of patients204344142010691830
Median age, y32393140.5384531.93630
OS, %         
 5 y9587100NRNRNR95
 10 y8177757485
 20 y60
PFS, %         
 5 y95NR72931-y FFR, 94%NRNRNR63
 10 y678035
No. observed upfront17/2010/4338/444/1420/209/1027/698/1819/30
 5-y PFS in observation group, %94NRNRNRNRNRNRNR42
 5-y OS in observation group, %10089NRNRNRNR89NR91
No. who received postoperative XRT3/2033/436/4410/140/200/1042/6910/1811/30
 5-y PFS in XRT group, %100NRNRNRNRNRNRNR91
 5-y OS in XRT group, %6785NRNRNRNR85NR100
Median follow up, y1011.24.610.22.752.811.22.77.3

Our 5-year PFS rate of 42% in the observation group may be considered below what we would expect for pilocytic astrocytoma; however, only 1 other series has reported outcomes based on initial observation versus radiotherapy in adults.5 Brown et al reported an exceptional 5-year PFS rate of 94% with observation; however, there were few patients in that series (n = 20); therefore, the results reported may not be representative of the population as a whole. Furthermore, for children, in whom the disease is most common, lower 5-year PFS rates of approximately 70% have been reported specifically for pilocytic astrocytoma.6 When considering the patients who received initial radiation in our series, the 5-year PFS rate was 91%, consistent with the reported literature for pilocytic astrocytoma postradiation.5, 7-9 By comparison, among patients who predominantly had grade 2 glioma (low-grade glioma [LGG]), the 5-year PFS rates were 35% in patients who were observed and 55% in patients who received radiation in a randomized trial that tested observation alone versus radiation led by the European Organization for the Research and Treatment of Cancer10 (EORTC). Therefore, once again, our results are consistent for pilocytic astrocytoma, and the 5-year PFS rate was better in our observation group than that reported in patients with LGG. Moreover, with immediate radiation, better outcomes were observed in our series, such that only a minority of patients progressed (5-year PFS rate, 91%) as opposed to the 55% expected control rate with LGG and immediate radiation. When OS is compared, our results again are consistent for pilocytic astrocytoma, with a 5-year OS rate of 91% for our observation group compared with an approximately 70% expected 5-year OS rate for patients who were observed in the LGG study.10 We also report a 10-year OS rate of 85% for the entire cohort, and this observed long-term survival also is consistent for pilocytic astrocytoma.

There are several limitations to the current study that should be considered, because it was a retrospective review. First, there is potential for pathologic misdiagnosis, although all patients had their pathology results reviewed at the time of referral by neuropathology at the University Health Network (a large tertiary referral center for neuro-oncology). Furthermore, no modifications in pathologic criteria for pilocytic astrocytoma have been suggested during the study period, and all cases were typical for pilocytic astrocytoma in terms of tumor location and radiographic appearance. Therefore, misdiagnosis is unlikely but possible. Second, in our series, all 3 patients who underwent biopsy only were observed postoperatively and eventually progressed. Given our small sample size, this may have affected the results toward a lower PFS in the observation group versus the radiation group. However, we observed no statistical difference in outcome based on type of initial surgery (biopsy, STR, or GTR; data not shown). This finding has also been supported by other investigators, who reported that more extensive surgery failed to yield better outcomes.11-13 For example, Minehan et al reported on 43 adult patients with pilocytic astrocytomas cases, including 30 patients who underwent biopsy only, 10 patients who underwent STR, and 3 patients who underwent GTR.13 Those authors concluded that the extent of surgery did not have a significant impact on PFS or OS. However, the opposite also has been reported from series in which more extensive surgery resulted in better outcomes. For example, based on a series of 44 patients, Stuer et al concluded that the extent of resection was a significant predictor of PFS, and a trend was observed toward a better OS.8 Therefore, although the impact of surgical debulking is controversial, the observation group in our series may have been at a disadvantage with respect to PFS. Conversely, the group that received treatment with up-front radiation may have been at an advantage with respect to PFS, because 3 patients who underwent GTR also received adjuvant radiation. Modern practice supports the use of observation after GTR given the expected long-term control that can be achieved with GTR alone.5 However, once again, we note that results regarding the impact of surgery on PFS are conflicting; in fact, 2 patients in our observation group developed disease progression after undergoing GTR. The lack of mature outcome analyses in the literature for patients with this disease who undergo observation also limits the ability to make comparisons with the current series. We contend that our results simply may be more reflective of the actual PFS in adults with pilocytic astrocytoma who are observed as opposed to those who receive radiation. The clinical implication of the finding that with observation approximately 60% of patients progress within 5 years, and approximately 80% progress by 10 years, may be that patients who have tumors in eloquent areas of the brain or incompletely resected, in which disease progression may yield significant neurologic deficits, should be considered for up-front adjuvant radiation.

