Article Tools
Neurooncology
Article Tools
Role of Extent of Resection in the Long-Term Outcome of Low-Grade Hemispheric Gliomas
Abstract
The prognostic role of extent of resection (EOR) of low-grade gliomas (LGGs) is a major controversy. We designed a retrospective study to assess the influence of EOR on long-term outcomes of LGGs.
The study population (N = 216) included adults undergoing initial resection of hemispheric LGG. Region-of-interest analysis was performed to measure tumor volumes based on fluid-attenuated inversion-recovery (FLAIR) imaging.
Median preoperative and postoperative tumor volumes and EOR were 36.6 cm3 (range, 0.7 to 246.1 cm3), 3.7 cm3 (range, 0 to 197.8 cm3) and 88.0% (range, 5% to 100%), respectively. There was no operative mortality. New postoperative deficits were noted in 36 patients (17%); however, all but four had complete recovery. There were 34 deaths (16%; median follow-up, 4.4 years). Progression and malignant progression were identified in 95 (44%) and 44 (20%) cases, respectively. Patients with at least 90% EOR had 5- and 8-year overall survival (OS) rates of 97% and 91%, respectively, whereas patients with less than 90% EOR had 5- and 8-year OS rates of 76% and 60%, respectively. After adjusting each measure of tumor burden for age, Karnofsky performance score (KPS), tumor location, and tumor subtype, OS was predicted by EOR (hazard ratio [HR] = 0.972; 95% CI, 0.960 to 0.983; P < .001), log preoperative tumor volume (HR = 4.442; 95% CI, 1.601 to 12.320; P = .004), and postoperative tumor volume (HR = 1.010; 95% CI, 1.001 to 1.019; P = .03), progression-free survival was predicted by log preoperative tumor volume (HR = 2.711; 95% CI, 1.590 to 4.623; P ≤ .001) and postoperative tumor volume (HR = 1.007; 95% CI, 1.001 to 1.014; P = .035), and malignant progression-free survival was predicted by EOR (HR = 0.983; 95% CI, 0.972 to 0.995; P = .005) and log preoperative tumor volume (HR = 3.826; 95% CI, 1.632 to 8.969; P = .002).
With the exception of patients with intractable epilepsy1 or symptomatic mass effect, the role of surgery in the treatment of infiltrative low-grade gliomas (LGGs) is a major controversy.2-7 At present, the only agreed-upon surgical standard for adults with suspected or known supratentorial nonoptic-pathway LGG is to obtain a tissue diagnosis before active treatment commences.8 The decision to operate beyond a simple biopsy rests on the presumption that reduction of tumor volume will improve long-term survival or quality of life by delaying recurrence and malignant progression.2 The role of surgical resection in the long-term outcome of LGG patients is complicated by a literature that lacks prospective randomized controlled trials addressing these issues and significant methodologic limitations of most studies.5
Although there is growing evidence suggesting that more extensive resection at the time of initial diagnosis may be a favorable prognostic factor,5 several management strategies, including simple biopsy without resection, will continue to be theoretically acceptable and will continue to be practiced until better evidence is available. It is unlikely that prospective randomized studies will be conducted to address the role of extent of resection (EOR) on outcome in LGG patients because of the relatively limited numbers of patients, the typically long survival times, and a general lack of equipoise with regard to treatment options among care providers. Nonetheless, recognizing the need to better define the optimal treatment strategy, and after carefully reviewing the limitations of previous studies,5 we have designed a retrospective study to volumetrically assess the influence of EOR on long-term outcomes in adult patients with hemispheric LGGs who are candidates for surgical resection.
Adult patients (age ≥ 18 years) who underwent initial surgery at University of California, San Francisco (UCSF; San Francisco, CA) for a hemispheric low-grade infiltrative glioma between 1989 and 2005 and who had preoperative and immediately postoperative magnetic resonance imaging were identified. Patients undergoing biopsy only were excluded because this small subset of cases typically had complicating factors, and inclusion of these cases would have inappropriately biased the results. Patients with gemistocytic histology9 or gliomatosis cerebri were excluded. Central pathology review was performed based on WHO guidelines.10 Clinical data were collected from patient records and telephone interviews. Clinical interpretation of follow-up imaging was predominantly by two neuro-oncologists (S.M.C. and M.D.P.) and reviewed by a neurosurgeon (J.S.S.). This study was approved by the UCSF Committee on Human Research.
Three outcome measures were assessed: overall survival (OS), progression-free survival (PFS), and malignant progression–free survival (MPFS). OS was defined as the time between initial surgery and death. PFS was defined as the time between initial surgery and demonstration of unequivocal increase in tumor size on follow-up imaging, malignant progression, and/or death. MPFS was defined as the time between initial surgery and demonstration of gadolinium enhancement on follow-up imaging and/or higher-grade tumor on subsequent biopsy or death. Patients with no known progression/malignant progression were censored as of their last scan date.
