CT imaging-guided dose painting in radiation therapy

Cancer Letters 355 (2014) 169–175

Contents lists available at ScienceDirect

Cancer Letters j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / c a n l e t

Mini-review

PET/CT imaging-guided dose painting in radiation therapy Xiaorong Shi a,b,1, Xue Meng a,1, Xindong Sun a, Ligang Xing a, Jinming Yu a,* a Department of Radiation Oncology, Key Laboratory of Radiation Oncology of Shandong Province, Shandong Cancer Hospital, Shandong University, Jinan, Shandong Province, China b Department of Oncology, Tangdu Hospital, Fourth Military Medical University, Xi’an, Shanxi Province, China

A R T I C L E

I N F O

Article history: Received 18 May 2014 Received in revised form 17 July 2014 Accepted 26 July 2014 Keywords: Functional imaging RT treatment planning Dose painting by contours Dose painting by numbers

A B S T R A C T

Application of functional imaging to radiotherapy (RT) is a rapidly expanding field with the development of new modalities and techniques. Functional imaging of PET in conjunction with RT provides new avenues towards the clinical application of dose painting – a new RT strategy delivering optimized dose redistribution according to the functional imaging information to further improve tumour control. Two prototypical strategies of dose painting are reviewed: dose painting by contours (DPBC) and dose painting by numbers (DPBN). DPBN set a linear correlation of the boost dose and image intensity of this same voxel while homogeneous dose is given to the subvolume contoured by a threshold created in PET images in DPBC. Both comply with strict organs at risk (OAR) constraints and are alternatives for boosting subvolumes in clinical practice. This review focuses on the rationale, target validation, dose prescription verification and evaluation and recent clinical achievements in the field of integrating PET imaging into RT treatment planning. Further research is necessary in order to investigate unresolved problems in its routine clinical application thoroughly. © 2014 Elsevier Ireland Ltd. All rights reserved.

Introduction It is illustrated that there is a strong correlation between local-regional control (LRC) and overall survival (OS) [1–4]. A straightforward way to improve local tumour control that proved by large bodies of evidence is to increase the radiation dose further [1,5–7]. However, we seem passive in the battle with cancer since the OS is not in keep with the clear improvement in various new modalities and techniques of treating cancer over decades. Current treatment prescriptions are already close to patient tolerance making it nearly impossible to raise the dose prescription of the whole target tumour. The failure of Radiation Therapy Oncology Group (RTOG) 0617 which beyond our widest expectation is another robust evidence. The side effects of pulmonary or cardiopulmonary from high thoracic radiation dose are the most likely explanation of the failure

Abbreviations: RT, radiotherapy; DPBC, dose painting by contours; DPBN, dose painting by numbers; OAR, organs at risk; LRC, local-regional control; OS, overall survival; RTOG, Radiation Therapy Oncology Group; PTV, planning target volume; LRF, local-regional failure; GTV, gross target volume; BTV, biological target volume; SBRT, Stereotactic Body Radiation Therapy; NSCLC, non-small cell lung cancer; SUV, standardized uptake values; MLC, multi leaf collimator; LCR, local control rates; BED, bioequivalent dose; IMRT, Intensity-modulated radiation therapy; MLD, mean lung dose; TCP, tumour control probability; NTCP, normal tissue complication probability. * Corresponding author. Tel.: +86 531 87984729; fax:+86 531 87984079. E-mail address: [email protected] (J. Yu). 1 Equal contributors (Co-first author). http://dx.doi.org/10.1016/j.canlet.2014.07.042 0304-3835/© 2014 Elsevier Ireland Ltd. All rights reserved.

based on the finding of the RTOG 0617 [8]. It illustrates the need for a more stringent application of the dose escalation as the potential for heightened toxicity of the higher dose. Then limited by tolerances of organs at risk (OAR), the idea of raising the target dose only to the functional subvolumes within the tumour, which are supposed to be more radio-resistant, seems promising and practical [9]. New biological imaging methodologies, mainly based on positron emission tomography/computed tomography (PET/CT), magnetic resonance imaging (MRI) and magnetic resonance spectroscopy (MRS) imaging, in conjunction with radiotherapy (RT), make dose painting possible. It can be used to draw a three-dimensional map of radiobiological relevant parameters as its inherent potential to trace the real target volume – volume consisting of tumour cells that requires a therapeutic dose to control the disease. PET/CT is outstanding and widely used in daily clinical practice. It offers molecular biological information about the tumour microenvironment in addition to anatomical imaging and shows significant biological heterogeneity of tumours, such as metabolism, proliferation, hypoxia, radio-resistance cell density, and perfusion [10]. Dose escalation to the definite radio-resistant subvolumes present on PET imaging makes the burden of the normal tissues constant, or even decreased. Exploration is hot regarding this new field. The new combination will revolutionize the way that RT is prescribed and planned and may improve the therapeutic outcome in terms of LRC with current available clinical data. The article here is meant to review the current status of implementation of biologic heterogeneity that showed in PET/CT imaging in dose planning and to identify whether the new combination can make a practical clinical profit. Also, the

