Irradiated volume as a predictor of brain radionecrosis after linear accelerator stereotactic radiosurgery

Significance of target location relative to the depth from the brain surface and high-dose irradiated volume in the development of brain radionecrosis after micromultileaf collimator-based stereotactic radiosurgery for brain metastases

The objective of this study was to investigate the factors that potentially lead to brain radionecrosis (RN) after micromultileaf collimator-based stereotactic radiosurgery (SRS) for brain metastases. We retrospectively evaluated 131 lesions with a minimum follow-up of 6 months, 43.5% of which received prior whole-brain radiotherapy (WBRT). The three-tiered location grade (LG) was defined, as follows, for each target by considering mainly the depth from the brain surface: grade 1 (superficial), involving the region at a depth of ≤5 mm from the brain surface; grade 2 (deep), located at a depth of >5 mm from the brain surface; and grade 3 (central), located in the brainstem, cerebellar peduncle, diencephalon, or basal ganglion. The predictive factors for RN, including high-dose irradiated isodose volumes (IIDVs) and LG, were evaluated by univariate and multivariate analysis. Symptomatic RN (S-RN) and asymptomatic RN (A-RN) were observed in 8.4% and 6.9% of cases, respectively. Multivariate analysis indicated that the significant factors for both types of RN were LG, V12 Gy, and V22 Gy in all cases; V22 Gy and LG for the non-WBRT cases; and V15 Gy and LG for the WBRT cases. For the non-WBRT cases, the cutoff values of V22 Gy were 2.62 and 2.14 cm(3) for S-RN and both RN, respectively. For the WBRT cases, the cutoff values of V15 Gy were 5.61 and 5.20 cm(3) for S-RN and both RN, respectively. In addition to the IIDV data, LG helps predict the risk of RN. High-dose IIDV, V22 Gy, was also significantly correlated with RN, particularly for patients treated with SRS alone.

Integrated boost IMRT with FET-PET-adapted local dose escalation in glioblastomas. Results of a prospective phase II study

PURPOSE

Dose escalations above 60 Gy based on MRI have not led to prognostic benefits in glioblastoma patients yet. With positron emission tomography (PET) using [(18)F]fluorethyl-L-tyrosine (FET), tumor coverage can be optimized with the option of regional dose escalation in the area of viable tumor tissue.

METHODS AND MATERIALS

In a prospective phase II study (January 2008 to December 2009), 22 patients (median age 55 years) received radiochemotherapy after surgery. The radiotherapy was performed as an MRI and FET-PET-based integrated-boost intensity-modulated radiotherapy (IMRT). The prescribed dose was 72 and 60 Gy (single dose 2.4 and 2.0 Gy, respectively) for the FET-PET- and MR-based PTV-FET((72 Gy)) and PTV-MR((60 Gy)). FET-PET and MRI were performed routinely for follow-up. Quality of life and cognitive aspects were recorded by the EORTC-QLQ-C30/QLQ Brain20 and Mini-Mental Status Examination (MMSE), while the therapy-related toxicity was recorded using the CTC3.0 and RTOG scores.

RESULTS

Median overall survival (OS) and disease-free survival (DFS) were 14.8 and 7.8 months, respectively. All local relapses were detected at least partly within the 95% dose volume of PTV-MR((60 Gy)). No relevant radiotherapy-related side effects were observed (excepted alopecia). In 2 patients, a pseudoprogression was observed in the MRI. Tumor progression could be excluded by FET-PET and was confirmed in further MRI and FET-PET imaging. No significant changes were observed in MMSE scores and in the EORTC QLQ-C30/QLQ-Brain20 questionnaires.

CONCLUSION

Our dose escalation concept with a total dose of 72 Gy, based on FET-PET, did not lead to a survival benefit. Acute and late toxicity were not increased, compared with historical controls and published dose-escalation studies.