Analyses were done to assess whether conclusions were dependent on specific definitions. In addition to calculating the time to assessed end points on the basis of the date of initial surgery, we also calculated on the basis of the start of symptoms. With rare exception, date of symptom onset was readily determined from medical records. The results were not substantially different from those using outcomes measures based on initial surgical date. Therefore, only the results based on the initial data are presented.
Since some patients were not followed regularly with scans, sensitivity analyses were performed to assess whether irregular imaging intervals significantly affected outcome measures. If these patients progressed or died, alternative assumptions were made as to when progression occurred. First, analyses were done assuming that progression occurred at the time of death or at the time of the scan showing progression. As a check on the assumptions, two alternatives were also assessed–1) assignment of progression date one day after the last scan showing no progression and 2) assignment of progression date midway between the last scan showing freedom from progression and the date of progression/death used in the first analysis. Conclusions about the role of the three volumetric measures as prognostic factors did not change. Therefore only the initial planned analyses are presented.
Manual segmentation was performed with region-of-interest analysis to measure tumor volumes (cm3) on the basis of fluid-attenuated inversion-recovery (FLAIR) axial slices, as previously described.11 EOR was calculated as:
(preoperative tumor volume − postoperative tumor volume)/preoperative tumor volume
When present, the volume of enhancing tumor was also measured on preoperative imaging. Determination of volumes was made without knowledge of clinical outcome.
Frequency distributions and summary statistics were calculated for all variables. For categoric variables, cross-tabulations were generated, and Wilcoxon tests (for ordered variables) and χ2 tests (for nonordered variables) were used to compare their distributions. The Kaplan-Meier method12 was used to estimate OS, PFS, and MPFS.
Univariate OS, PFS, and MPFS analyses using proportional hazards models13 were used to assess the prognostic significance of multiple variables. Each measure of tumor burden was assessed separately for predictive value for OS, PFS, and MPFS after adjusting for the effects of age, Karnofsky performance score (KPS), tumor location, and tumor subtype. For outcome measures that were significantly predicted by more than one measure of tumor burden, forward step-wise proportional hazards modeling was performed to assess the relative and independent prognostic capacity of each parameter. All P values were from two-sided tests with statistical significance defined as P < .05. For analyses involving multiple categories, to minimize the problem of overstatement of statistical differences resulting from multiple comparisons, individual comparisons were considered for statistical significance only if the overall test was statistically significant. For analyses in which preoperative tumor volume was treated as a continuous variable, log transformation (base 10) was performed to help prevent inappropriate weighting of the results by tumors with large volumes.
In excess of 800 patients with LGGs were treated at UCSF between 1989 and 2005. Characteristics of the 216 patients who met inclusion criteria are shown in Table 1. Median time between symptom onset and time of surgical resection was 0.42 years (range, 0.01 to 3.57 years). Motor and speech mapping were performed in 154 (71%) and 75 (35%) cases, respectively. There was no operative mortality. New postoperative deficits were noted in 36 patients (17%); all but four subsequently had complete recovery (Table 2). There was no significant association between EOR and presence of new postoperative deficit (P = .36). The majority of surgical procedures (74%) were performed by a single surgeon (M.S.B.).
There were 34 deaths (16%), with a median follow-up for the remaining patients of 4.4 years. Of those still alive at last follow-up, 158 (87%) were followed for more than 2 years. Progression and malignant progression were identified in 95 (44%) and 44 (20%) cases, respectively. Median time to progression was 5.5 years (95% CI, 4.6 to 6.5 years), and median time to malignant progression was 10.1 years (95% CI, 8.8 to 11.4 years). Of the 44 cases of malignant progression, 42 (95%) demonstrated new contrast enhancement on follow-up imaging. The two cases of malignant progression that did not demonstrate contrast enhancement on follow-up imaging demonstrated rapid increases in FLAIR volume, prompting repeat surgery that demonstrated malignant progression. A total of 32 of the 44 cases with malignant progression underwent repeat resection, and in all cases a higher-grade tumor was identified. Of the 12 cases of malignant progression that were not biopsy proven, five received no external-beam radiation therapy (EBRT) before malignant progression. For the remaining cases, malignant progression occurred an average of 53 months (median, 51 months; range, 19 to 79 months) after EBRT, suggesting that these imaging changes were not a result of radiation effects.
Measures of tumor burden are summarized in Table 1. Notably, nearly one half had at least 90% resection of FLAIR abnormality. Figure 1 provides a graphical summary of preoperative tumor volume versus EOR.
Contrast enhancement on preoperative imaging was identified in 16 cases (7%; median volume, 0.28 cm3; range, 0.03 to 7.78 cm3), including 13 oligodendrogliomas, two mixed oligoastrocytomas, and one astrocytoma. In each case, there was complete resection of all contrast-enhancing tumor tissue. Areas of contrast enhancement were sent for pathology review, and no high-grade tumor was identified. Preoperative presence of contrast-enhancing tumor was not predictive of OS (P = .7) or MPFS (P = .5), but was predictive of shorter PFS (hazard ratio [HR] = 2.99; 95% CI, 1.41 to 6.35; P = .004). Of these 16 patients, only two had enhancement on recurrence; both of these patients underwent subsequent biopsy that showed higher-grade tumor.