170

X. Shi et al./Cancer Letters 355 (2014) 169–175

key challenges involved in daily clinical application of the new RT strategies will be addressed. Validation of dose painting target volume in PET/CT imaging Firstly, the validation of a dose painting target necessarily requires a complete analysis of the correlation between the patterns of failure and the planning target volume (PTV) on molecular imaging of PET/CT. There are several emerging data analyses on the relationship of the exact anatomical site of local-regional failure (LRF) and the pretreatment PET/CT defined tumour site. In a study of locally-advanced head and neck cancers treated using PET/CT based IMRT planning, the results were very impressive: the gross target volume (GTV)-recurrent was completely encompassed by PTV (70 Gy) in 19/31 cases; 8/31 patients had recurrent disease extending from inside PTV into the low dose region, 3/31 patients failed in the low dose region alone (50.4 Gy) [11]. Then Madani et al. figured out that only 33% (3/9) LRF of head and neck cancer was outside the PET-biological target volume (PET-BTV) [12]. A more encouraging evidence was given by Soto and Dirix. Soto et al. declared that 100% (9/9) of failures were inside the GTV for patients with head and neck cancer [13]. Similarly, Dirix et al. concluded that all localregional recurrences were within the (18) F-FDG-avid regions on baseline (18) F-FDG PET [14]. Most failures after Stereotactic Body Radiation Therapy (SBRT) treatment for recurrent head and neck cancer were partial or complete within the PTV, i.e. In-field (>75% inside PTV) – 12.3%, Overlap (20–75% inside PTV) – 24.6%, Marginal (<20% inside PTV but closest edge within 1 cm of PTV) – 36.8% [15]. There were some similar results shown in the non-small cell lung cancer (NSCLC): the location of residual metabolic-active areas after therapy which indicates worse survival corresponded with the high FDG-avid regions pre-radiotherapy [16]. And in a study of Abramyuk et al., 10 patients of NSCLC who received curative RT (66 Gy) with local-regional relapse afterwards were analysed retrospectively, 6 of 10 recurrences were localized partially or completely in the irradiated target volume with high initial metabolic activities on PET images [17]. In literature, there are various segmentation strategies to define the area with high-affinity tracer uptake and metabolism, including thresholding, region growing, classifiers, clustering, edge detection, Markov random field models, deformable models and many other approaches. But in clinical trials, thresholding is commonly used: standardized uptake values (SUV) > 2.5, 38%, 42%, 47% and 50% of the maximum standardized uptake values (SUVmax) in FDG PET/CT; 80% SUVmax in FLT PET/CT; 60% and 70% SUVmax in 11C-choline PET/CT; SUV 1.4 in FMISO or FLT PET/CT et al. More researches are needed to identify the optimal image segmentation approaches which reproducibly and accurately identify the high recurrent-risk regions to match the specificity of each particular cancer imaging probe and tumour type. Dose painting strategies in PET/CT imaging For a nonuniform radiosensitivity distribution of tumour, a distribution that delivers a relatively higher proportion of the integral tumour dose to the more resistant regions of the tumour (the core of dose painting) seems more logical compared with an uniform dose distribution. Different dose painting strategies to shape the radiation dose according to the functional image information have been proposed recently: dose escalation and dose redistribution. Dose escalation applies an additional dose to the functional subvolumes of the target whereas dose redistribution consists of increasing the dose to the radio-resistant areas while reducing the dose to the rest of the tumour in a way to keep the mean dose constant. The researches now mainly focus on dose escalation.

There are two strategies for the realization of dose escalation: dose painting by contours (DPBC) and dose painting by numbers (DPBN). DPBN intends to increase the additional dose gradually, according to the local PET voxel intensities, while a homogeneous dose of BTV which contoured by a threshold created in PET image is given in DPBC. DPBC creates a boost subvolume within the tumour by a certain threshold. The areas of relatively lower and higher risk for recurrence are set fixed to voxels with a corresponding standardized uptake value (SUV) < threshold and SUV > threshold, respectively. In general, treatment application and also planning for DPBC can be realized by using the simultaneous integrated boost (SIB) technique. Troost et al. thinks that dose escalation to a relatively small subvolume by DPBC can be realized with meeting the limited dose criterion of the surrounding normal tissues and thus might be better tolerated by patients [18]. The strategy that a homogeneous boost dose is assigned to the subvolume has been fulfilled in many clinical trials. DPBN is a method in which a continuously increasing relationship is set between the voxel values of the functional imaging and the risk of local recurrence by using a linear correlation of the boost dose and image intensity of this same voxel. The technical feasibility of DPBN is already shown by many groups [19–21]. Dose prescription with applicable steep gradients can be delivered to the numerous mini-subvolumes by means of a conventional linear accelerator equipped with a high resolution micro multi leaf collimator (MLC) [22]. Rickhey et al. showed that the DPBN approach in brain tumours by 18F-FET-PET was achievable with high accuracy [23]. Alternatively, 18[F]-FDG-PET-guided DPBN using currently available technique was proved to be feasible in phase I clinical trial by Berwouts et al. in head and neck RT [24]. For both DPBC and DPBN plans, strict planning constraints set for the OAR should be complied, in addition, the feasibility of technique is proved by the evidence mentioned above, thus both are therefore considered clinically acceptable and alternatives for boosting subvolumes. Verification of maximum dose and evaluation of dose painting planning With the current evidence, it tends to believe that higher dose prescription leads to lower risk of local failure. Given the fact that a dose range of 1–5 Gy per fraction is suitable in modelling of tumour response by the linear–quadratic equation and the dose–response curves presents a negative correlation between cell survival with dose, the higher dose of subvolumes in dose painting could prominently decrease the amount of tumour cells and thus improve the local control [25]. In the viewpoints of Bradley [6] and Fowler et al. [26], local control rates (LCR) over 90% can be reached when 120 Gy is given to the tumour. Similar opinions have been expressed that the potency of the dose regimen was significantly associated with LCR, patients received bioequivalent dose (BED) < 100 Gy had a 16.7% risk of local failure compared with 2.3% for patients BED >100 Gy, suggesting that high dose regimens (BED >100 Gy) should be adopted because of lower risk for local failure [27]. In a previous study by Shaw et al., it was shown that larger target volume had a potential of being delivered with lesser dose because of normal issues tolerance [28]. Thus, we attempt to push the dose of small subvolumes with acceptable toxicity to resolve the paradox. DPBC results in relatively steeper dose gradients as it is binary, contrary to that, the escalated dose in DPBN takes place within numerous subvolumes inside the target volume according to the level of tracer uptake. The dose gradients at the boundary of the target volume remain almost the same as in conventional IMRT treatment planning. Therefore, the dose to the tracer-avid areas on PET imaging can be raised remarkably at an acceptable normal tissue burden simultaneously [23].

X. Shi et al./Cancer Letters 355 (2014) 169–175

So what is the safe dose with a therapeutic effect to control the disease? For now, arbitrary dose prescriptions are used in most planning and clinical trials with confined dose load of the OARs. Max tolerance dose in DPBC The theoretical dose escalation level was addressed in many planning studies. Witte et al. showed that the dose ranges from 66 to 86 Gy can be reached for one lung cancer patient, 70~85 Gy was reached for one oropharynx cancer patient in the planning study [19]. Choi et al. delivered 30 daily doses of 2.6 Gy (total of 78 Gy) to the hypoxic tumour volumes of the head and neck cancer without increases in normal tissue doses using 18F-FMISO PET/CT-guided IMRT [29]. Also, the safety of the boost plans in subvolumes by 6 Gy to regions with SUV above 2.0 using IMRT for postoperative local recurrent rectal cancer was proved [30]. Furthermore, in head and neck cancer Lee et al. maximally escalated the dose to hypoxic subvolumes to 84 Gy in 10 patients and 105 Gy in 1 patient while sparing normal tissues [31]. In a recently planning study of Moller et al. for 10 NSCLC patients, a dose level of 82 Gy in nine patients was proved feasible [32]. Madani et al. put forward that a boost-dose of 3.0 Gy/fraction within 10 cm3 of the 18F-FDG-PET-avid regions could be safe during the first 2 weeks of treatment on the basis of a phase I clinical trial for head and neck cancer [12]. In a phase II trial, Kong et al. escalated a dose to the residual tumour of locally advanced NSCLC on mid-treatment FDG-PET. The median physical dose reached was 84 Gy (range 63–86 Gy), equivalent to 90 Gy in 2 Gy fractions (biological effective dose 108 Gy) [33]. In the RTOG 1106 (an ongoing phase II randomized clinical trial), the dose per fraction to patients of inoperable NSCLC which varies between 2.2 and 3.8 Gy will be delivered for the mid-treatment residual FDG-PET-avid subvolumes. If the highest dose per fraction (up to 3.8 Gy/Fx) that meets the 95% coverage of the mid-treatment PTV with the total MLD ≤ 20 Gy and acceptable dose of other OARs, it would be adopted in the patient. In another randomized phase II trial for advanced NSCLC, the dose levels of boost region with high FDG uptake (>50% SUVmax) inside the primary tumour were on an average of 86.9 ± 14.9 Gy [34]. As mentioned above, the size of the subvolume is a vital factor in determining the level of the escalated dose, thus, the delineation method used for contouring is crucial to the applicable maximum dose. The optimal delineation method which could reproducibly and accurately identify the high recurrent-risk region needs further investigated. Max tolerance dose in DPBN There are several classic formulations widely used to calculate the maximum tolerance dose in DPBN. The linear correlation of the boost dose (D) and image intensity (I) of the same voxel are implemented in these formulae. The minimum doses are set fixed to all voxels in the target volume with a therapeutic dose used in the current routine clinical practice to avoid underdosing, the Ihigh and Ilow are also fixed in order to alleviate the steepness between the corresponding Dhigh and Dlow. The dose increases with the increasing intensity of the corresponding volex, and thus forms a maximum dose. For instance, in the work performed by Vanderstraeten et al. [22], the boost prescription was created by the formula below:

DI = Dlow +

I − Ilow (Dhigh − Dlow ) Ihigh − Ilow

Ilow = 0.25 × I95%·Ihigh = I95%·Dlow = 69 Gy. Another classic formula which is slightly different was put forth by Flynn et al. [35]:

D ( I ) = DCTV +

171

I ⋅ Dboost Imean

With a ratio of I of a certain volume to Imean, the mean boost dose of the subvolumes was kept equal to a previously fixed uniform boost dose, Dboost. Deveau et al. applies the same formula in his DPBN planning [20]. Several other similar formulae have been proposed [23,36], but the standard mathematical form of the dose prescriptions in DPBN is still unknown, since the formulae are obviously oversimplifications of the more complex multifactorial biological characters. Higher peak-dose may result in steeper dose gradients of the whole target volume according to the formulae mentioned above. Even meeting the confined dose of OARs, the maximum dose is also limited by the maximal steepness of the deliverable dose gradients in the technique. When the dose boosted to the functional volume only to segmental fractions in DPBN, the feasibility of dose summation of each voxel in the whole treatment plan that adopts deformable image coregistration is showed to minimize the anatomical mismatch [37]. Currently most researches use bulk tumour characteristics measured by PET such as maximum or mean SUV or subvolumes contoured by a certain SUV-threshold. It does not give any information to verify the optimal dose which can result in a positive radioresponsiveness at the voxel level. Tumour control probability (TCP) models that integrate quantitative imaging can be used to determine the optimal dose painting prescriptions [38,39] in the sense of providing a maximal TCP. A modest boost (120–150% of the primary dose) was reported to increase TCP [29]. Using biological parameters – TCP and normal tissue complication probability (NTCP) – in the dose painting optimization process appears to be a promising approach towards optimal individualized treatment plans [40,41]. Synthesizing all these evidence, the optimal dose prescription is not verified, the dose from the theoretical research and current clinical trials can only provide an estimation, the definite dose prescription with maximum therapeutic ratio can only be extracted from the sufficient data of clinical controlled trials.

Optimization and evaluation of dose painting planning Generally, the present commercial treatment planning software can easily optimize the planning of DPBC but do not allow the optimization of DPBN. Dirscherl et al. [38] and Bogner et al. [42] recommend their in-house treatment planning software Direct Monte Carlo Optimization (DMCO) to create plans in DPBN. The conventional dose-volume histograms (DVH), D98%, D2% for the boost volume can be used to assess coverage of DPBC, in contrast, DVH is not suitable for evaluating the dose coverage of the target volume as the dose prescription is heterogeneous in DPBN. Consequently, dose difference histogram, Q (quality)-volume histograms and quality factor (QF) is brought up to evaluate plan conformity for DPBN. The dose difference between prescribed and planned at each voxel is calculated and visualized by means of a dose difference histogram [23,38]. The standard deviation and mean value of the dose difference of the whole target volume can be used to identify a plan with good accuracy. The concept of ΔVH which takes geometric uncertainties into account has been put forward by Witte et al. [19], it is similar to dose difference histogram and proves to be applicable in plan evaluation. Evaluation by the Q-volume histograms in which Q was defined as a ratio of the actual D obtained dose ( Dobtained) to the plan dose (Dplan) Q = obtained , and a Dplan 1 quality factor (QF) of entire target volume QF = ∑ Q i − 1 has been n i applied successfully to evaluate plan conformity in DPBN by Deveau

172

X. Shi et al./Cancer Letters 355 (2014) 169–175

et al. [20], Vanderstraeten et al. [22], Madani et al. [43] and Duprez et al. [44]. Perfect dose painting planning would yield a Q = 1 for each voxel, in reality, an obtained dose of each voxel within 5% of the prescription (Q 0.95–1.05) is acceptable [22,43]. Q value of each voxel can be calculated and visualized in a Q-volume histogram (QVH). Timing of integration with dose painting planning There is confusion about when the additional dose should be given: before the radiotherapy or after several fractions of radiotherapy. The choice of the timing of functional imaging that accurately presents a biological feature of tumour is of crucial importance. In 18F-FDG PET/CT images, the relatively bigger subvolumes may make it harder when adding a higher boost pretreatment, but the RT-induced inflammation in treatment phase could be an interference that cannot create actual radiobiological maps and results in the inaccurate volumes boosted subsequently. In a prospective study, Aerts et al. [45] showed that the location of the low and high FDG uptake areas within the tumour for NSCLC remained stable during RT. The mean overlap fraction of FDG-avid area at Day 0 with Days 7 and 14 was 82.8% ± 8.1% and 84.3% ± 7.6%, respectively. Later, in his other two studies [16,46], the post-treatment residual metabolic activity within the tumour largely corresponded with the FDGavid area of the pretreatment scan, the overlap was as high as 70%. These data tend to support that the pretreatment or early phase of the treatment images can potentially be applied in dose painting planning. Duprez et al. created a DPBN plan based on a 18F-FDGPET/CT scan acquired after the eighth fraction of the RT [44]. Madani et al. conducted a phase I clinical trial for head and neck cancer by using 18F-FDG/CT image at pretreatment and 2 week of the treatment [43]. However, Kong et al. demonstrated that the peak tumour FDG-activity decreased significantly after 45 Gy, and found a significant correlation in tumour metabolic response between duringRT scans and 3 months post-RT scans in patients with NSCLC [47]. Then Kong and her colleagues conduct two clinical studies (including the ongoing RTOG 1106) using FDG PET/CT images in late phase (at around 40~50 Gy) of the RT [33]. Gillham et al. also used the FDG PET/CT images performed following 50 or 60 Gy in his study to boost [48]. With regard to 18F-FLT PET/CT images, a strong correlation between FLT uptake and Ki-67 score measured by immunohistochemistry in brain, lung and breast cancer is proved by a metaanalysis recently [49], but direct clinical evidence of the geographic correlation between FLT-avid regions and the subsequent local failure areas is nearly absent. Given the close competition between cell killing, inhibiting and accelerated cell repopulation during the radiotherapy, the reduced FLT uptake shortly after a dose of radiation (mitotic delay) can cause a biased contouring of the proliferation target volume [50] and in late treatment course a rapid reduction of the tumour cell density may render the PET signal intensities too low to contour a subvolume. In the work performed by Yue [51], tumoural SUVmax of FLT-PET/CT and proliferation target volume decreased gradually during the treatment. An almost complete absence of detectable tumour proliferation at 30 Gy and a complete absence at 40 Gy and beyond were found in inoperable locally advanced squamous cell carcinoma of the oesophagus. More recently a study by Troost et al. delivered the boost in the first 4 week by using 18FFLT/CT image at pretreatment and second week of the treatment [18]. Concerning the FMISO-PET/CT imaging, the temporal and spatial stability, such as the time window of complete reoxygenation and the conflict of whether the location of the hypoxic subvolumes varied during the treatment, has long been in concern [52,53]. Hypoxia showed on FMISO-PET/CT before radiotherapy in glioblastoma

multiforme was strongly associated with poorer TTP (time to progression) and survival [54]. The same opinion was presented by Schutze et al. in their tumour models and they supported doseescalation with inhomogeneous dose distribution based on pretreatment [18F]FMISO PET/CT scans [55]. Chang et al. [56], Choi et al. [29] and Lee et al. [31] used FMISO-PET/CT imaging before treatment to create dose-escalation plans. Zips et al. thought that FMISOPET/CT imaging at 1 or 2 weeks during RT could be a promising way to dose-escalated treatment [57]. Hendrickson et al. created DPBC plans by FMISO-PET imaging obtained before therapy for 10 patients who exhibited significant hypoxia [58]. It seems that dose painting to the hypoxic areas (acute hypoxia) may be more applicable with repeated PET-scans during treatment. Lin et al. designed a DPBC plan which boost a dose of 14 Gy to the hypoxic volume on the basis of twice FMISO scan imaging in RT [59]. To date, no clinical studies have been published about dose escalation to radioresistent hypoxic subvolumes confined by PET imaging parameters. Progress of new strategies in clinical application We assumed that an increased precision of target subvolume localization by PET/CT information plus a desired therapeutic dose that delivered by dose painting may have an additional effect to improve local control and further reduce the potential of tumour recurrence. In theory, dose painting with PET imaging is verified to yield significant TCP gains with acceptable toxicity [38,40,60]. Primary prognostic outcomes of the clinically application of PET imaging and dose painting in the combined modality have been demonstrated recently (Table 1). Does dose painting ultimately lead to higher cure rates? According to our best knowledge, the answer tends to be positive. Several clinical trials present prognostic outcome of patients with the administration of DPBC successfully. Madani et al. conducted a phase I clinical trial by PET-guided dose escalation, showed that actuarial 1-year local control was 85% and 87%, and 1-year OS was 82% and 54% (P = 0.06), at dose level of 72.5 and 77.5 Gy, respectively. He explained some negative prognostic factors of patient characters may contribute to the worse survival at dose level of 77.5 Gy [12]. In a PET-plan pilot trial [61] of locally advanced NSCLC with the prescribed total doses in range of 66.6–73.8 Gy, the estimated median survival time was 19.3 months after a median followup time of 27.2 months. It confirmed a low risk of out-of-field isolated nodal recurrences and an improved tumour control. An ongoing randomized controlled phase II trial for head and neck cancer – Adaptive and innovative Radiation Treatment FOR improving Cancer treatment outcomE (ARTFORCE) was designed by Heukelom et al. [62]. It aimed to detect a 15% improvement in LRC with a power of 80% at a significance level of 0.05. Kong et al. have risen the level of the dose prescription to the residual tumour on mid-treatment (at 40~50 Gy of the treatment) in NSCLC with an improvement of 2-year LRC. The 2-year rates of in-field LRC, overall LRC were 84% (63%–94%), 68% (47%–82%), compared with the stagematched patients with standard-dose RT, the OS is also astonishing – 51% (34%–65%) [33]. An ongoing phase II randomized clinical trial by Kong et al. (RTOG1106) is proposed to verify the improvement of LRC further. Another randomized phase II trial of FDG PET Boost in Stage IB, II and III NSCLC is intended to observe the local progression-free survival (LPFS) at 1 year (NCT01024829). The novel strategy which PET/CT imaging combined with DPBN was applied in the following trials recently. In a phase I clinical trial by Berwouts et al., a median prescription dose of 70.2 Gy to the gross tumour volume (GTV) consists of three DPBN plans by using PET/ CT imaging at pretreatment and after 8 and 18 fractions only caused a mild acute toxicity without a decrease in tumour response at 3 months post-treatment assessment in a small study populations, 9 of 10 patients did not have evidence of disease at a median

X. Shi et al./Cancer Letters 355 (2014) 169–175

173

Table 1 Clinical progress of PET/CT guided dose painting.

DPBC

DPBN

Authors

Status

Phase

Tumor type

Level of dose escalation

Conclusion

Madani et al.2007 [12] Fleckenstein et al. 2011 [61] Kong et al. 2013 [33] Heukelom et al. 2013 [62]

Completed

I

N. 39

Head and neck cancer

72.5 and 77.5 Gy

Completed

Pilot trial

32

Range of 66.6–73.8 Gy

Completed

II

42

Locally advanced NSCLC NSCLC

Actuarial 1-year local control was 85% and 87%, and 1-year OS was 82% and 54% (P = 0.06) The estimated median survival time was 19.3 months

Ongoing

II

268

Head and neck cancer

Kong et al. (RTOG1106) van Elmpt et al. [34] (NCT01024829)

Ongoing

II

NR

NSCLC

Ongoing

II

NR

NSCLC

Berwouts et al. 2013 [24]

Completed

I

10

Head and neck cancer

Madani et al. 2011 [43]

Completed

I

21

Non-metastatic head and neck cancer

Median physical dose 84 Gy (BED108 Gy) Boost region 77 Gy, PTV outside the Boost region 67 Gy Total dose of 80.4 Gy Mean total dose 77.3 ± 7.9 Gy (arm A) and 77.5 ± 10.1 Gy (arm B). Boost region 86.9 ± 14.9 Gy A median prescription dose of GTV 70.2 Gy (68.7 ± 2.6, 80.7 ± 1.2) Median total dose of 80.9 and 85.9 Gy

The 2-year rates of in-field LRC, overall LRC were 84% and 68%, the OS was51% (34%–65%) It aimed to detect a 15% improvement in LRC with a power of 80% at a significance level of 0.05. NR NR

Disease control in 9/10 patients at a median follow-up of 13 months An actuarial 2-year LRC and freedom from distant metastasis were 95%, 93% and 68%, respectively

Abbreviations: DPBC, dose painting by contours; DPBN, dose painting by numbers; N, numbers of patients; OS, overall survival; LRC, local-regional control; NSCLC, non-small cell lung cancer; NR, not reported.

follow-up of 13, range 7–22 months [24]. In another phase I trial for non-metastatic head and neck cancer, a median total dose of 80.9 Gy in the high-dose clinical target volume in 7 patients and 85.9 Gy in the gross tumour volume of 14 patients came into reality, an actuarial 2-year local and regional control and freedom from distant metastasis in all patients were 95%, 93% and 68%, respectively [43]. Conclusion In this study, we showed an explicit review on the current state of PET/CT imaging-guided dose painting in RT. The incorporation of PET/CT imaging into RT has the potential to allow dose painting with tolerant toxicity of normal tissue and to control the disease in a therapeutic dose. The data of clinical studies on the integration of PET/CT in dose painting for lung and head-and-neck cancers are presented above. Systematic experimental and clinical trials are necessary to validate the results of planning studies on brain, rectal tumours and prostate cancer. For other entities like gastrointestinal cancer, lymphomas, sarcomas, etc., the data of the feasibility of the new paradigm are yet scarce. Dose painting with functional MRI which allows characterization of the biological behaviour of glioblastoma and prostate tumour tissue is a new exciting area and under hot investigation, the primary planning studies are demonstrated in prostate cancer [63]. Functional MRI characterized by exquisite soft tissue contrast can provide valuable information for further design of promising dose strategies in dose painting. Although encouraging primary clinical data have presented a promising prospect, there are many pending issues which may challenge the involvement of the dose painting with PET/CT imaging in daily clinical application. 1. Since dose painting is on the most fundamental assumption that voxels with a higher PET intensity require a higher dose, the validated, robust evidence of the correlation between PET imaging of various tracers and real biological feature has not given whereas to date. 2. The consistent positioning of the patient in the consecutive image acquisition and whole RT treatment procedures needs to be more stringently guaranteed, especially in DPBN as the dose differs dramatically between voxels.

3. Some inherent limitations such as coarse spatial resolution and the related partial volume effects and false positive/negative readings affect the reliability of PET imaging [64]. Standardization in methodological problems like the choice of segmentation and reconstruction algorithms remains to be created in order to guarantee high levels of reproducibility. 4. There are numerous studies that showed functional image parameters of pretreatment [65–67] and during treatment [68] correlate to the prognostic outcome, identification of patients who would most benefit from dose escalation need a comprehensive analysis of patient data in the following clinical trials. 5. Assessment of cost effectiveness of this new combined RT of (multiple) re-imaging, re-delineation and re-planning that is urgently needed.More randomized controlled trials are expected to be presented to provide more robust evidence of the effectiveness of dose painting in disease control and the toxicity rates against standard RT strategy. The application is not ready for routine clinical practice for now.

Conflict of interest The authors have no potential conflicts of interest relevant to the content of this manuscript.

Acknowledgments This work was supported by The National Science Foundation of China (81201827) and Wu Jieping Medicine Foundation (320.6750.12243).

References [1] M. Machtay, K. Bae, B. Movsas, et al., Higher biologically effective dose of radiotherapy is associated with improved outcomes for locally advanced non-small cell lung carcinoma treated with chemoradiation: an analysis of the Radiation Therapy Oncology Group, Int. J. Radiat. Oncol. Biol. Phys. 82 (1) (2012) 425–434. [2] M. Machtay, R. Paulus, J. Moughan, et al., Defining local-regional control and its importance in locally advanced non-small cell lung carcinoma: a radiation therapy oncology group analysis, J. Thorac. Oncol. 7 (4) (2012) 716–722.

174

X. Shi et al./Cancer Letters 355 (2014) 169–175

[3] C. Schaake-Koning, W. van den Bogaert, O. Dalesio, et al., Radiosensitization by cytotoxic drugs. The EORTC experience by the Radiotherapy and Lung Cancer Cooperative Groups, Lung Cancer 10 (Suppl. 1) (1994) S263–S270. [4] A. Auperin, P.C. Le, E. Rolland, et al., Meta-analysis of concomitant versus sequential radiochemotherapy in locally advanced non-small-cell lung cancer, J. Clin. Oncol. 28 (13) (2010) 2181–2190. [5] F.M. Kong, H.R.K. Ten, M.J. Schipper, et al., High-dose radiation improved local tumor control and overall survival in patients with inoperable/unresectable non-small-cell lung cancer: long-term results of a radiation dose escalation study, Int. J. Radiat. Oncol. Biol. Phys. 63 (2) (2005) 324–333. [6] J. Bradley, A review of radiation dose escalation trials for non-small cell lung cancer within the Radiation Therapy Oncology Group, Semin. Oncol. 32 (2 Suppl. 3) (2005) S111–S113. [7] H.J. Aerts, P. Lambin, D.D. Ruysscher, FDG for dose painting: a rational choice, Radiother. Oncol. 97 (2) (2010) 163–164. [8] J.D. Cox, Are the results of RTOG 0617 mysterious, Int. J. Radiat. Oncol. Biol. Phys. 82 (3) (2012) 1042–1044. [9] C.C. Ling, J. Humm, S. Larson, et al., Towards multidimensional radiotherapy (MD-CRT): biological imaging and biological conformality, Int. J. Radiat. Oncol. Biol. Phys. 47 (3) (2000) 551–560. [10] S.K. Das, H.R.K. Ten, Functional and molecular image guidance in radiotherapy treatment planning optimization, Semin. Radiat. Oncol. 21 (2) (2011) 111–118. [11] J.M. Glanzman, D.M. Mejía, W.F. Mourad, et al., Patterns of local failure in locally-advanced head-and-neck cancers treated using PET-CT-based IMRT planning: does area of recurrence correlate with preradiation PET?, Int. J. Radiat. Oncol. Biol. Phys. 87 (2) (2013) S476. [12] I. Madani, W. Duthoy, C. Derie, et al., Positron emission tomography-guided, focal-dose escalation using intensity-modulated radiotherapy for head and neck cancer, Int. J. Radiat. Oncol. Biol. Phys. 68 (1) (2007) 126–135. [13] D.E. Soto, M.L. Kessler, M. Piert, A. Eisbruch, Correlation between pretreatment FDG-PET biological target volume and anatomical location of failure after radiation therapy for head and neck cancers, Radiother. Oncol. 89 (1) (2008) 13–18. [14] P. Dirix, V. Vandecaveye, F. De Keyzer, et al., Dose painting in radiotherapy for head and neck squamous cell carcinoma: value of repeated functional imaging with (18)F-FDG PET, (18)F-fluoromisonidazole PET, diffusion-weighted MRI, and dynamic contrast-enhanced MRI, J. Nucl. Med. 50 (7) (2009) 1020–1027. [15] K. Wang, D.E. Heron, J.C. Flickinger, et al., A retrospective, deformable registration analysis of the impact of PET-CT planning on patterns of failure in stereotactic body radiation therapy for recurrent head and neck cancer, Head Neck Oncol. 4 (2012) 12. [16] H.J. Aerts, A.A. van Baardwijk, S.F. Petit, et al., Identification of residual metabolic-active areas within individual NSCLC tumours using a preradiotherapy (18)Fluorodeoxyglucose-PET-CT scan, Radiother. Oncol. 91 (3) (2009) 386–392. [17] A. Abramyuk, S. Tokalov, K. Zophel, et al., Is pre-therapeutical FDG-PET/CT capable to detect high risk tumor subvolumes responsible for local failure in non-small cell lung cancer, Radiother. Oncol. 91 (3) (2009) 399–404. [18] E.G. Troost, J. Bussink, A.L. Hoffmann, et al., 18F-FLT PET/CT for early response monitoring and dose escalation in oropharyngeal tumors, J. Nucl. Med. 51 (6) (2010) 866–874. [19] M. Witte, G. Shakirin, A. Houweling, et al., Dealing with geometric uncertainties in dose painting by numbers: introducing the DeltaVH, Radiother. Oncol. 100 (3) (2011) 402–406. [20] M.A. Deveau, S.R. Bowen, D.C. Westerly, et al., Feasibility and sensitivity study of helical tomotherapy for dose painting plans, Acta Oncol. (Madr) 49 (7) (2010) 991–996. [21] D. Thorwarth, M. Alber, Implementation of hypoxia imaging into treatment planning and delivery, Radiother. Oncol. 97 (2) (2010) 172–175. [22] B. Vanderstraeten, W. Duthoy, W. De Gersem, et al., [18F]fluoro-deoxy-glucose positron emission tomography ([18F]FDG-PET) voxel intensity-based intensitymodulated radiation therapy (IMRT) for head and neck cancer, Radiother. Oncol. 79 (3) (2006) 249–258. [23] M. Rickhey, O. Koelbl, C. Eilles, et al., A biologically adapted dose-escalation approach, demonstrated for 18F-FET-PET in brain tumors, Strahlenther. Onkol. 184 (10) (2008) 536–542. [24] D. Berwouts, L.A. Olteanu, F. Duprez, et al., Three-phase adaptive dose-paintingby-numbers for head-and-neck cancer: initial results of the phase I clinical trial, Radiother. Oncol. 107 (3) (2013) 310–316. [25] J.M. Brown, D.J. Carlson, D.J. Brenner, The tumor radiobiology of SRS and SBRT: are more than the 5 Rs involved, Int. J. Radiat. Oncol. Biol. Phys. 88 (2) (2014) 254–262. [26] J.F. Fowler, W.A. Tome, J.D. Fenwick, M.P. Mehta, A challenge to traditional radiation oncology, Int. J. Radiat. Oncol. Biol. Phys. 60 (4) (2004) 1241–1256. [27] Z. Allibhai, M. Taremi, A. Bezjak, et al., The impact of tumor size on outcomes after stereotactic body radiation therapy for medically inoperable early-stage non-small cell lung cancer, Int. J. Radiat. Oncol. Biol. Phys. 87 (5) (2013) 1064–1070. [28] E. Shaw, C. Scott, L. Souhami, et al., Single dose radiosurgical treatment of recurrent previously irradiated primary brain tumors and brain metastases: final report of RTOG protocol 90-05, Int. J. Radiat. Oncol. Biol. Phys. 47 (2) (2000) 291–298. [29] W. Choi, S.W. Lee, S.H. Park, et al., Planning study for available dose of hypoxic tumor volume using fluorine-18-labeled fluoromisonidazole positron emission tomography for treatment of the head and neck cancer, Radiother. Oncol. 97 (2) (2010) 176–182.

[30] K. Jingu, H. Ariga, T. Kaneta, et al., Focal dose escalation using FDG-PET-guided intensity-modulated radiation therapy boost for postoperative local recurrent rectal cancer: a planning study with comparison of DVH and NTCP, BMC Cancer 10 (2010) 127. [31] N.Y. Lee, J.G. Mechalakos, S. Nehmeh, et al., Fluorine-18-labeled fluoromisonidazole positron emission and computed tomography-guided intensity-modulated radiotherapy for head and neck cancer: a feasibility study, Int. J. Radiat. Oncol. Biol. Phys. 70 (1) (2008) 2–13. [32] D.S. Moller, A.A. Khalil, M.M. Knap, et al., A planning study of radiotherapy dose escalation of PET-active tumour volumes in non-small cell lung cancer patients, Acta Oncol. 50 (6) (2011) 883–888. [33] F.M. Kong, R. Haken, M.J. Schipper, et al., A phase II trial of mid-treatment FDG-PET adaptive, individualized radiation therapy plus concurrent chemotherapy in patients with non-small cell lung cancer (NSCLC), J. Clin. Oncol. 31 (Suppl.) (2013) abstr 7522. [34] W. van Elmpt, D. De Ruysscher, A. van der Salm, et al., The PET-boost randomised phase II dose-escalation trial in non-small cell lung cancer, Radiother. Oncol. 104 (1) (2012) 67–71. [35] R. Flynn, D. Barbee, S. Bowen, et al., Dose painting with intensity modulated proton therapy and intensity modulated x-ray therapy: a comparison, Med. Phys. 34 (2007) 2522–2523. [36] G. Meijer, J. Steenhuijsen, M. Bal, et al., Dose painting by contours versus dose painting by numbers for stage II/III lung cancer: practical implications of using a broad or sharp brush, Radiother. Oncol. 100 (3) (2011) 396–401. [37] L.A. Olteanu, I. Madani, W. De Neve, et al., Evaluation of deformable image coregistration in adaptive dose painting by numbers for head-and-neck cancer, Int. J. Radiat. Oncol. Biol. Phys. 83 (2) (2012) 696–703. [38] T. Dirscherl, M. Rickhey, L. Bogner, Feasibility of TCP-based dose painting by numbers applied to a prostate case with (18)F-choline PET imaging, Z. Med. Phys. 22 (1) (2012) 48–57. [39] Y. Kim, W.A. Tome, On the impact of functional imaging accuracy on selective boosting IMRT, Phys. Med. 25 (1) (2009) 12–24. [40] Y. Kim, W.A. Tome, Dose-painting IMRT optimization using biological parameters, Acta Oncol. 49 (8) (2010) 1374–1384. [41] J.H. Chang, J.D. Lim, S.T. Lee, et al., Intensity modulated radiation therapy dose painting for localized prostate cancer using (1)(1)C-choline positron emission tomography scans, Int. J. Radiat. Oncol. Biol. Phys. 83 (5) (2012) e691–e696. [42] L. Bogner, M. Alt, T. Dirscherl, et al., Fast direct Monte Carlo optimization using the inverse kernel approach, Phys. Med. Biol. 54 (13) (2009) 4051–4067. [43] I. Madani, F. Duprez, T. Boterberg, et al., Maximum tolerated dose in a phase I trial on adaptive dose painting by numbers for head and neck cancer, Radiother. Oncol. 101 (3) (2011) 351–355. [44] F. Duprez, W. De Neve, W. De Gersem, et al., Adaptive dose painting by numbers for head-and-neck cancer, Int. J. Radiat. Oncol. Biol. Phys. 80 (4) (2011) 1045–1055. [45] H.J. Aerts, G. Bosmans, A.A. van Baardwijk, et al., Stability of 18F-deoxyglucose uptake locations within tumor during radiotherapy for NSCLC: a prospective study, Int. J. Radiat. Oncol. Biol. Phys. 71 (5) (2008) 1402–1407. [46] H.J. Aerts, J. Bussink, W.J. Oyen, et al., Identification of residual metabolic-active areas within NSCLC tumours using a pre-radiotherapy FDG-PET-CT scan: a prospective validation, Lung Cancer 75 (1) (2012) 73–76. [47] F.M. Kong, K.A. Frey, L.E. Quint, et al., A pilot study of [18F]fluorodeoxyglucose positron emission tomography scans during and after radiation-based therapy in patients with non small-cell lung cancer, J. Clin. Oncol. 25 (21) (2007) 3116–3123. [48] C. Gillham, D. Zips, F. Ponisch, et al., Additional PET/CT in week 5–6 of radiotherapy for patients with stage III non-small cell lung cancer as a means of dose escalation planning, Radiother. Oncol. 88 (3) (2008) 335–341. [49] A. Chalkidou, D.B. Landau, E.W. Odell, et al., Correlation between Ki-67 immunohistochemistry and 18F-fluorothymidine uptake in patients with cancer: a systematic review and meta-analysis, Eur. J. Cancer 48 (18) (2012) 3499–3513. [50] S.M. Bentzen, V. Gregoire, Molecular imaging-based dose painting: a novel paradigm for radiation therapy prescription, Semin. Radiat. Oncol. 21 (2) (2011) 101–110. [51] J. Yue, L. Chen, A.R. Cabrera, et al., Measuring tumor cell proliferation with 18F-FLT PET during radiotherapy of esophageal squamous cell carcinoma: a pilot clinical study, J. Nucl. Med. 51 (4) (2010) 528–534. [52] M.I. Bittner, N. Wiedenmann, S. Bucher, et al., Exploratory geographical analysis of hypoxic subvolumes using (18)F-MISO-PET imaging in patients with head and neck cancer in the course of primary chemoradiotherapy, Radiother. Oncol. 108 (3) (2013) 511–516. [53] W.J. Koh, K.S. Bergman, J.S. Rasey, et al., Evaluation of oxygenation status during fractionated radiotherapy in human nonsmall cell lung cancers using [F18]fluoromisonidazole positron emission tomography, Int. J. Radiat. Oncol. Biol. Phys. 33 (2) (1995) 391–398. [54] A.M. Spence, M. Muzi, K.R. Swanson, et al., Regional hypoxia in glioblastoma multiforme quantified with [18F]fluoromisonidazole positron emission tomography before radiotherapy: correlation with time to progression and survival, Clin. Cancer Res. 14 (9) (2008) 2623–2630. [55] C. Schutze, R. Bergmann, K. Bruchner, et al., Effect of [(18)F]FMISO stratified dose-escalation on local control in FaDu hSCC in nude mice, Radiother. Oncol. 111 (1) (2014) 81–87. [56] J.H. Chang, M. Wada, N.J. Anderson, et al., Hypoxia-targeted radiotherapy dose painting for head and neck cancer using (18)F-FMISO PET: a biological modeling study, Acta Oncol. 52 (8) (2013) 1723–1729.

X. Shi et al./Cancer Letters 355 (2014) 169–175

[57] D. Zips, K. Zophel, N. Abolmaali, et al., Exploratory prospective trial of hypoxiaspecific PET imaging during radiochemotherapy in patients with locally advanced head-and-neck cancer, Radiother. Oncol. 105 (1) (2012) 21–28. [58] K. Hendrickson, M. Phillips, W. Smith, et al., Hypoxia imaging with [F-18] FMISO-PET in head and neck cancer: potential for guiding intensity modulated radiation therapy in overcoming hypoxia-induced treatment resistance, Radiother. Oncol. 101 (3) (2011) 369–375. [59] Z. Lin, J. Mechalakos, S. Nehmeh, et al., The influence of changes in tumor hypoxia on dose-painting treatment plans based on 18F-FMISO positron emission tomography, Int. J. Radiat. Oncol. Biol. Phys. 70 (4) (2008) 1219–1228. [60] S. Vinjamuri, J. Uzan, S. Riley, et al., 18Fluoro-ethyl choline PET CT for radiobiologically optimised prostate dose-painting (BIOPROP), Radiother. Oncol. 103 (Suppl. 1) (2012) S330–S331. [61] J. Fleckenstein, D. Hellwig, S. Kremp, et al., F-18-FDG-PET confined radiotherapy of locally advanced NSCLC with concomitant chemotherapy: results of the PET-PLAN pilot trial, Int. J. Radiat. Oncol. Biol. Phys. 81 (4) (2011) e283–e289. [62] J. Heukelom, O. Hamming, H. Bartelink, et al., Adaptive and innovative Radiation Treatment FOR improving Cancer treatment outcomE (ARTFORCE); a randomized controlled phase II trial for individualized treatment of head and neck cancer, BMC Cancer 13 (2013) 84.

175

[63] S.F. Riches, G.S. Payne, N.M. Desouza, et al., Effect on therapeutic ratio of planning a boosted radiotherapy dose to the dominant intraprostatic tumour lesion within the prostate based on multifunctional MR parameters, Br. J. Radiol. 87 (1037) (2014) 20130813. [64] W.A. Weber, Quantitative analysis of PET studies, Radiother. Oncol. 96 (3) (2010) 308–310. [65] J.A. Abelson, J.D. Murphy, N. Trakul, et al., Metabolic imaging metrics correlate with survival in early stage lung cancer treated with stereotactic ablative radiotherapy, Lung Cancer 78 (3) (2012) 219–224. [66] D.S. Lee, Y.S. Kim, Y. IeR, et al., Long-term clinical experience of high-dose ablative lung radiotherapy: high pre-treatment [18F]fluorodeoxyglucosepositron emission tomography maximal standardized uptake value of the primary tumor adversely affects treatment outcome, Lung Cancer 80 (2) (2013) 172–178. [67] C. Tang, J.D. Murphy, B. Khong, et al., Validation that metabolic tumor volume predicts outcome in head-and-neck cancer, Int. J. Radiat. Oncol. Biol. Phys. 83 (5) (2012) 1514–1520. [68] C. Schiepers, M. Dahlbom, W. Chen, et al., Kinetics of 3’-deoxy-3’-18Ffluorothymidine during treatment monitoring of recurrent high-grade glioma, J. Nucl. Med. 51 (5) (2010) 720–727.