Radiation necrosis following treatment of high grade glioma--a review of the literature and current understanding

Radiation therapy is an integral part of the standard treatment paradigm for malignant gliomas, with proven efficacy in randomized control trials. Radiation treatment is not without risk however, and radiation injury occurs in a certain proportion of patients. Difficulties in differentiating recurrence from radiation injury complicate the treatment course and can compromise care. These complexities are compounded by the recent distinction of two types of radiation injury: pseudoprogression and radiation necrosis, which are likely the result of radiation injury to the tumor and normal tissue, respectively. A thorough understanding of radiation-induced injury offers insights to guide further therapies. We detail the current knowledge of the mechanisms of radiation injury, along with potential targets for therapeutic intervention. Various diagnostic modalities are also described, in addition to the multiple options for treatment within the context of their pathophysiology and clinical efficacy. Radiation therapy is an integral part of the multidisciplinary management of gliomas, and the optimal diagnosis and management of radiation injury is paramount to improving patient outcomes.

Plan quality and treatment planning technique for single isocenter cranial radiosurgery with volumetric modulated arc therapy

PURPOSE

To demonstrate plan quality and provide a practical, systematic approach to the treatment planning technique for single isocenter cranial radiosurgery with volumetric modulated arc therapy (VMAT; RapidArc, Varian Medical systems, Palo Alto, CA).

METHODS AND MATERIALS

Fifteen patients with 1 or more brain metastases underwent single isocenter VMAT radiosurgery. All plans were normalized to deliver 100% of the prescription dose to 99%-100% of the target volume. All targets per plan were treated to the same dose. Plans were created with dose control tuning structures surrounding targets to maximize conformity and dose gradient. Plan quality was evaluated by calculation of conformity index (CI = 100% isodose volume/target volume) and homogeneity index (HI = maximum dose/prescription dose) scores for each target and a Paddick gradient index (GI = 50% isodose volume/100% isodose volume) score for each plan.

RESULTS

The median number of targets per patient was 2 (range, 1-5). The median number of non-coplanar arcs utilized per plan was 2 (range, 1- 4). Single target plans were created with 1 or 2 non-coplanar arcs while multitarget plans utilized 2 to 4 non-coplanar arcs. Prescription doses ranged from 5-16 Gy in 1-5 fractions. The mean conformity index was 1.12 (± SD, 0.13) and the mean HI was 1.44 (± SD, 0.11) for all targets. The mean GI per plan was 3.34 (± SD, 0.42).

CONCLUSIONS

We have outlined a practical approach to cranial radiosurgery treatment planning using the single isocenter VMAT platform. One or 2 arc single isocenter plans are often adequate for treatment of single targets, while 2-4 arcs may be more advantageous for multiple targets. Given the high plan quality and extreme clinical efficiency, this single isocenter VMAT approach will continue to become more prevalent for linac-based radiosurgical treatment of 1 or more intracranial targets and will likely replace multiple isocenter techniques.

Evaluation of volumetric modulated arc therapy for cranial radiosurgery using multiple noncoplanar arcs

PURPOSE

To evaluate a commercial volumetric modulated arc therapy (VMAT), using multiple noncoplanar arcs, for linac-based cranial radiosurgery, as well as evaluate the combined accuracy of the VMAT dose calculations and delivery.

METHODS

Twelve patients with cranial lesions of variable size (0.1-29 cc) and two multiple metastases patients were planned (Eclipse RapidArc AAA algorithm, v8.6.15) using VMAT (1-6 noncoplanar arcs), dynamic conformal arc (DCA, ∼4 arcs), and IMRT (nine static fields). All plans were evaluated according to a conformity index (CI), healthy brain tissue doses and volumes, and the dose to organs at risk. A 2D dose distribution was measured (Varian Novalis Tx, HD120 MLC, 1000 MU/min, 6 MV beam) for the ∼4 arc VMAT treatment plans using calibrated film dosimetry.

RESULTS

The CI (0-1 best) average for all plans was best for ∼4 noncoplanar arc VMAT at 0.86 compared with ∼0.78 for IMRT and a single arc VMAT and 0.68 for DCA. The volumes of healthy brain receiving 50% of the prescribed target coverage dose or more (V(50%)) were lowest for the four arc VMAT [RA(4)] and DCA plans. The average ratio of the V(50%) for the other plans to the RA(4) V(50%) were 1.9 for a single noncoplanar arc VMAT [RA(1nc)], 1.4 for single full coplanar arc VMAT [RA(1f)] and 1.3 for IMRT. The V(50%) improved significantly for single isocenter multiple metastases plan when two noncoplanar VMAT arcs were added to a full single coplanar one. The maximum dose to 5 cc of the outer 1 cm rim of healthy brain which one may want to keep below nonconsequential doses of 300-400 cGy, was 2-3 times greater for IMRT, RA(1nc) and RA(1f) plans compared with the multiple noncoplanar arc DCA and RA(4) techniques. Organs at risk near (0-4 mm) to targets were best spared by (i) single noncoplanar arcs when the targets are lateral to the organ at risk and (ii) by skewed nonvertical planes of IMRT fields when the targets are not lateral to the organ at risk. The highest dose gradient observed between an organ at risk and a target at the edge of a VMAT arc plane or plane of IMRT fields was 17%/mm. The average absolute percent difference between the measured and calculated central axis dose for all the VMAT plans was 3.6 ± 2.2%. The measured perpendicular profile widths and shifts were on average within 0.5 mm of planned values. The average total MUs for VMAT plans was double the DCA average and similar to the IMRT average.