When comparing the observation group with the adjuvant radiation group, our analysis yielded a significant difference in PFS with up-front radiation and no difference in OS. This finding is in keeping with patients with LGG according to the EORTC randomized study of patients who initially received observation versus radiation.10 However, for patients with pilocytic astrocytoma this is a controversial finding; because, to date, no similar randomized study has been performed in this patient population. Moreover, few reports have even compared PFS according to up-front treatment. In the series by Brown et al, 5-year PFS rates of 94% with observation and 100% with radiation were reported; however, the difference was not statistically significant. In that series, only 3 of 20 patients received adjuvant radiotherapy; therefore, comparisons cannot be made.5 Others have reported analyses based on OS comparisons, but not PFS comparisons. For example, Minehan et al reported on 61% of 69 patients who received postoperative radiotherapy as opposed to observation. They concluded that there was no significant OS benefit on multivariate analysis for up-front radiation, and PFS was not reported. Kim et al observed a longer median survival with postoperative radiotherapy in 10 of 18 patients who received radiation; however, the difference was not statistically significant. In that series, once again, PFS was not specified.11 Our current study is consistent, in that no significant benefit in OS was observed for postoperative radiation. However, we also analyzed PFS and observed a significant PFS benefit in favor of adjuvant radiation. Again, this is not unreasonable if you consider the therapeutic impact of up-front radiation for LGG9 and that pilocytic astrocytoma represents a much more indolent glioma with a known response to radiation.5, 8, 9, 12, 14

Outcomes after first progression also are reported in our series, and there are few similar data in the literature. Although we report on patients who received treatment for progression after observation (n = 11) a prolonged 2-year PFS rate (68%), as opposed to the 2-year PFS rate (33%) in patients who received treatment for progression after radiation (n = 3), the difference was not significant (Fig. 4). In addition, no significant difference was observed in OS (Fig. 4). Because few patients belonged to this category, these data are descriptive and represent the positive effect of salvage therapies in yielding prolonged survival despite progression. We performed an exploratory analysis comparing salvage radiotherapy alone versus surgery plus any additional salvage therapy, and we compared salvage surgery alone versus radiotherapy plus any additional salvage therapy. We were unable to reach any significant conclusions regarding the potential benefits of 1 salvage therapy compared with another. This may reflect the small sample size and the heterogeneity in the salvage therapies that were used (Fig. 3). Therefore, these data can conclude only that salvage therapy is effective; because, despite disease progression, 80% of patients remained at 10 years (the median OS had not been reached), and the 5-year PFS rate was 39% (the median PFS after progression was 3.7 years). Therefore, salvage therapy has to be individualized based on tumor factors (ie, location of the tumor and degree of progression), patient factors (ie, neurology at time of progression), and treatment factors (ie, amenable to further surgery or not). For second progressors (n = 7), we also observed long-term survival. Their 5-year OS rate was 63%, and their median survival had not been reached. Therefore, all possible salvage therapies should be offered, because long-term survival still can be expected.

Other limitations to this study include our inability to determine whether any of the presenting signs or symptoms became better or worse after definitive therapy and whether definitive therapy resulted in a detriment to quality of life. In particular, we were unable to ascertain the long-term neurocognitive sequelae of radiation, which may negate gains in PFS, especially because OS remains the same. However, in this regard, radiation tends to be tolerated better in adults than in children.15 Therefore, it is unknown whether the side effects of radiation may be worse in the long-term than allowing for slow progression of the disease and delaying radiation. Finally, because the period of this study spanned the years from 1971 to 2007, the patients received treatment and were imaged with a mixture of old and new surgical, radiation, and imaging techniques (early patients in this series were in the pre-MRI era). Therefore, the conclusions offered by this study may differ with current technology.

In conclusion, the results from our series of 30 patients with pilocytic astrocytoma suggest that approximately 60% of patients are at risk of disease progression at 5 years if they initially are observed postsurgery and that postoperative radiotherapy can delay disease progression significantly but does not have an impact on OS. For patients who have tumors located in eloquent areas of the brain in which progression may render significant neurologic deficits, postoperative radiation should be considered. Salvage therapy is effective, and treatment options must be individualized to the unique circumstances of each patient.

CONFLICT OF INTEREST DISCLOSURES

The authors made no disclosures.

Ancillary