Issues in optimization for planning of intensity-modulated radiation therapy

Knowledge-based multi-institution plan prediction of whole breast irradiation with tangential fields

PURPOSE

To quantify inter-institute variability of Knowledge-Based (KB) models for right breast cancer patients treated with tangential fields whole breast irradiation (WBI).

MATERIALS AND METHODS

Ten institutions set KB models by using RapidPlan (Varian Inc.), following previously shared methodologies. Models were tested on 20 new patients from the same institutes, exporting DVH predictions of heart, ipsilateral lung, contralateral lung, and contralateral breast. Inter-institute variability was quantified by the inter-institute SD_int of predicted DVHs/Dmean. Association between lung sparing vs PTV coverage strategy was also investigated. The transferability of models was evaluated by the overlap of each model's geometric Principal Component (PC1) when applied to the test patients of the other 9 institutes.

RESULTS

The overall inter-institute variability of DVH/Dmean ipsilateral lung dose prediction, was less than 2% (20%-80% dose range) and 0.55 Gy respectively (1SD) for a 40 Gy in 15 fraction schedule; it was < 0.2 Gy for other OARs. Institute 6 showed the lowest mean dose prediction value and no overlap between PTV and ipsilateral lung. Once excluded, the predicted ipsilateral lung Dmean was correlated with median PTV D99% (R² = 0.78). PC1 values were always within the range of applicability (90th percentile) for 7 models: for 2 models they were outside in 1/18 cases. For the model of institute 6, it failed in 7/18 cases. The impact of inter-institute variability of dose calculation was tested and found to be almost negligible.

CONCLUSIONS

Results show limited inter-institute variability of plan prediction models translating in high inter-institute interchangeability, except for one of ten institutes. These results encourage future investigations in generating benchmarks for plan prediction incorporating inter-institute variability.

Using individual patient anatomy to predict protocol compliance for prostate intensity-modulated radiotherapy

If a prostate intensity-modulated radiation therapy (IMRT) or volumetric-modulated arc therapy (VMAT) plan has protocol violations, it is often a challenge knowing whether this is due to unfavorable anatomy or suboptimal planning. This study aimed to create a model to predict protocol violations based on patient anatomical variables and their potential relationship to target and organ at risk (OAR) end points in the setting of definitive, dose-escalated IMRT/VMAT prostate planning. Radiotherapy plans from 200 consecutive patients treated with definitive radiation for prostate cancer using IMRT or VMAT were analyzed. The first 100 patient plans (hypothesis-generating cohort) were examined to identify anatomical variables that predict for dosimetric outcome, in particular OAR end points. Variables that scored significance were further assessed for their ability to predict protocol violations using a Classification and Regression Tree (CART) analysis. These results were then validated in a second group of 100 patients (validation cohort). In the initial analysis of the hypothesis-generating cohort, percentage of rectum overlap in the planning target volume (PTV) (%OR) and percentage of bladder overlap in the PTV (%OB) were highlighted as significant predictors of rectal and bladder dosimetry. Lymph node treatment was also significant for bladder outcomes. For the validation cohort, CART analysis showed that %OR of < 6%, 6% to 9% and > 9% predicted a 13%, 63%, and 100% rate of rectal protocol violations respectively. For the bladder, %OB of < 9% vs > 9% is associated with 13% vs 88% rate of bladder constraint violations when lymph nodes were not treated. If nodal irradiation was delivered, plans with a %OB of < 9% had a 59% risk of violations. Percentage of rectum and bladder within the PTV can be used to identify individual plan potential to achieve dose-volume histogram (DVH) constraints. A model based on these factors could be used to reduce planning time, improve work flow, and strengthen plan quality and consistency.

Elective lymph node irradiation with intensity-modulated radiotherapy: is conventional dose fractionation necessary?

PURPOSE

Intensity-modulated radiation therapy (IMRT) is the standard of care for head-and-neck cancer (HNC). We treated patients with HNC by delivering either a moderate hypofractionation (MHF) schedule (66 Gy at 2.2 Gy per fraction to the gross tumor [primary and nodal]) with standard dose fractionation (54-60 Gy at 1.8-2.0 Gy per fraction) to the elective neck lymphatics or a conventional dose and fractionation (CDF) schedule (70 Gy at 2.0 Gy per fraction) to the gross tumor (primary and nodal) with reduced dose to the elective neck lymphatics. We analyzed these two cohorts for treatment outcomes.