CONCLUSIONS

For the aforementioned planning and delivery system and cranial lesions greater than 7 mm in diameter, multiple noncoplanar arc VMAT consistently provides accurate and high quality cranial radiosurgery dose distributions with low doses to healthy brain tissue and high dose conformity to the target. These qualities may make multiple noncoplanar arc VMAT suitable for a greater range of prescription doses or larger and more irregular lesions. For smaller and/or rounder lesions there are other clinically acceptable treatment techniques that may involve fewer couch angles or arcs and reduce treatment times.

Frameless linac-based stereotactic radiosurgery (SRS) for brain metastases: analysis of patient repositioning using a mask fixation system and clinical outcomes

Purpose

To assess the accuracy of patient repositioning and clinical outcomes of frameless stereotactic radiosurgery (SRS) for brain metastases using a stereotactic mask fixation system.

Patients and Methods

One hundred two patients treated consecutively with frameless SRS as primary treatment at University of Rome Sapienza Sant'Andrea Hospital between October 2008 and April 2010 and followed prospectively were involved in the study. A commercial stereotactic mask fixation system (BrainLab) was used for patient immobilization. A computerized tomography (CT) scan obtained immediately before SRS was used to evaluate the accuracy of patient repositioning in the mask by comparing the isocenter position to the isocenter position established in the planning CT. Deviations of isocenter coordinates in each direction and 3D displacement were calculated. Overall survival, brain control, and local control were estimated using the Kaplan-Meier method calculated from the time of SRS.

Results

The mean measured isocenter displacements were 0.12 mm (SD 0.35 mm) in the lateral direction, 0.2 mm (SD 0.4 mm) in the anteroposterior, and 0.4 mm (SD 0.6 mm) in craniocaudal direction. The maximum displacement of 2.1 mm was seen in craniocaudal direction. The mean 3D displacement was 0.5 mm (SD 0.7 mm), being maximum 2.9 mm. The median survival was 15.5 months, and 1-year and 2-year survival rates were 58% and 24%, respectively. Nine patients recurred locally after SRS, with 1-year and 2-year local control rates of 91% and 82%, respectively. Stable extracranial disease (P = 0.001) and KPS > 70 (P = 0.01) were independent predictors of survival.

Conclusions

Frameless SRS is an effective treatment in the management of patients with brain metastases. The presented non-invasive mask-based fixation stereotactic system is associated with a high degree of patient repositioning accuracy; however, a careful evaluation is essential since occasional errors up to 3 mm may occur.

Proposal for a new prognostic score for linac-based radiosurgery in cerebral arteriovenous malformations

PURPOSE

We evaluate patient-, angioma-, and treatment-specific factors for successful obliteration of cerebral arteriovenous malformations (AVM) to develop a new appropriate score to predict patient outcome after linac-based radiosurgery (RS).

METHODS AND MATERIALS

This analysis in based on 293 patients with cerebral AVM. Mean age at treatment was 38.8 years (4-73 years). AVM classification according Spetzler-Martin was 55 patients Grade I (20.5%), 114 Grade II (42.5%), 79 Grade III (29.5%), 19 Grade IV (7.1%), and 1 Grade V (0.4%). Median maximum AVM diameter was 3.0 cm (range, 0.3-10 cm). Median dose prescribed to the 80% isodose was 18 Gy (range, 12-22 Gy). Eighty-five patients (29.1%) had prior partial embolization; 141 patients (51.9%) experienced intracranial hemorrhage before RS. Median follow-up was 4.2 years.

RESULTS

Age at treatment, maximum diameter, nidus volume, and applied dose were significant factors for successful obliteration. Under presumption of proportional hazard in the dose range between 12 and 22 Gy/80% isodose, an increase of obliteration rate of approximately 25% per Gy was seen. On the basis of multivariate analysis, a prediction score was calculated including AVM maximum diameter and age at treatment. The prediction error up to the time point 8 years was 0.173 for the Heidelberg score compared with the Kaplan-Meier value of 0.192. An increase of the score of 1 point results in a decrease of obliteration chance by a factor of 0.447.

CONCLUSION

The proposed score is linac-based radiosurgery-specific and easy to handle to predict patient outcome. Further validation on an independent patient cohort is necessary.

Cyberknife hypofractionated stereotactic radiosurgery (HSRS) of resection cavity after excision of large cerebral metastasis: efficacy and safety of an 800 cGy × 3 daily fractions regimen

Development of hypofractionated stereotactic radiosurgery (HSRS) has expanded the size of lesion that can be safely treated by focused radiation in a limited number of treatment sessions. However, clinical data regarding the efficacy and morbidity of HSRS in the treatment of cerebral metastasis is lacking. Here, we review our experience with CyberKnife(®) HSRS for this indication. From 2005 to 2010, we identified 37 patients with large (>3 cm in diameter) cerebral metastases resection cavity that was treated with HSRS. This constituted approximately 8% of all treated resection cavities. We reviewed dose regimens, local control, distal control, and treatment associated morbidities. Primary sites for the metastatic lesions included: lung (n = 10), melanoma (n = 12), breast (n = 9), kidney (n = 4), and colon (n = 2). All patients underwent resection of the cerebral metastasis and received 800 cGy × 3 daily fractions to the resection cavity. Of the 37 patients treated, one-year follow-up data was available for 35 patients. The median survival was 5.5 months. Actuarial local control rate at 6 months was 80%. Local failures did not correlate with prior WBRT, or tumor histology. Distant recurrence occurred in 7 of the 35 patients. Morbidities associated with HSRS totaled 9%, including radiation necrosis (n = 1, 2.9%), prolonged steroid use (n = 1, 2.9%), and new-onset seizures (n = 1, 2.9%). This study demonstrates the safety and efficacy of an 800 cGy × 3 daily fractions CyberKnife(®) HSRS regimen for irradiation of large resection cavity. The efficacy compares favorably to historical data derived from patients undergoing WBRT, SRS, or brachytherapy.

Brain metastases from renal cell carcinoma. Should we change the current standard?

Renal cell carcinoma (RCC) is one of the most common sources of brain metastases, with an incidence that varies widely from 4% to 48% according to different studies. In addition, asymptomatic metastases occur in up to 33% of patients with metastatic RCC, further complicating the decision-making process in this poor prognosis population. The purpose of this review is to cover in depth the present state of knowledge on the diagnosis and management of patients with brain metastases from RCC, in order to assess whether the current standard should be challenged. The existing systems to predict response and survival will be reviewed, as well as the available therapeutic options regarding local treatment and systemic therapy, all within the context of updated data from clinical trials. In this regard, the role of novel targeted agents for the treatment of brain metastases from RCC, such as the multi-targeted receptor tyrosine kinase inhibitors sunitinib and sorafenib, will be updated and discussed.

Stereotactic radiosurgery for brain metastases: analysis of outcome and risk of brain radionecrosis

Purpose

to investigate the factors affecting survival and toxicity in patients treated with stereotactic radiosurgery (SRS), with special attention to volumes of brain receiving a specific dose (V10 - V16 Gy) as predictors for brain radionecrosis.

Patients and Methods

Two hundred six consecutive patients with 310 cerebral metastases less than 3.5 cm were treated with SRS as primary treatment and followed prospectively at University of Rome La Sapienza Sant'Andrea Hospital. Overall survival, brain control, and local control were estimated using the Kaplan-Meier method calculated from the time of SRS. Univariate and multivariate analysis using a Cox proportional hazards regression model were performed to determine the predictive value of prognostic factors for treatment outcome and SRS-related complications.

