Neo-adjuvant chemo-radiation of rectal cancer with Volumetric Modulated Arc Therapy: summary of technical and dosimetric features and early clinical experience

Pre-operative chemo-radiotherapy of rectal cancer in locally advanced stage has become a widely accepted treatment modality as reported by Roh et al [1] or by Gollins et al [2] and references there-in. Locally advanced rectal cancer treated with neoajuvant chemoradiation therapy is expected to show positive response with tumour downstaging in about 45%-47% of patients [3, 4]. Although no effective method has been identified so far to predict outcome from molecular biomarkers or other methodology, research is actively performed to identified clinically valuable predictors (e.g. serum carinoembryonic antigen was reported to be predictor of pathologic tumour response by Yoon et al. [4]). Hystopatological downstaging was also reported to be poorly correlated with tumor volume reduction as measured with magnetic resonance imaging after treatment [5]. The management of advanced rectal cancer is generally approached with two different radiotherapy scheduling, with a short or a long course for preoperative treatments [6]. Longer courses have been more frequently adopted for advanced stages and, although presenting, on the tolerance side, a higher rate of reversible acute toxicity, these schemes showed lower rate of late gastrointestinal toxicity. A Phase I trial with hypofractionated intensity-modulated radiotherapy and simultaneous boost in association with capecitabine chemotherapy was reported as unacceptably toxic and was interrupted [7] questioning at the same time the role of advanced treatment modalities like intensity modulation and the application of alterated fractionations to locally advanced rectal cancer.

In summary, the current challenge is to improve outcomes whereas minimizing morbidity and maximizing the potential for more complete surgery resection and sphincter preservation.

Aim of the present study is to report the technical and dosimetric aspects of the treatments as well as to summarize early acute toxicity findings and histo-pathological results after surgery. The introduction of an advanced technology as volumetric modulated arc therapy in clinical practice follows the hypothesis that this shall lead to dosimetric and clinical results not inferior and possibly superior to previous treatment modalities. For this reason, to prove the hypothesis at a basic level, as a first approach to the new treatment modality, the first group of patients treated with this new modality at our institute, was planned with the same planning objectives of the previously adopted conformal modality, in order to demonstrate at least the clinical equivalence of the new approach with conformal data, and to possibly give information about the potential improvements on toxicity rates or on logistic side. To enable a qualitative comparison of the treatment features and early outcome, a group of patients with similar characteristics but irradiated with a 3D conformal technique, was pooled out of the institutional database and analysed in parallel. A similar investigaton was performed at planning level by Engels et al [8] comparing Helical Tomotherapy and 3D-conformal plans showing that the combination of helical tomotherapy and in their case the usage of daily adaped margins based on MV-CT imaging significantly decreases the probability of gastro-intestinal toxicity. Consistently, a study from the William Beaumont Hospital group [9] showed a clear and highly significant dose-volume relationship between bowel irradiation and acute grade 3 diarrhoea suggesting the need of reducing as much as possible the involvement of the relatively distant organ at risk during pre-operative irradiation of rectal cancer. Similar results were reported by Tho et al. [10] from University of Glasgow showing a strong dose-volume relationship with acute diarrhoea at all dose levels. These results show that, although 3D conformal treatments are technically adequate for pre-operative rectal cancer patients, the application of advanced techniques shall be carefully explored being potentially beneficial for the patients.

The Volumetric Modulated Arc Therapy technique adopted for this study is RapidArc^® (RA), the technical solution realised by Varian Medical Systems (Palo Alto, California) and based on the original investigation of K. Otto [11].

RapidArc^® was recently introduced in clinical practice in several institutes after an intensive validation at planning level where it was compared to IMRT or other approaches, in a series of studies on brain tumours, prostate, head and neck, anal canal, cervix uteri and other indications [12‐18]. At our institute, as per end of September 2009, 160 patients have been treated with RA for a variety of indications. Among these, twenty-five received RA treatment as part of their multidisciplinary management of rectal adenocarcinoma, most of them in pre-operative neo-adjuvant chemo-radiation setting. RA is implemented at planning level by means of the Progressive Resolution Optimisation (PRO) algorithm in the Eclipse planning system by Varian. The optimisation process is based on an iterative inverse planning process aiming to simultaneously optimise the instantaneous multi leaf collimator (MLC) positions, the dose rate, and the gantry rotation speed to achieve the desired dose distribution. Delivery is performed on Varian Clinacs as single or multiple arcs with dynamic continuous modulation of MLC shapes, gantry speed and dose rate.

Figure 1 shows examples of dose distributions for two patients treated with RA or with 3DC plans. Colourwash is in the interval from 10 to 47 Gy. PTV, femurs and bladder are outlined as solid lines in the images. Figures 2 and 3 report the average dose volume histograms of the two groups of patients for PTV, bladder, bowels, femurs and healthy tissue. Dashed lines represent the inter-patient variability at one standard deviation.

Table 2 summarises the technical features of the treatment characteristics. Table 3 reports results of the DVH analysis for the primary course of 44/45 Gy. Tables 4 and 5 record the clinical data of the treatments as early acute reactions and surgical outcome whenever available.

From technical features, it is evident how pure beam on time for RA is slightly longer than the corresponding quantity for 3DC but this difference is compensated by the usage of single arc against the usage of multiple static fields with wedges (normally mechanical being oriented in left-right direction and no collimator rotation is applied to keep MLC leaves in the needed direction to generate conformal shapes). This means that a 3DC plan requires a substantially longer time to be delivered. Higher monitor units were observed for 3DC plans because of the usage of wedges. Fixed dose rate of 300 MU/min was selected for 3DC treatments from institutional policy for conventional treatments, while the effective average dose rate for RA plans resulted of ~200 MU/min, associated to a constant maximum gantry speed rotation (dose rate does not affect delivery time for RapidArc as mentioned in the methods). Field size, defined by jaws, resulted larger for RA but, considering the effective aperture size of the continuously adapting control points (elementary apertures), the situation is reversed with RA resulting in smaller average apertures compared to 3DC.

Pre-treatment quality assurances of RA plans resulted in an average gamma agreement index GAI 3% superior to the acceptance threshold of 95% (range: 95.4 - 99.9%), set as references in our institute.

Dosimetric data showed that RA offered a significant improvement in treatment conformality (CI_90% and CI_95%) with a trend to significance (p < 0.10) for homogeneity (D_5-95%) and for V_107%. No other significant differences between RA and 3DC have been observed. PTV volumes were equivalent between the two groups. Concerning femurs, RA showed a significant reduction in both mean and maximum dose. Bladder data, although a significant difference in the organ's volume was observed (not correlated to any change in patient preparation protocols), no statistically significant differences were observed, with RA delivering low doses to a larger relative volume than 3DC. For bowels, RA presented a systematic additional sparing over the entire dose range but statistically significant only for the maximum significant dose D_1%.

Significant differences were instead observed for Healthy Tissue with RA showing a lower integral and mean dose.

Concerning clinical outcome, only early results are available. Total treatment duration (including week ends and holidays) resulted in a similar number of days with a wider span for 3DC due to some unscheduled interruptions prolonging the course of treatment for this group. The clinical data showed (Table 4) that, after surgery, a T down-staging was observed in 7 patients over 17 with RA. For 3DC the T down-staging was observed in 5 patients over 19. N down-staging was observed in 2/17 patients in the RA group and 4/19 in the 3DC group of patients. Some up-staging for both T and N were observed in both RA and 3DC patients. Distant metastasis progression was observed in one patient in the RA group and in no patients in the 3DC group.

Chemotherapy is one of the most challenging components of the treatment scheme and was interrupted in a similar proportion of patients between the two groups (12% and 15% respectively) for toxicity. Similar patterns of acute toxicity were observed, as expected, between the two groups. All patients in the 3DC group received chemotherapy while two patients did not received any chemotherapy in the RA group (8%, both post-operative patients).

Concerning quantitative results on acute toxicity (table 5), about 50% of patients showed diarrhoea up to grade 3 (G3 in the 8% of RA patients and 5% in the 3DC patients). 28% of RA patients manifested local erithema of grade 1 or 2, 15% in the 3DC group of patients; no grade 3 erithema or higher were observed. Two patients in the 3DC group developed disuria/incontinency, none for RA.

Based on the results of an intensive program of pre-clinical investigations performed at planning level [12‐18] to assess its reliability and potential benefit, RA, a Volumetric Modulated Arc Therapy, was introduced in clinical practice since September 2008 at our institute for a variety of indications. The present study reports about the early findings from the treatment of a group of 25 patients affected by advanced rectal adenocarcinoma irradiated with RA. Most of the patients received concomitant chemotherapy with capecitabine in a pre-operative neo-adjuvant scheme.

The main objective of the first phase of clinical introduction of RA is the assessment of the possibility to administer to patients treatments not inferior and possibly superior to previously adopted conventional modalities. These results should be achieved without introducing elements of potential confusion like alterations of the fractionation schemes (acceleration or hypo-fractionation for example) or like the attempt to maximise dosimetric performances (e.g. enhanced protection of organs at risk). Further studies will assess the elements of improvement once the safety of the new approach is consolidated in routine practice.

Under this perspective, RA showed, according to the here presented data, the capability to reliably reproduce the dosimetric quality of conventional conformal plans, previously used as the standard of treatment for this class of patients, with some already observable improvement in i) conformality of treatments, ii) reduction of hot spots inside target volume, iii) reduction of distant organs at risk involvement like femurs or bowels, iv) global reduction of healthy tissue involvement. Although most of the trends observed and reported in the tables are minor, not statistically significant and likely not clinically relevant, the three points mentioned above (statistically significant) suggest how the advanced technique with volumetric modulation with arcs might be beneficial and confirm general findings from in-silico investigations [12‐18]. The dosimetric improvement achieved with RA compared to 3DC is quite obvious and might be similar to what achievable with other IMRT approaches. In our institute the selection of RA instead of, e.g., fixed gantry IMRT was based also on logistic issues as discussed below.

These results enabled the activation of a second phase, aiming to push RA towards improved sparing of organs at risk, particularly the bladder and the bowels, as already potentially proven in planning investigation for cervix uteri treatments [13].

Having achieved the same quality of treatments of previously adopted techniques, RA confirmed also some advantages at a logistical level. It allowed a significant reduction of effective treatment time, defined as the time needed to deliver a single fraction with the exclusion of time needed to position the patients and to acquire data for image guidance. The measured treatment time reduction with RA was about 40%. If the absolute benefit is a reduction of about 1.4 minutes which seems to be limited, it shall be noted that this was obtained from a comparison against a 3DC technique rather than IMRT with fixed gantry fields. IMRT might allow abetter dosimetric quality but at the expense of prolonged treatment time. In addition, RA allows avoidance of usage of wedges. Mechanical wedges were used in our institute instead of dynamic wedges to allow useful orientation of them without conflicting with MLC leaves movement, but this required manual operation by radiographers with multiple entries into in the treatment room. Therefore, the logistic benefit derived from RA is significant for a department and reduces also the risk of human error.

From a more general perspective, optimization and calculation of RA plans is obviously a longer process than the forward calculation of 3DC plans in most of the cases. Nevertheless, although the number of treatment machines in a department cannot be easily increased, the planning power of a radiotherapy facility is much more easily customizable and addition of more planning station, availability of faster hardware and distribution of calculation burden over the planning network are all methods of ''adaptation" achievable at reasonable costs. Therefore, the longer time objectively needed for any IMRT planning process, particularly for RA, has a potentially smaller impact on the clinical throughput than the time-slots to be allocated per patients at the treatment units.

From the clinical point of view, data presented here show that RA can be considered as a safe modality for this category of patients with a good potential for improving OAR sparing and acute toxicity. On the clinical outcome, although interesting, it is possible that the observed increased in down-staging is due to statistical fluctuations from the very small group of patients. In theory, having not introduced any fractionation alteration (like dose escalation) an improved downstaging is not expected and should be investigated on larger samples. To notice anyway that the downstaging rate observed for patients treated with RA is consistent with reported data [3]. The possibility of improving organs at risk sparing shall be explored with RA in order to compensate or mitigate known negative effects from concomitant chemotherapy (e.g. reduction of diarrhoea). The smoother process of RA and its potential reduction in acute toxicity could also lead to a more uniform duration of treatments reducing the risk of unscheduled and undue interruptions.

It is obvious that the present study cannot be considered as conclusive and that long term observation of patients is needed to measure outcome and late toxicity. These preliminary results are anyway encouraging for further experience in this field.

In summary, twenty-five patients with advanced rectal cancer were treated preoperatively within and neoadjuvant radio-chemotherapy scheme with radiation course delivered with Volumetric Modulated Arc Therapy according to the RapidArc^® implementation in a clinical feasibility protocol. Quality of treatments resulted comparable with conformal modality used for benchmarking with improved conformality and reduced treatment times. Clinical outcome and early acute toxicity and assessment of tumour stage at surgery showed similar results. Anyway this study was not comparative and the study is under power to draw significant conclusion. Future investigations will aim to increase sparing of organs at risk and to look to long term outcome having the first phase achieved the primary goal to demonstrate safety and efficacy of RA.