Tumor involvement of eloquent tissue was significantly predictive of OS, PFS, and MPFS (Table 3). No patients with complete resection of all FLAIR abnormality received EBRT. Of the patients whose tumors were subtotally resected, those who underwent postoperative EBRT demonstrated longer PFS (P = .022) but shorter OS (P = .022). The latter observation may reflect an association between greater postoperative tumor burden and application of EBRT because subtotally resected tumors that were treated with postoperative EBRT had a mean EOR of 55%, whereas those not treated with postoperative EBRT had a mean EOR of 75% (P < .001).
After adjusting for the effects of age, KPS, tumor location, and tumor subtype, log preoperative tumor volume remained a significant predictor of OS (HR = 4.442; 95% CI, 1.601 to 12.320; P = .004), PFS (HR = 2.711; 95% CI, 1.590 to 4.623; P < .001), and MPFS (HR = 3.826; 95% CI, 1.632 to 8.969; P = .002; Table 4). For graphical display, patients were grouped based on preoperative tumor volume (Fig 2A).
After adjusting for the effects of age, KPS, tumor location, and tumor subtype, postoperative tumor volume remained a significant predictor of OS (HR 1.010; 95% CI, 1.001 to 1.019; P = .03) and PFS (HR = 1.007; 95% CI, 1.001 to 1.014; P = .035), but was not predictive of MPFS (P = .32; Table 4). Patients with complete resection of FLAIR abnormality had a significantly longer OS compared with patients having any residual FLAIR abnormality (HR = 0.094; 95% CI, 0.023 to 0.39; P = .001; Fig 2B). Subgroupings were created on the basis of postoperative tumor volume to specifically address the risk of relatively small volumes of residual tumor. Patients with residual FLAIR abnormality volume between 0.1 and 5.0 or between 5.1 and 15.0 cm3 demonstrated significantly shorter OS compared with patients who had complete resection of FLAIR abnormality (log-rank P < .001; three-group comparison; Fig 2C).
After adjusting for the effects of age, KPS, tumor location and tumor subtype, EOR remained a significant predictor of OS (HR = 0.972; 95% CI, 0.960 to 0.983; P < .001) and MPFS (HR = 0.983; 95% CI, 0.972 to 0.995; P = .005), and demonstrated a predictive trend for PFS (HR = 0.992; 95% CI, 0.984 to 1.001; P = .088; Table 4). EOR remained a statistically significant predictor of OS even when the set of patients analyzed was confined to those with EOR of at least 80% (HR = 0.894; 95% CI, 0.816 to 0.979; P = .016). For the purposes of display, patients were grouped by EOR (Fig 2D). As an example of differences in predicted outcome based on EOR, patients with at least 90% EOR had 5- and 8-year OS, PFS, and MPFS rates of 97% and 91%, 75% and 43%, and 93% and 76%, respectively, whereas patients with less than 90% EOR had 5- and 8-year OS, PFS, and MPFS rates of 76% and 60%, 40% and 21%, and 72% and 48%, respectively. Patients with complete resection of all FLAIR abnormality had 5- and 8-year OS, PFS, and MPFS rates of 98% and 98%, 78% and 48%, and 96% and 79%, respectively. As a second example of differences in predicted outcome related to EOR, the relative HR for 100% versus 50% resection is 0.24, based on the proportional hazards model.
We sought to address whether improvements in surgical techniques and supportive care from 1989 to 2005 may have changed sufficiently to significantly affect both the EOR attempted and postoperative survival. Data were separately analyzed for patients operated between 1989 and 1998. Regression analysis demonstrated a significant association between EOR and survival (HR = 0.973; 95% CI, 0.955 to 0.992; P = .005). Thus, the association between EOR and outcome does not simply reflect improvements in technology.
Because each outcome measure was significantly predicted by at least two measures of tumor burden (Table 4), we sought to assess the relative contributions of the tumor burden measures. Multivariate modeling was performed for each outcome measure using the following parameters: age, KPS, tumor location, involvement of eloquent brain regions, tumor subtype, log preoperative tumor volume, postoperative tumor volume, and EOR. Of the three measures of tumor burden, forward stepwise modeling selected only EOR (HR = 0.982; 95% CI, 0.971 to 0.994; P = .002) as predictive of OS, log preoperative volume (HR = 2.074; 95% CI, 1.243 to 3.462; P = .005) as predictive of PFS, and log preoperative volume (HR = 3.567; 95% CI, 1.754 to 7.257; P < .001) as predictive of MPFS. For OS, tumor subtype (P = .027) and involvement of eloquent brain regions (HR = 4.27; 95% CI, 1.17 to 15.56; P = .028) also entered the model. For PFS, involvement of eloquent brain regions (HR = 1.92; 95% CI, 1.17 to 3.14; P = .01) also entered the model. The only parameter selected for MPFS was log preoperative tumor volume.
The optimal role of surgery for low-grade infiltrative glioma patients remains a major controversy.2-7 We provide clear evidence that a more aggressive resection does predict significant improvement in OS compared with a simple debulking procedure. Furthermore, a significantly improved OS is predicted, with pushing the EOR near the limits of complete resection. Predicted OS is negatively influenced by even residual tumor volumes on the order of 10 cm3 (comparable to a cube with 2.15-cm dimensions; Fig 2C). Although previous studies have suggested an association between greater resection and improved outcome, none has offered the resolution to demonstrate the benefit of resecting the last few cubic centimeters, a step that may require the use of sophisticated mapping techniques. Collectively, these data argue strongly in favor of achieving a maximal resection of LGGs.
We encountered remarkably minimal morbidity and no mortality. Only two new permanent deficits were identified, and the outcomes of two additional new deficits could not be verified. Assuming these latter two deficits were permanent, our incidence of new permanent neurological deficits is less than 2%, and none of these deficits was profound. A subset of patients experienced temporary neurological deficits that improved back to baseline, which we attribute to edema adjacent to the resection cavity.
Previous reports have suggested that larger preoperative tumor volume is a negative prognostic indicator in LGG patients and have suggested that the biology of smaller lesions may differ from that of larger lesions.14-16 In the present analysis, greater preoperative tumor volume was significantly associated with shorter OS, PFS, and MPFS. This suggests that tumors that are larger at presentation may have an inherently faster growth rate, and thus recur faster in the setting of a gross total resection or continue to grow faster in the setting of a subtotal resection. This increased growth rate may in turn account for the shorter OS and MPFS in patients with larger preoperative tumor volume.
Keles et al5 reviewed the literature regarding the role of EOR in LGGs, identifying 30 studies, including European Organisation for Research and Treatment of Cancer (EORTC)17 and US Cooperative Group18 prospective studies. All but five19-23 were excluded from further review because of significant methodologic limitations. Although these five studies suggested that more extensive surgical resection may favor longer survival, methodologic concerns in each of these studies limit the applicability of the results.5 Most notably, none of the studies included volumetric analysis, and four relied solely on the surgeons’ intraoperative impression of EOR. Seven subsequent reports provide assessment of the role of EOR in LGGs, with a subset suggesting improved patient outcome with greater resection16,24,25 and the remaining suggesting no association.26-29 These additional studies also have significant limitations similar to those described by Keles et al.5 For example, only one reports provides volumetric analysis, and in that report, the analyzed population is only 42 patients.16
Inconsistent classifications of EOR in previous reports complicate detailed comparisons of patient outcome with the present series. However, in three of the five reports on EOR in LGGs that met basic design criteria in a review by Keles et al,5 OS rates are provided separately for cases with gross total resection. The 5- and 8-year OS rates in the present series (98% and 98%, respectively) compare favorably with the 5- and 10-year rates of Rajan et al22 (90% and 68%, respectively), the 5- and 10-year rates of Soffietti et al 21 (51% and 11%, respectively), and the 5-year rate of Phillippon et al 20 (80%).
The present study was designed after carefully considering the limitations of previous studies. Most importantly, our methodology includes volumetric assessment of preoperative and postoperative tumor volumes, rather than relying on nonquantitative estimates or estimates based solely on the intraoperative interpretation of the surgeon. Nonetheless, our study also has limitations. The retrospective design did not permit us to standardize patient follow-up. However, sensitivity analyses suggest that this did not affect our overall conclusions. In addition, there were relatively few deaths in our series, which limited extensive multivariate analyses. The present study also does not include assessment of neurocognitive outcomes30; it will be important to collect this data in the future. Furthermore, the present study does not offer direct comparison between patients undergoing biopsy only and patients undergoing maximal surgical resection because the goal for the patients included in our series was to do the most complete resection possible.
The end points PFS and MPFS have several potential limitations. New contrast enhancement can appear after surgery, EBRT, or intraparenchymal drug therapies. Postsurgical enhancement has been well characterized and has not been reported to resolve and subsequently recur.11 In the present series, for a diagnosis of malignant progression to be made on the basis of imaging alone, there had to be evidence of new enhancement, and the earliest this occurred was 19 months after initial surgery. In some cases, it can be difficult to distinguish radiation effects from progression. However, in the present series, the seven patients diagnosed with malignant progression on the basis of imaging alone and who had had prior EBRT were far enough out from EBRT for imaging effects to have been expected to have resolved. None of the patients who demonstrated malignant progression had been treated with intraparenchymal drug therapy.
Improved outcome among adult patients with hemispheric LGG is predicted by greater EOR, even at the limits near complete resection. When used with functional mapping techniques, increased EOR is not associated with additional morbidity. We recommend that, when feasible, management of adult hemispheric LGG include maximal surgical resection to improve long-term patient outcomes.
The author(s) indicated no potential conflicts of interest.
