Adaptive Radiotherapy for H&N Cancers: What Forms of Rescheduling? Debates Surrounding a Clinical Case

Transcription

1 clinical cases Adaptive Radiotherapy for H&N Cancers: What Forms of Rescheduling? Debates Surrounding a Clinical Case Ovidiu Veresezan 1, Catherine Dejean 2, Mathieu Gautier 2, Juliette Thariat 3 1) Radiotherapy Department, Henri Becquerel Cancer Center, Rouen; 2) Physics Unit, Antoine Lacassagne Cancer Center, Nice; 3) Radiotherapy Department, Antoine Lacassagne Cancer Center, Nice, France This work was realized in the cadre of the Inter-university diploma of the High Technicity Radiation Therapy at Antoine Lacassagne Cancer Center, Nice, France Introduction: Many studies have suggested that anatomical changes of risk organs and target volumes occur during irradiation, especially for H&N cancers. We strived to analyze the clinically relevant thresholds and technical means currently available from treatment planning systems and additional software to adapt radiotherapy in the course of treatment. Material and method: We will describe the replanning process through the example of a patient, irradiated for oropharyngeal squamous cell carcinoma, who lost 10% of his body weight. Shape and volume changes of the target volumes and organs at risk / normal tissues were analyzed for the need to reschedule the initial dosimetry in order to avoid tumor under-coverage and excess toxicities. Results: Replanning represent a series of challenging issues for radiation oncologists and physicists: current treatment planning systems either allow accumulating two dosimetries using one single reference scanner or adding dose matrices obtained through different CT acquisitions, with uncertainties regarding periphery dose. Neither of the two systems respond in a satisfactory manner to all the challenges of rescheduling and do not represent the continuous variation of the patient anatomy. Conclusions: The routine use of adaptive radiotherapy requires optimization of current dosimetry systems in order to allow the real-time addition of doses per different scanners. Another prerequisite would be the evaluation of the actual administered doses with the aid of daily CBCTs. Keywords: adaptive radiotherapy, H&N cancers, anatomical changes, replanning. Introduction Given its steep gradients, modern intensity modulated radiation therapy (IMRT) techniques need insurance as per their capability to fit initial treatments plans to daily dose delivery. In the case of H&N cancer patients, some authors have recorded important volume, shape or position-related changes of organs at risk and/or target volumes in the course of irradiation (1-3). These findings raise numerous questions regarding the doses actually administered at the level of these structures. Iterative rescheduling in the course of treatment was proposed in order to make sure irradiation is performed with maximum precision and perfectly overlapped on the planned dosimetry. This approach raises the problem of objective methodology to add different treatment plans, considering one or more CT acquisitions. Material and method The phases of rescheduling in the case of a patient treated with static beam orientation IMRT for a right side oropharynx epidermoid carcinoma who lost 10% of his body Journal of Radiotheraphy & Medical Oncology December 2012 Vol. XVII No 2: Address for correspondence: Ovidiu Veresezan Radiotherapy Department Henri Becquerel Cancer Center 1 Rue d Amiens, 76038, Rouen, France veresezanovidiu@yahoo.com weight thus demanding a new contention mask and a new provisional dosimetry at the 44 Gy dose will be described. The doses administered to various structures have been evaluated using the Eclipse computation system of Varian and the comparison of dose matrixes using the Artiview program of the Acquilab laboratory. Results Mr. B, age 75 presented a diagnosis of stage T3N2bM0 right side oropharynx epidermoid carcinoma. An exclusive radiotherapy treatment was proposed. The initial provisional dosimetry was calculated based on a CT in treatment position. The treatment was delivered by a IMRT sliding window technique, with seven beams. Dose-volume histograms are illustrated in Fig. 1. During weekly evaluation, at the 44 Gy dose, weight loss was 7kg (10% of the initial body weight) despite careful nutritional support concerned with significant changes of target volumes and organs at risk. We thus performed a cone beam CT (CBCT) which suggested a 5mm neck thickness loss and thermoplastic mask unfit for new skin patient contours. Our current CBCT cannot be used for planning, because the Hounsfield units are not stable enough. A new planning CT scan in treatment position, with a new contention mask was thus performed. A fusion of the two scans with the aid of the Eclipse program of Varian indicated important changes of the structures (Fig. 2). A manual registration of the 2 CT scans