METHODS AND MATERIALS

Between November 2001 and February 2009, 89 patients with primary carcinomas of the oral cavity, larynx, oropharynx, hypopharynx, and nasopharynx received definitive IMRT with or without concurrent chemotherapy. Twenty patients were treated using the MHF schedule, while 69 patients were treated with the CDF schedule. Patient characteristics and dosimetry plans were reviewed. Patterns of failure including local recurrence (LR), regional recurrence (RR), distant metastasis (DM), disease-free survival (DFS), overall survival (OS), and toxicities, including rate of feeding tube placement and percentage of weight loss, were reviewed and analyzed.

RESULTS

Median follow-up was 31.2 months. Thirty-five percent of patients in the MHF cohort and 77% of patients in the CDF cohort received chemotherapy. No RR was observed in either cohort. OS, DFS, LR, and DM rates for the entire group at 2 years were 89.3%, 81.4%, 7.1%, and 9.4%, respectively. Subgroup analysis showed no significant differences in OS (p = 0.595), DFS (p = 0.863), LR (p = 0.833), or DM (p = 0.917) between these two cohorts. Similarly, no significant differences were observed in rates of feeding tube placement and percentages of weight loss.

CONCLUSIONS

Similar treatment outcomes were observed for MHF and CDF cohorts. A dose of 50 Gy at 1.43 Gy per fraction may be sufficient to electively treat low-risk neck lymphatics.

Direct aperture optimization as a means of reducing the complexity of Intensity Modulated Radiation Therapy plans

Intensity Modulated Radiation Therapy (IMRT) is a means of delivering radiation therapy where the intensity of the beam is varied within the treatment field. This is done by dividing a large beam into many small beamlets. Dose constraints are assigned to both the target and sensitive structures and computerised inverse optimization is performed to find the individual weights of this large number of beamlets. The computer adjusts the intensities of these beamlets according to the required planning dose objectives. The optimized intensity patterns are then decomposed into a series of deliverable multi leaf collimator (MLC) shapes in the sequencing step.

One of the main problems of IMRT, which becomes even more apparent as the complexity of the IMRT plan increases, is the dramatic increase in the number of Monitor Units (MU) required to deliver a fractionated treatment. The difficulty with this increase in MU is its association with increased treatment times and a greater leakage of radiation from the MLCs increasing the total body dose and the risk of secondary cancers in patients. Therefore one attempts to find ways of reducing these MU without compromising plan quality.

The design of inverse planning systems where the beam is divided into small beamlets to produce the required intensity map automatically introduces complexity into IMRT treatment planning. Plan complexity is associated with many negative factors such as dosimetric uncertainty and delivery issues A large search space is required necessitating much computing power. However, the limitations of the delivery technology are not taken into consideration when designing the ideal intensity map therefore a further step termed the sequencing step is required to convert the ideal intensity map into a deliverable one.

Many approaches have been taken to reduce the complexity. These include setting intensity limits, putting penalties on the cost function and using smoothing filters Direct Aperture optimization (DAO) incorporates the limitations of the delivery technology at the initial design of the intensity map thereby eliminating the sequencing step. It also gives control over the number of segments and hence control over the complexity of the plan although the design of the segments is independent of the person preparing the plan.

In vitrostudy of cell survival following dynamic MLC intensity-modulated radiation therapy dose deliverya)

The possibility of reduced cell kill following intensity-modulated radiation therapy (IMRT) compared to conventional radiation therapy has been debated in the literature. This potential reduction in cell kill relates to prolonged treatment times typical of IMRT dose delivery and consequently increased repair of sublethal lesions. While there is some theoretical support to this reduction in cell kill published in the literature, direct experimental evidence specific to IMRT dose delivery patterns is lacking. In this study we present cell survival data for three cell lines: Chinese hamster V79 fibroblasts, human cervical carcinoma, SiHa and colon adenocarcinoma, WiDr. Cell survival was obtained for 2.1 Gy delivered as acute dose with parallel-opposed pair (POP), irradiation time 75 s, which served as a reference; regular seven-field IMRT, irradiation time 5 min; and IMRT with a break for multiple leaf collimator (MLC) re-initialization after three fields were delivered, irradiation time 10 min. An actual seven-field dynamic MLC IMRT plan for a head and neck patient was used. The IMRT plan was generated for a Varian EX or iX linear accelerator with 120 leaf Millenium MLC. Survival data were also collected for doses 1X, 2X, 3X, 4X, and 5x 2.1 Gy to establish parameters of the linear-quadratic equation describing survival following acute dose delivery. Cells were irradiated inside an acrylic cylindrical phantom specifically designed for this study. Doses from both IMRT and POP were validated using ion chamber measurements. A reproducible increase in cell survival was observed following IMRT dose delivery. This increase varied from small for V79, with a surviving fraction of 0.8326 following POP vs 0.8420 following uninterrupted IMRT, to very pronounced for SiHa, with a surviving fraction of 0.3903 following POP vs 0.5330 for uninterrupted IMRT. When compared to IMRT or IMRT with a break for MLC initialization, cell survival following acute dose delivery was significantly different, p < 0.05, in three out of six cases. In contrast, when cell survival following IMRT was compared to that following IMRT with a break for MLC initialization the difference was always statistically insignificant. When projected to a 30 fraction treatment, dose deficit to bring cell survival to the same value as in POP was calculated as 4.1, 24.9, and 31.1 Gy for V79, WiDr, and SiHa cell lines, respectively. The dose deficit did not relate to the alpha/beta ratio obtained in this study for the three cell lines. Clinical data do not show reduction in local control following IMRT. Possible reasons for this are discussed. The obtained data set can serve as a test data set for models designed to explore the effect of dose delivery prolongation/fractionation in IMRT on radiation therapy outcome.

Hypoxia imaging-directed radiation treatment planning

Increasing evidence supports the role of the tumor microenvironment in modulating cancer behavior. Tissue hypoxia, an important and common condition affecting the tumor microenvironment, is well established as a resistance factor in radiotherapy. Increasing evidence points to the ability of hypoxia to induce the expression of gene products, which confer aggressive tumor behavior and promote broad resistance to therapy. These factors suggest that determining the presence or absence of tumor hypoxia is important in planning cancer therapy. Recent advances in PET hypoxia imaging, conformal radiotherapy, and imaging-directed radiotherapy treatment planning now make it possible to perform hypoxia-directed radiotherapy. We review the biological aspects of tumor hypoxia and PET imaging approaches for measuring tumor hypoxia, along with methods for conformal radiotherapy and image-guided treatment, all of which provide the underpinnings for hypoxia-directed therapy. As a case example, we review emerging data on PET imaging of hypoxia to direct radiotherapy.

IMRT optimization: Variability of solutions and its radiobiological impact

We aim at (1) defining and measuring a "complexity" index for the optimization process of an intensity modulated radiation therapy treatment plan (IMRT TP), (2) devising an efficient approximate optimization strategy, and (3) evaluating the impact of the complexity of the optimization process on the radiobiological quality of the treatment. In this work, for a prostate therapy case, the IMRT TP optimization problem has been formulated in terms of dose-volume constraints. The cost function has been minimized in order to achieve the optimal solution, by means of an iterative procedure, which is repeated for many initial modulation profiles, and for each of them the final optimal solution is recorded. To explore the complexity of the space of such solutions we have chosen to minimize the cost function with an algorithm that is unable to avoid local minima. The size of the (sub)optimal solutions distribution is taken as an indicator of the complexity of the optimization problem. The impact of the estimated complexity on the probability of success of the therapy is evaluated using radiobiological indicators (Poissonian TCP model [S. Webb and A. E. Nahum, Phys. Med. Biol. 38(6), 653-666 (1993)] and NTCP relative seriality model [Kallman et al., Int. J. Radiat. Biol. 62(2), 249-262 (1992)]). We find in the examined prostate case a nontrivial distribution of local minima, which has symmetry properties allowing a good estimate of near-optimal solutions with a moderate computational load. We finally demonstrate that reducing the a priori uncertainty in the optimal solution results in a significant improvement of the probability of success of the TP, based on TCP and NTCP estimates.