Conception and design: Justin S. Smith, Edward F. Chang, Kathleen R. Lamborn, Susan M. Chang, Michael D. Prados, Soonmee Cha, Tarik Tihan, Scott VandenBerg, Michael W. McDermott, Mitchel S. Berger
Administrative support: Mitchel S. Berger
Provision of study materials or patients: Susan M. Chang, Michael D. Prados, Soonmee Cha, Tarik Tihan, Scott VandenBerg, Mitchel S. Berger
Collection and assembly of data: Justin S. Smith, Edward F. Chang, Tarik Tihan, Scott VandenBerg, Mitchel S. Berger
Data analysis and interpretation: Justin S. Smith, Edward F. Chang, Kathleen R. Lamborn, Susan M. Chang, Michael D. Prados, Soonmee Cha, Tarik Tihan, Scott VandenBerg, Michael W. McDermott, Mitchel S. Berger
Manuscript writing: Justin S. Smith, Edward F. Chang, Kathleen R. Lamborn, Susan M. Chang, Michael D. Prados, Soonmee Cha, Tarik Tihan, Scott VandenBerg, Michael W. McDermott, Mitchel S. Berger
Final approval of manuscript: Justin S. Smith, Edward F. Chang, Kathleen R. Lamborn, Susan M. Chang, Michael D. Prados, Soonmee Cha, Tarik Tihan, Scott VandenBerg, Michael W. McDermott, Mitchel S. Berger
Fig 1. Graphical summary of preoperative tumor volume (median, 36.6 cm3; range, 0.7 to 246.1 cm3) versus extent of tumor resection (median, 88.0%; range 5.0% to 100%) for 216 adult low-grade hemispheric gliomas. Although there was a tendency toward higher extent of resection for smaller tumors, there were a substantial number of large tumors for which complete, or nearly complete, resection was accomplished.
Fig 2. Associations between tumor burden and patient outcome. (A) Patients with larger preoperative tumor volumes have significantly shorter progression-free survival (Cox proportional hazards model based on log transformation of preoperative tumor volume; P < .001; HR = 2.711; 95% CI, 1.590 to 4.623). On the basis of the distribution of preoperative tumor volumes, cases were grouped for the purposes of visual display. The numbers of patients and events in the less than 25, 25 to 50 and more than 50 cm3 categories were 74 and 25, 59 and 24, and 83 and 46, respectively. (B) Patients with complete resection of fluid-attenuated inversion-recovery (FLAIR) abnormality (75 patients, two events) had a significantly longer overall survival compared with patients having any residual FLAIR abnormality (141 patients, 32 events; HR = 0.094; 95% CI, 0.023 to 0.39; P = .001). (C) Patients with even small volumes of residual FLAIR abnormality demonstrated shorter overall survival compared with patients with no residual FLAIR abnormality (Cox proportional hazards model only including patients with 15 cm3 or less of residual FLAIR abnormality; P = .001; HR = 1.166; 95% CI, 1.068 to 1.274). For the purposes of visual display, cases were grouped based on residual FLAIR abnormality: 0 cm3 (74 patients, two events), 0.1 to 5.0 (37 patients, three events), and 5.1 to 15.0 cm3 (38 patients, eight events). (D) Patients with a greater percentage of tumor resection had a significantly longer overall survival (Cox proportional hazards model P < .001; HR = 0.972; 95% CI, 0.960 to 0.983). For the purposes of visual display, cases were grouped by extent of tumor resection: 100% (75 patient, two events), 90% to 99% (26 patients, three events), 41% to 89% (94 patients, 18 events), and less than 40% (21 patients, 11 events). HR, hazard ratio.
|
| Characteristic | No. of Patients | % | |
|---|---|---|---|
| Sex | |||
| Male | 134 | 62 | |
| Female | 82 | 38 | |
| Age at diagnosis, years | |||
| < 40 | 119 | 55 | |
| 40-60 | 92 | 43 | |
| > 60 | 5 | 2 | |
| Median | 38 | ||
| Range | 19-72 | ||
| KPS at diagnosis | |||
| 80 | 7 | 3 | |
| 90 | 194 | 90 | |
| 100 | 15 | 7 | |
| Median | 90 | ||
| Range | 80-100 | ||
| Presentation | |||
| Seizure | 179 | 83 | |
| Incidental | 13 | 6 | |
| Headache | 11 | 5 | |
| Miscellaneous neurologic complaints | 13 | 6 | |
| Side of tumor | |||
| Left | 117 | 54 | |
| Right | 99 | 46 | |
| Tumor location | |||
| Frontal lobe | 112 | 52 | |
| Insula | 56 | 26 | |
| Temporal lobe | 30 | 14 | |
| Parietal lobe | 18 | 8 | |
| Eloquent tumor location* | 125 | 58 | |
| Tumor subtype | |||
| Astrocytoma | 93 | 43 | |
| Oligodendroglioma | 91 | 42 | |
| Mixed oligoastrocytoma | 32 | 15 | |
| Postoperative EBRT† | |||
| No | 83 | 58 | |
| Yes | 45 | 32 | |
| Unknown | 14 | 10 | |
| Postoperative chemotherapy† | |||
| No | 78 | 55 | |
| Yes | 51 | 36 | |
| Unknown | 13 | 9 | |
| Pre-operative tumor volume, cm3 | |||
| < 25.0 | 74 | 34 | |
| 25-50 | 59 | 27 | |
| 51-100 | 51 | 24 | |
| 101-250 | 32 | 15 | |
| Median | 36.6 | ||
| Range | 0.7-246.1 | ||
| Postoperative tumor volume, cm3 | |||
| 0 | 75 | 35 | |
| 0.1-10.0 | 58 | 27 | |
| 10.1-50.0 | 65 | 30 | |
| 50.1-200.0 | 18 | 8 | |
| Median | 3.7 | ||
| Range | 0.0-197.8 | ||
| Extent of tumor resection (%) | |||
| 0-40 | 21 | 10 | |
| 41-69 | 39 | 18 | |
| 70-89 | 55 | 25 | |
| 90-99 | 26 | 12 | |
| 100 | 75 | 35 | |
| Median | 88.0 | ||
| Range | 5.0-100.0 | ||
Abbreviations: KPS, Karnofsky performance score; EBRT, external-beam radiation therapy.
*Tumor involved one or more of the following: internal capsule, basal ganglia, language cortex, sensory cortex, motor cortex, thalamus, and hypothalamus.
†Includes only patients with subtotal tumor resection.
|
| Condition | Total No. of Patients | % | Outcome | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Complete Recovery | Partial Recovery | No Recovery | Unknown | ||||||||||||||
| No. | % | No. | % | No. | % | No. | % | ||||||||||
| No new deficits | 180 | 83 | |||||||||||||||
| Speech deficits | |||||||||||||||||
| Expressive aphasia | 10 | 5 | 10 | 100 | |||||||||||||
| Dysarthria | 2 | 1 | 2 | 100 | |||||||||||||
| Motor deficits | |||||||||||||||||
| SMA syndrome | 8 | 4 | 8 | 100 | |||||||||||||
| Facial droop | 8 | 4 | 8 | 100 | |||||||||||||
| Hemiparesis | 2 | 1 | 2 | 100 | |||||||||||||
| Hand clumsiness/weakness | 5 | 2 | 5 | 100 | |||||||||||||
| Sensory deficits | |||||||||||||||||
| Upper and/or lower extremity | 2 | 1 | 2 | 100 | |||||||||||||
| Decreased proprioception (isolated foot) | 1 | 0.5 | 1 | 100 | |||||||||||||
| Partial Gerstmann syndrome | 3 | 1 | 2 | 67 | 1 | 33 | |||||||||||
| Quadrantanopsia | 1 | 0.5 | 1 | 100 | |||||||||||||
| Post-operative EDH* | 1 | 0.5 | 1 | 100 | |||||||||||||
Abbreviations: SMA,supplementary motor area; EDH, epidural hematoma.
*Patient developed new mild to moderate hemiparesis that improved almost back to baseline following evacuation.
|
| Assessment | Overall Survival | Progression-Free Survival | Malignant Progression-Free Survival |
|---|---|---|---|
| Age at diagnosis | |||
| HR | 1.002 | 0.990 | 0.974 |
| 95% CI | 0.966 to 1.040 | 0.966 to 1.014 | 0.940 to 1.010 |
| P | .91 | .41 | .15 |
| Sex, male v female | |||
| HR | 0.66 | 0.91 | 1.21 |
| 95% CI | 0.31 to 1.37 | 0.60 to 1.37 | 0.67 to 2.19 |
| P | .26 | .65 | .53 |
| KPS at diagnosis | |||
| HR | 0.88 | 0.95 | 0.88 |
| 95% CI | 0.76 to 1.01 | 0.87 to 1.03 | 0.78 to 1.00 |
| P | .073 | .19 | .058 |
| Tumor side, left v right | |||
| HR | 0.53 | 0.85 | 0.73 |
| 95% CI | 0.27 to 1.05 | 0.57 to 1.28 | 0.40 to 1.32 |
| P | .071 | .44 | .29 |
| Tumor location* | |||
| Overall P | .089 | .078 | .22 |
| Temporal lobe | |||
| HR | 0.79 | 1.32 | 1.34 |
| 95% CI | 0.18 to 3.47 | 0.66 to 2.63 | 0.49 to 3.64 |
| P | .071 | .43 | .57 |
| Parietal lobe | |||
| HR | 2.44 | 1.32 | 2.35 |
| 95% CI | 0.69 to 8.62 | 0.59 to 2.95 | 0.78 to 7.07 |
| P | .17 | .50 | .13 |
| Insula | |||
| HR | 2.34 | 1.86 | 1.87 |
| 95% CI | 1.12 to 4.90 | 1.17 to 2.95 | 0.95 to 3.70 |
| P | .024 | .009 | .072 |
| Eloquence, yes v no† | |||
| HR | 8.59 | 2.57 | 2.60 |
| 95% CI | 2.62 to 28.14 | 1.64 to 4.04 | 1.31 to 5.16 |
| P | < .001 | < .001 | .006 |
| Tumor subtype‡ | |||
| Overall P | 0.071 | 0.44 | 0.044 |
| Oligodendroglioma | |||
| HR | 0.27 | 0.77 | 0.52 |
| 95% CI | 0.11 to 0.67 | 0.49 to 1.20 | 0.27 to 1.01 |
| P | .005 | .25 | .054 |
| Mixed OA | |||
| HR | 0.72 | 0.78 | 0.35 |
| 95% CI | 0.29 to 1.80 | 0.43 to 1.42 | 0.12 to 1.00 |
| P | .48 | .41 | .050 |
| Postoperative EBRT, yes v no§ | |||
| HR | 2.80 | 0.54 | 0.86 |
| 95% CI | 1.16 to 6.77 | 0.32 to 0.91 | 0.43 to 1.72 |
| P | .022 | .022 | .67 |
| Postoperative chemotherapy, yes v no§ | |||
| HR | 0.86 | 0.77 | 1.26 |
| 95% CI | 0.33 to 2.23 | 0.45 to 1.31 | 0.59 to 2.69 |
| P | .76 | .33 | .55 |
NOTE. Boldfacing represents statistical significance.P values are from two-sided tests (Cox regression) and were statistically significant when < .05.