2 Adaptive radiotherapy for H&N cancers 79 was made by using a rigid fusion on the bony structure. Organs at risk and target volumes were marked off on the new CT acquisition and, during a first phase, the initial irradiation ballistic was applied for this new anatomical acquisition. We found a maximum spinal cord dose of 55.4 Gy versus a 43.9 Gy for the initial dosimetry. Obviously, rescheduling was necessary as the dose administered for the spinal cord was unacceptable. We evaluated the doses of the organs at risk and target volumes obtained in the different hypothetical options, using worst case scenarios for the spinal cord dose, thus avoiding the risk of radiation myelitis in any situation. At the same time, we sought to optimize the dosimetry so as to ensure an optimum coverage of the target volumes for the remnant course of irradiation. The most pessimistic option for the spinal dose was when the initial ballistic was applied for the replanning scan and added with the ballistic optimized using Eclipse applied to the same rescheduling scan (rescheduled dosimetry). Without compromising the irradiation of target volumes, we chose the safer option to protect against post radiation myelitis in the final computation of the rescheduled dosimetry are illustrated in Fig. 3. A comparison of dose ratios considering initial doses on the initial scan with the same beam geometry applied for the rescheduled scan is presented in Table I. The maximum dose obtained at PRV spinal cord level goes from 43.9 to 55.4 Gy. We also noticed a moderate increase of median doses in the parotids, especially on the contralateral side. There are no significant differences regarding the irradiation of target volumes, despite areas of over/undercoverage at this level, with a degradation of the homogeneity index for the initial beam geometry applied for the rescheduled CT scan. The theoretical maximum dose in the spinal cord needed a new optimization so as to avoid the risk of myelitis. PTV 70 PTV 56 Brainstem L-Parot PRV spinal cord Spinal cord R-Parot PTV 63 (L-Parot)-PTV (R-Parot)-PTV Fig. 1. DVH corresponding with the initial dosimetry of patient Fig. 2. Fusion of the initial and rescheduling scan

3 80 Veresezan et al PTV 56 PTV 63 PTV 70 Spinal cord Brainstem PRV spinal cord PRV Brainstem Fig. 3. DVH corresponding to the rescheduled dosimetry Table I. A comparison of the dose ratio considering the initial ballistic on the initial scan with the same ballistic applied for the rescheduling scan Dosi 0-70Gy initial CT Dosi 0-70Gy rescheduling scan Dmin Dmax Dmed Dmin Dmax Dmed PRV spinal cord PRV brain stem Spinal cord Brain stem Right Parotid Left Parotid PTV PTV PTV Table II presents dose comparisons between the rescheduled dosimetry and addition of dose matrices in Artiview (initial dosimetry with the initial scan: 0-44 Gy, added to the new ballistic optimization based on the rescheduling scan: Gy). Irradiation of target volumes was comparable in both cases, with lesser homogeneity with cumulated dosimetry in Eclipse. Table III presents dose comparisons between the initial provisional dosimetry (0-70 Gy, initial CT), and the rescheduled dosimetry. Despite the greater dose at the level of the parotids and spinal cord for the first 44 Gy, we could compensate, at least partially, by optimizing the dosimetry for the subsequent radiation course. Table II. A comparison of the dose ratio between the rescheduled dosimetry and the dosimetry resulting through the addition of the dose matrixes in Artiview Rescheduled dosimetry (Eclipse) Summation D matrixes in Artiview Dmin Dmax Dmed Dmin Dmax Dmed PRV spinal cord PRV brain stem Spinal cord Brain stem Right Parotid Left Parotid PTV PTV PTV