Results

Median overall survival and brain control were 14.1 months and 10 months, respectively. The 1-year and 2-year survival rates were 58% and 24%, and respective brain control were 43% and 22%. Sixteen patients recurred locally after SRS, with 1-year and 2-year local control rates of 92% and 84%, respectively. On multivariate analysis, stable extracranial disease and KPS >70 were associated with the most significant survival benefit. Neurological complications were recorded in 27 (13%) patients. Severe neurological complications (RTOG Grade 3 and 4) occurred in 5.8% of patients. Brain radionecrosis occurred in 24% of treated lesions, being symptomatic in 10% and asymptomatic in 14%. On multivariate analysis, V10 through V16 Gy were independent risk factors for radionecrosis, with V10 Gy and V12 Gy being the most predictive (p = 0.0001). For V10 Gy >12.6 cm³ and V12 Gy >10.9 cm³ the risk of radionecrosis was 47%.

Conclusions

SRS alone represents a feasible option as initial treatment for patients with brain metastases, however a significant subset of patients may develop neurological complications. Lesions with V12 Gy >8.5 cm³ carries a risk of radionecrosis >10% and should be considered for hypofractionated stereotactic radiotherapy especially when located in/near eloquent areas.

Differentiation between intra-axial metastatic tumor progression and radiation injury following fractionated radiation therapy or stereotactic radiosurgery using MR spectroscopy, perfusion MR imaging or volume progression modeling

OBJECTIVE

To determine the accuracy of magnetic resonance spectroscopy (MRS), perfusion MR imaging (MRP), or volume modeling in distinguishing tumor progression from radiation injury following radiotherapy for brain metastasis.

METHODS

Twenty-six patients with 33 intra-axial metastatic lesions who underwent MRS (n=41) with or without MRP (n=32) after cranial irradiation were retrospectively studied. The final diagnosis was based on histopathology (n=4) or magnetic resonance imaging (MRI) follow-up with clinical correlation (n=29). Cho/Cr (choline/creatinine), Cho/NAA (choline/N-acetylaspartate), Cho/nCho (choline/contralateral normal brain choline) ratios were retrospectively calculated for the multi-voxel MRS. Relative cerebral blood volume (rCBV), relative peak height (rPH) and percentage of signal-intensity recovery (PSR) were also retrospectively derived for the MRPs. Tumor volumes were determined using manual segmentation method and analyzed using different volume progression modeling. Different ratios or models were tested and plotted on the receiver operating characteristic curve (ROC), with their performances quantified as area under the ROC curve (AUC). MRI follow-up time was calculated from the date of initial radiotherapy until the last MRI or the last MRI before surgical diagnosis.

RESULTS

Median MRI follow-up was 16 months (range: 2-33). Thirty percent of lesions (n=10) were determined to be radiation injury; 70% (n=23) were determined to be tumor progression. For the MRS, Cho/nCho had the best performance (AUC of 0.612), and Cho/nCho >1.2 had 33% sensitivity and 100% specificity in predicting tumor progression. For the MRP, rCBV had the best performance (AUC of 0.802), and rCBV >2 had 56% sensitivity and 100% specificity. The best volume model was percent increase (AUC of 0.891); 65% tumor volume increase had 100% sensitivity and 80% specificity.

CONCLUSION

Cho/nCho of MRS, rCBV of MRP, and percent increase of MRI volume modeling provide the best discrimination of intra-axial metastatic tumor progression from radiation injury for their respective modalities. Cho/nCho and rCBV appear to have high specificities but low sensitivities. In contrast, percent volume increase of 65% can be a highly sensitive and moderately specific predictor for tumor progression after radiotherapy. Future incorporation of 65% volume increase as a pretest selection criterion may compensate for the low sensitivities of MRS and MRP.

A 10-year experience of radiosurgical treatment for cerebral arteriovenous malformations: a perspective from a series with large malformations. Clinical article

OBJECT

The purpose of this study was to describe a 10-year experience in the use of radiosurgery (RS) for patients with arteriovenous malformations (AVMs) in Puerto Rico.

METHODS

This retrospective analysis was performed for all patients with AVMs treated with RS by the senior author (R.H.B.) in Puerto Rico. Between February 1999 and December 2009, a total of 83 patients underwent the procedure. All charts were reviewed for recollection of demographic data, and AVM and treatment characteristics. Clinical and radiographic follow-up information was collected retrospectively.

RESULTS

Eighty-three patients were treated and 86 RS procedures for AVMs were performed during a 10-year period. Eight patients were lost to follow-up. The remaining 75 patients included 36 males and 39 females, whose median age was 34.5 years. Hemorrhage was the initial presentation in 40% of patients. Fifty-seven AVMs (73%) were treated previously with endovascular neurosurgery, without success. The median volume of the malformation was 17.7 ml. Nearly 65% of the malformations were considered large (≥ 10 ml) in volume. Forty patients had AVMs with largest diameter ≥ 3.5 cm. The overall obliteration rate was 56.4%, and the median time for obliteration was 29 months. The AVMs ≥ 3.5 cm in diameter had a greater latency period than those < 3.5 cm (31 months vs 46 months, respectively; p = 0.01). In addition, AVM obliteration was inversely associated with its volume, especially in large lesions (p = 0.037). In bivariate analysis, patients achieving obliteration had lower Spetzler-Martin scores compared with patients in whom obliteration was not achieved (p = 0.009). Postradiosurgery hemorrhages were seen in 9 cases. Eleven patients underwent surgery after RS. Major neurological deficits developed in 9 patients, whereas 17 had only minor deficits. The occurrence of neurological deficits was significantly associated with lesions with volume ≥ 10 ml.

CONCLUSIONS

Radiosurgery is a reasonable treatment option for AVMs in the majority of cases, in spite of the large, difficult-to-treat malformations.

Characterisation of dose distribution in linear accelerator-based intracranial stereotactic radiosurgery with the dynamic conformal arc technique: consideration of the optimal method for dose prescription and evaluation

OBJECTIVES

The purpose of this study was to characterise dose distribution in linear accelerator-based intracranial stereotactic radiosurgery using the dynamic conformal arc technique, and to validate the pertinence of dose prescription to the specific percentage isodose surface (IDS).

METHODS

73 plans for brain metastases were reviewed and replanned with a uniform method for target definition and treatment planning.

RESULTS

In all cases except 1 the dose prescription to the 80% IDS satisfied the criteria of the standardised prescription IDS as previously proposed. However, both of the planning target volume (PTV) coverage values for the 80% and 90% IDSs and the PTV D99 and D95 (IDS receiving at least 99% or 95% of the PTV) were inconsistent and significantly increased as a function of the PTV size. The 80% IDS for a PTV of more than 5 cm(3) achieved adequate PTV coverage without a leaf margin. The dose conformity for 80% IDS gradually worsened as the PTV increased, whereas that for the PTV D99 or D95 improved as a function of the PTV size. The addition of a leaf margin attained 100% PTV coverage for 80% IDS, while leading to a poorer dose conformity.

CONCLUSION

The dose prescription to the specific percentage IDS does not necessarily guarantee consistent target coverage, D99 and D95, and desirable dose conformity in proportion to the target volume. The dose prescription and evaluation at the specific target coverage would therefore be preferable as an objective method in order to report the "marginal dose" and to clearly compare the planning parameters with those from other modalities.

Dose gradient analyses in Linac-based intracranial stereotactic radiosurgery using Paddick's gradient index: consideration of the optimal method for plan evaluation

The objective of our study was to describe the dose gradient characteristics of Linac-based stereotactic radiosurgery using Paddick's gradient index (GI) and to elucidate the factors influencing the GI value. Seventy-three plans for brain metastases using the dynamic conformal arcs were reviewed. The GI values were calculated at the 80% and 90% isodose surfaces (IDSs) and at the different target coverage IDSs (D99, D95, D90, and D85). The GI values significantly decreased as the target coverage of the reference IDS increased (the percentage of the IDS decreased). There was a significant inverse correlation between the GI values and target volume. The plans generated with the addition of a 1-mm leaf margin had worse GI values both at the D99 and D95 relative to those without leaf margin. The number and arrangement of arcs also affected the GI value. The GI values are highly sensitive to (1) the IDS selection variability for dose prescription or evaluation, (2) the target volume, and (3) the planning method. To objectively compare the quality of dose gradient between rival plans, it would be preferable to employ the GI defined at the reference IDS indicating the specific target coverage (e.g., D95), irrespective of the intended marginal dose. The modified GI (mGI), defined in this study, substituting the denominator of the original GI with the target volume, would be useful to compensate for the false superior GI value in cases of target over-coverage with the reference IDS and to objectively evaluate the dose gradient outside the target boundary.