Abbreviations: HR, hazard ratio; KPS, Karnofsky performance score; OA, oligoastrocytoma; EBRT, external-beam radiation therapy.
*Relative to tumors with frontal lobe location.
†Tumor involved eloquent region (precentral gyrus, postcentral gyrus, speech cortex, visual cortex, hypothalamus, thalamus, internal capsule, and/or basal ganglia).
‡Relative to tumors with pure astrocytoma histology.
§Analyses include only patients with subtotal tumor resection.
|
| Measure | Overall Survival | Progression-Free Survival | Malignant Progression–Free Survival | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| P | Hazard Ratio | P | Hazard Ratio | P | Hazard Ratio | |||||||
| Point Estimate | 95% CI | Point Estimate | 95% CI | Point Estimate | 95% CI | |||||||
| Log preoperative tumor volume, cm3 | .004 | 4.442 | 1.601 to 12.320 | < .001 | 2.711 | 1.590 to 4.623 | .002 | 3.826 | 1.632 to 8.969 | |||
| Postoperative tumor volume, cm3 | .03 | 1.010 | 1.001 to 1.019 | .035 | 1.007 | 1.001 to 1.014 | .32 | 1.005 | 0.996 to 1.014 | |||
| Extent of resection | < .001 | 0.972 | 0.960 to 0.983 | .088 | 0.992 | 0.984 to 1.001 | .005 | 0.983 | 0.972 to 0.995 | |||
NOTE. For each measure of tumor burden, a forward Cox proportional hazards model was generated for each outcome measure. Each model included correction for the effects of patient age (continuous variable), Karnofsky performance score (continuous variable), tumor location (frontal, temporal, parietal, insula), and tumor subtype (astrocytoma, oligodendroglioma, mixed oligoastrocytoma). Boldfacing indicates statistical significance.
Authors’ disclosures of potential conflicts of interest and author contributions are found at the end of this article.
We thank the patients and families who generously provided follow-up information beyond that available in our records. We also thank our clinical colleagues and support staff who helped us to acquire this clinical experience.
| 1. | Chang EF, Keles GE, Potts MB, et al: Seizure characteristics and control following surgical resection in 332 patients with low-grade glioma. J Neurosurg (in press) Google Scholar |
| 2. | Berger MS, Rostomily RC: Low grade gliomas: Functional mapping resection strategies, extent of resection, and outcome. J Neurooncol 34::85,1997-101, Crossref, Medline, Google Scholar |
| 3. | Lang FF, Gilbert MR: Diffusely infiltrative low-grade gliomas in adults. J Clin Oncol 24::1236,2006-1245, Link, Google Scholar |
| 4. | Cavaliere R, Lopes MB, Schiff D: Low-grade gliomas: An update on pathology and therapy. Lancet Neurol 4::760,2005-770, Crossref, Medline, Google Scholar |
| 5. | Keles GE, Lamborn KR, Berger MS: Low-grade hemispheric gliomas in adults: A critical review of extent of resection as a factor influencing outcome. J Neurosurg 95::735,2001-745, Crossref, Medline, Google Scholar |
| 6. | Piepmeier J, Baehring JM: Surgical resection for patients with benign primary brain tumors and low grade gliomas. J Neurooncol 69::55,2004-65, Crossref, Medline, Google Scholar |
| 7. | Grier JT, Batchelor T: Low-grade gliomas in adults. Oncologist 11::681,2006-693, Crossref, Medline, Google Scholar |
| 8. | Low-Grade Glioma Guidelines Team in association with the Guidelines and Outcomes Committee of the American Association of Neurological Surgeons: Practice parameters in adults with suspected or known supratentorial nonoptic pathway low-grade glioma. Neurosurg Focus 4::e10,1998, Google Scholar |
| 9. | Tihan T, Vohra P, Berger MS, et al: Definition and diagnostic implications of gemistocytic astrocytomas: A pathological perspective. J Neurooncol 76::175,2006-183, Crossref, Medline, Google Scholar |
| 10. | Kleihues P, Louis DN, Scheithauer BW, et al: The WHO classification of tumors of the nervous system. J Neuropathol Exp Neurol 61::215,2002-229, Crossref, Medline, Google Scholar |
| 11. | Smith JS, Cha S, Mayo MC, et al: Serial diffusion-weighted magnetic resonance imaging in cases of glioma: Distinguishing tumor recurrence from postresection injury. J Neurosurg 103::428,2005-438, Crossref, Medline, Google Scholar |
| 12. | Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53::457,1958-481, Crossref, Google Scholar |
| 13. | Cox DR: Regression models and life tables. J R Stat Soc 34::187,1972-220, Google Scholar |
| 14. | Berger MS, Deliganis AV, Dobbins J, et al: The effect of extent of resection on recurrence in patients with low grade cerebral hemisphere gliomas. Cancer 74::1784,1994-1791, Crossref, Medline, Google Scholar |
| 15. | Kreth FW, Faist M, Rossner R, et al: Supratentorial World Health Organization Grade 2 astrocytomas and oligoastrocytomas: A new pattern of prognostic factors. Cancer 79::370,1997-379, Crossref, Medline, Google Scholar |
| 16. | Mariani L, Siegenthaler P, Guzman R, et al: The impact of tumour volume and surgery on the outcome of adults with supratentorial WHO grade II astrocytomas and oligoastrocytomas. Acta Neurochir (Wien) 146::441,2004-448, Crossref, Medline, Google Scholar |
| 17. | Karim AB, Maat B, Hatlevoll R, et al: A randomized trial on dose-response in radiation therapy of low-grade cerebral glioma: European Organization for Research and Treatment of Cancer (EORTC) study 22844. Int J Radiat Oncol Biol Phys 36::549,1996-556, Crossref, Medline, Google Scholar |
| 18. | Eyre HJ, Crowley JJ, Townsend JJ, et al: A randomized trial of radiotherapy versus radiotherapy plus CCNU for incompletely resected low-grade gliomas: A Southwest Oncology Group study. J Neurosurg 78::909,1993-914, Crossref, Medline, Google Scholar |
| 19. | Leighton C, Fisher B, Bauman G, et al: Supratentorial low-grade glioma in adults: An analysis of prognostic factors and timing of radiation. J Clin Oncol 15::1294,1997-1301, Link, Google Scholar |
| 20. | Philippon JH, Clemenceau SH, Fauchon FH, et al: Supratentorial low-grade astrocytomas in adults. Neurosurgery 32::554,1993-559, Crossref, Medline, Google Scholar |
| 21. | Soffietti R, Chio A, Giordana MT, et al: Prognostic factors in well-differentiated cerebral astrocytomas in the adult. Neurosurgery 24::686,1989-692, Crossref, Medline, Google Scholar |
| 22. | Rajan B, Pickuth D, Ashley S, et al: The management of histologically unverified presumed cerebral gliomas with radiotherapy. Int J Radiat Oncol Biol Phys 28::405,1994-413, Crossref, Medline, Google Scholar |
| 23. | van Veelen ML, Avezaat CJ, Kros JM, et al: Supratentorial low grade astrocytoma: Prognostic factors, dedifferentiation, and the issue of early versus late surgery. J Neurol Neurosurg Psychiatry 64::581,1998-587, Crossref, Medline, Google Scholar |
| 24. | Claus EB, Horlacher A, Hsu L, et al: Survival rates in patients with low-grade glioma after intraoperative magnetic resonance image guidance. Cancer 103::1227,2005-1233, Crossref, Medline, Google Scholar |
| 25. | Duffau H, Lopes M, Arthuis F, et al: Contribution of intraoperative electrical stimulations in surgery of low grade gliomas: A comparative study between two series without (1985-96) and with (1996-2003) functional mapping in the same institution. J Neurol Neurosurg Psychiatry 76::845,2005-851, Crossref, Medline, Google Scholar |
| 26. | Shaw E, Arusell R, Scheithauer B, et al: Prospective randomized trial of low- versus high-dose radiation therapy in adults with supratentorial low-grade glioma: Initial report of a North Central Cancer Treatment Group/Radiation Therapy Oncology Group/Eastern Cooperative Oncology Group study. J Clin Oncol 20::2267,2002-2276, Link, Google Scholar |
| 27. | Lebrun C, Fontaine D, Ramaioli A, et al: Long-term outcome of oligodendrogliomas. Neurology 62::1783,2004-1787, Crossref, Medline, Google Scholar |
| 28. | Olson JD, Riedel E, DeAngelis LM: Long-term outcome of low-grade oligodendroglioma and mixed glioma. Neurology 54::1442,2000-1448, Crossref, Medline, Google Scholar |
| 29. | Pignatti F, van den Bent M, Curran D, et al: Prognostic factors for survival in adult patients with cerebral low-grade glioma. J Clin Oncol 20::2076,2002-2084, Link, Google Scholar |
| 30. | Meyers CA, Brown PD: Role and relevance of neurocognitive assessment in clinical trials of patients with CNS tumors. J Clin Oncol 24::1305,2006-1309, Link, Google Scholar |