4 Adaptive radiotherapy for H&N cancers 81 Table III. A comparison of the dose ratio between the initial, optimized plan in Eclipse for the initial scan and the dose ratio of the rescheduled dosimetry Dosi 0-70 Gy initial scan Rescheduled dosimetry (Eclipse) Dmin Dmax Dmed Dmin Dmax Dmed PRV spinal cord PRV brain stem Spinal cord Brain stem Right Parotid Left Parotid PTV PTV PTV Discussion Adaptive radiotherapy represents an active field of research for a radiotherapist for two reasons: first, it aims at optimizing radiation delivery in target volumes while taking into account real time inter-fraction changes; secondly, it allows for an adjustment of dose-volume histograms (DVHs) in the course of the treatment, to limit the doses administered to organs at risk. In case of significant anatomical deformations, as in head and neck tumours (tumour shrinking and decrease in volume of the salivary glands), replanning appears to be necessary, corresponding to the adaptive radiotherapy. This should ideally be monitored and possibly triggered based on a calculation of cumulative dose, session after session, compared to the initial planning dose, corresponding to the concept of dose-guided adaptive radiotherapy (4). The practical solutions of addition of replanning dosimetries are not standardized yet. Creating a model of optimally cumulated doses is the purpose of an ongoing Clinical Research Program conducted by Pr. De Crevoisier s team to standardize adaptive radiotherapy in HN cancer. Our goal was to analyze the technical solutions of dose addition currently available in routine practice. Mr. B s treatment was performed in accordance with the initial provisional dosimetry until the 44 Gy dose and a rescheduling was accomplished subsequently. Other scenarios were also used to evaluate actual doses administered: 1. We postulated that the actual dose distribution corresponded to the initial provisional dosimetry up to 44 Gy, and that for the subsequent 26 Gy such a dose is correctly evaluated using replanning. In reality, at least a part of the dose is falsely estimated by the initial provision, since the anatomical modifications precedes the rescheduled CT scan. This scenario underestimated the spinal cord dose. 2. We ignored the initial anatomical acquisition and considered that from the beginning, the anatomy of the patient overlapped that obtained with the rescheduled CT scan. We evaluated the effective dose by applying the initial beam geometry to the new plan. The maximal spinal cord dose was 55.4 Gy. A new dosimetry was necessary in order to avoid myelopathy. This scenario is certainly different from the real situation, but ensures a minimum risk of myelitis thus compelling us to impose strict dosimetry restrictions for the spinal cord for the remaining sessions of irradiation. 3. A third option would be to imagine that a part of the irradiation was accomplished with the initial beam geometry as per the initial anatomy, and the remaining treatment is performed on an anatomy corresponding to the rescheduled CT scan. Seeing that the patient lost 10% of the initial body weight, we could consider the threshold between the two anatomies at the time when the patient had lost 5% of the initial body weight. This moment corresponds to the 26 Gy dose. The clinician must answer many difficult questions when rescheduling treatment in the course of radiotherapy. The changes of different structures are progressive, but the radiotherapist must use one or more exact and virtual provisional dosimetries in order to evaluate the doses effectively administered on organs at risk and target volumes (5-8). Can we consider the delivered dose as being estimated by the initial dosimetry until rescheduling, and by the dosimetry rescheduled subsequently, without asking about the evolution of progressive modifications between two CT exams? Or is it more logical to consider that the actual anatomy is approximated more precisely by the rescheduled CT acquisition? The reality is likely situated between these two extremes, but how can we evaluate the curve of anatomical modifications between the two CT acquisitions? Is there a computerized application available to calculate the temporary set in time of various structures or a pattern that could generate an average anatomical acquisition? A last option would be to accomplish an elastic fusion of the first and second CT exams and attempt to find a formula that would allow the computation of the changes associated to each voxel. By doing so we would obtain, starting from the initial anatomy, the final anatomy corresponding to the second CT examination. Although theoretically attractive,

5 82 Veresezan et al this last option is not yet accessible in practice because, we do not presently have reliable criteria to estimate the real changes in time of the different anatomical structures and its implication on dose. If replacement of the initial beam geometry on iterative CBCT with real time reevaluation of DVHs can already be imagined in practice, optimal algorithms of additional dosimetries on a continuously changing anatomy have yet to be achieved (9-11). For the dosimetry program, medical images are a function of the number of Hounsfield Units, while for the physician these represent distinct anatomical structures in which the dose related objectives are extremely heterogeneous, such as a tumor tissue or highly radiosensitive organs at risk. Ideally, radiotherapists could add DVHs of the initial scan with the daily DVHs of different CBCTs while taking into account the progressive deformation of different structures, along with a perfect overlapping of the planned dose with the dose really administered at the treatment station. This is not yet possible in practice, as the Eclipse treatment planning system / type computational systems uses one reference CT scan for the addition of different treatment plans. Research in this direction are ongoing with suppliers of dosimetry related programs, but many meetings between physicists, computer scientists, mathematicians and physicians will be required so that each understands better the language, possibilities, objectives and expectations of others (12,13). The Artiview program allows the transfer of dose matrices obtained from a CT scan onto a different scan, and therefore the administered dose with the initial anatomy can be added with the rescheduled dosimetry starting from a new scan acquisition. Still, adding of dose matrices is only accomplished on the common volume of the two CT exams and we cannot estimate the peripheral doses. Presently, with Artiview we can accomplish a dose addition through an elastic fusion of the reference scan on daily CBCTs while evaluating the dose administered to each voxel corresponding to target volumes and organs at risk (9-11). An example is depicted in Fig. 4 for a spinal cord voxel. Dose matrices were transferred onto the new scan and the doses received by the same spinal cord voxel added on the two scans. Although this method is still underdeveloped, advances are anticipated in the near future. Conclusions The concept of adaptive radiotherapy has been used by radiotherapists since the beginning of the 20th century when fields were adapted on the modifications of the tumor mass in the course of treatment (14,15). Daily online CBCTs can be used in order to eliminate systematic and random setup errors (9-11). Some teams have developed their own programs of addition of iterative treatment plans accomplished with different CTs (16-18). Still, current dose computation programs do not allow this option in our current practice. These programs will probably proposed soon by firms and so, in the near future, we will be able to accomplish a highly precise adaptive radiotherapy by adding actually administered doses with the aid of daily CBCTs, of elastic fusion and iterative rescheduling of treatment. In this manner, we will be able to guarantee optimal overlapping of the envisioned dose with the effectively administered dose. Elastic transformation Dose (spinal cord) = D1+D2 additional DVH Fig. 4. Dose addition of a spinal cord voxel by an elastic fusion of the reference scan on daily CBCT.

6 Adaptive radiotherapy for H&N cancers 83 Because the continuous readjustment of the treatment will compensate for uncertainties related to repositioning and internal movement of structures, we should be able to reduce the current margins and overlap PTVs and CTVs with a considerable abatement of irradiated volumes and doses for organs at risk and therefore improve the therapeutic index. References 1. Barker JL Jr, Garden AS, Ang KK, et al. Quantification of volumetric and geometric changes occurring duringfractionated radiotherapy for head-and-neck cancer using an integrated CT/linear accelerator system. Int. J.Radiat. Oncol. Biol. Phys. 2004;59: Castadot P, Geets X, Lee JA, Christian N, Grégoire V. Assessment by a deformable registration method of thevolumetric and positional changes of target volumes and organs at risk in pharyngo-laryngeal tumors treatedwith concomitant chemo-radiation. Radiother Oncol. 2010;95: Hansen EK, Bucci MK, Quivey JM, Weinberg V, Xia P. Repeat CT imaging and replanning during the courseof IMRT for head-and-neck cancer. Int. J. Radiat. Oncol. Biol. Phys. 2006;64: Louvel G, et al. Image-guided and adaptive radiotherapy. Cancer Radiother (2012), 5. Bhide SA, Davies M, Burke K, et al. Weekly volume and dosimetric changes during chemoradiotherapy with intensity-modulated radiation therapy for head and neck cancer: a prospective observational study. Int. J. Radiat. Oncol. Biol. Phys. 2010;76: You SH, Kim SY, Lee CG, et al. Is There a Clinical Benefit to Adaptive Planning During Tomotherapy in Patients with Head and Neck Cancer at Risk for Xerostomia? American Journal of Clinical Oncology Jun; 35(3): Schwartz DL, Dong L. Adaptive radiation therapy for head and neck cancer-can an old goal evolve into a new standard? J Oncol. 2011;2011. pii: Epub 2010 Aug Cazoulat G, et al. De la radiothérapie guidée par l image à la radiothérapie guidée par la dose. Cancer Radiother 15 (2011) Søvik A, Rødal J, Skogmo HK, et al. Adaptive radiotherapy based on contrast enhanced cone beam CT imaging. Acta Oncol. 2010;49: Marchant TE, Amer AM, Moore CJ. Measurement of inter and intra fraction organ motion in radiotherapy using cone beam CT projection images. Phys Med Biol. 2008;53: Thomas THM, Devakumar D, Purnima S, Ravindran BP. The adaptation of megavoltage cone beam CT for use in standard radiotherapy treatment planning. Phys Med Biol. 2009;54: Vestergaard A, Søndergaard J, Petersen JB, Høyer M, Muren LP. A comparison of three different adaptive strategies in image-guided radiotherapy of bladder cancer. Acta Oncol. 2010;49: Wu QJ, Li T, Wu Q, Yin F-F. Adaptive radiation therapy: technical components and clinical applications. Cancer J. 2011;17: Bataini JP, Jaulerry C, Brunin F, Ponvert D, Ghossein NA. Significance and therapeutic implications of tumor regression following radiotherapy in patients treated for squamous cell carcinoma of the oropharynx and pharyngolarynx. Head Neck. 1990;12: Bataini JP, Bernier J, Jaulerry C, et al. Impact of neck node radioresponsiveness on the regional control probability in patients with oropharynx and pharyngolarynx cancers managed by definitive radiotherapy. International Journal of Radiation Oncology*Biology *Physics. 1987;13: Chao KSC, Bhide S, Chen H, et al. Reduce in variation and improve efficiency of target volume delineation by a computer-assisted system using a deformable image registration approach. Int. J. Radiat. Oncol. Biol. Phys.2007;68: Wang H, Dong L, O Daniel J, et al. Validation of an accelerated demons algorithm for deformable image registration in radiation therapy. Physics in Medicine and Biology. 2005;50: Lee C, Langen KM, Lu W, et al. Assessment of parotid gland dose changes during head and neck cancer radiotherapy using daily megavoltage computed tomography and deformable image registration. International Journal of Radiation Oncology, Biology, Physics. 2008;71: