Bilateral kidney preservation by volumetric-modulated arc therapy (RapidArc) compared to conventional radiation therapy (3D-CRT) in pancreatic and bile duct malignancies

Over the last decades, progress in treating pancreatic cancer has remained modest, and the disease is still associated with very poor prognosis regardless of stage. Although there is currently no clear standard therapy, chemotherapy (CT) alone or radiochemotherapy (RCT) are two treatment options available for patients with non metastatic locally advanced pancreatic cancer. Two systematic reviews have shown no difference between both treatment modalities in terms of overall survival, but an increased toxicity with RCT [1, 2]. Lack of benefit for RCT can be explained by the low compliance related to acute toxicity. In a recently published phase 3 trial comparing intensive induction chemoradiation followed by gemcitabine to gemcitabine alone, Barhoumi et al. reported that only 42% of patients were able to receive 75% of the planned chemoradiation schedule (60 Gy, 2 Gy/fraction and concomitant infusion of 5-fluorouracil and cisplatin) whereas 73% of patients were given 75% of the treatment of gemcitabine alone (1,000 mg/m2 weekly) [3]. Interestingly, the administration of induction chemotherapy before RCT has been shown to be a promising strategy for selected patients with non progressive disease, which may help to define the subset of patients likely to benefit from RCT while sparing those with rapidly progressive disease from potentially toxic radiotherapy [4]. In the case of resectable pancreatic cancer, it is now well established that RCT should be only considered in the adjuvant setting after insufficient pancreatic resection [5], whereas for bile duct malignancies, radiotherapy (RT) is likely to prolong survival in case of locally advanced disease as well as after R1 resection [6].

The occurrence of toxicity is clearly associated with the use of old RT techniques and the delineation of important target volumes. Besides, recent data regarding the treatment of pancreatic disease are in favor of RCT without prophylactic irradiation of peripancreatic lymph nodes since the use of small radiation fields seems to provide similar local recurrence rates and lower gastrointestinal (GI) toxicity [7, 8]. The goal of modern radiotherapy such as the intensity modulated radiation therapy (IMRT) or the volumetric-modulated arc therapy (VMAT) is to insure full delivery of the prescribed dose, even to allow dose escalation while reducing doses to surrounding critical organs (small bowel, kidneys, spinal cord). A few studies investigated IMRT for pancreatic cancers and showed a statistically significant improvement over three dimensional conformal radiotherapy (3D-CRT) in lowering dose to the liver, stomach and bowel [9‐12]. Therefore, advances in RT delivery with IMRT suggest a tolerability advantage including a lower incidence of acute toxicity [13, 14]. In addition to a favorable toxicity profile, the local control is not compromised with IMRT [15].

Volumetric-modulated arc therapy (RapidArc) is a new form of IMRT optimization combining one gantry rotation and the capability to vary dose-rate, gantry speed and dynamic multi leaf collimator (MLC) [16]. Details of the RapidArc process and specific quality assurance procedures have been described in several publications [16, 17]. The VMAT approach has a number of potential advantages compared to IMRT by significantly reducing the treatment time and the number of monitor units (MU), and improving normal tissue sparing while keeping the adequate target coverage [18‐20]. In a trial comparing VMAT and IMRT for pancreatic cancers, VMAT was shown to provide superior conformity indices to the target and to reduce the dose in organs at risk (OAR) volumes with a shorter delivery time [21].

The purpose of the present study was to perform a comparative dosimetric analysis of RapidArc and 3D-CRT techniques in the treatment of pancreatic and bile duct malignancies with a special focus on the preservation of bowel and bilateral kidney function.

The treatment of pancreatic cancer remains a major clinical challenge. Only 20% of patients are considered eligible for curative surgical resection, whereas radiotherapy given as definitive treatment or adjuvant treatment is a subject of controversy, essentially because of radiation-induced toxicity. It is now generally accepted that RA technology offers the potential to deliver better conformal radiation dose with less toxicity as compared with 3D-RT [9‐15]. In the present study, we chose to directly compare 3D-CRT with RA without testing IMRT for two main reasons. Firstly, several studies have already demonstrated benefit of IMRT over 3D-CRT [13‐15], and advantage of VMAT over IMRT in terms of conformity and OAR dose reduction [21]. Secondly, in our institution, the RA technique has been implemented as a standard of care for many tumor sites, and is preferred to IMRT because of shorter delivery time. That choice has direct impact on daily practice by improving patient comfort and reducing intrafraction changes in patient position [20].

In this dosimetric analysis, particular emphasis was placed on improving kidney sparing according to the Quantec constraints [22]. Much has changed since the papers by Emami et al., and later Cassady et al., suggesting that total doses of 18-23 Gy and 28 Gy in 0.5-1.25 Gy/fraction may be associated with a 5% and 50% risk of injury at 5 years, respectively [23, 24]. Nowadays, conventional treatments are delivered in 2 Gy daily fractions, so we need to take caution with these first published cut-offs. According to current recommendations for partial bilateral kidney radiotherapy, the mean dose should be lower than 15-18 Gy and the V20 lower than 32%, but substantial uncertainty still remains because few published data are available. Moreover, study follow-up periods are usually too short to draw any clear conclusions [22]. RT-induced kidney injury is generally subclinical. Signs and symptoms including arterial hypertension and renal failure usually do not develop until late and their incidence is often underestimated because of a long latency period and the short-life expectancy of the majority of the patients suffering from pancreas cancer. Conventional 3D-CRT allows total preservation of contralateral kidney, but has a detrimental effect on homolateral kidney. In the new era of arctherapy, it is crucial to minimize as much as possible the dose to both kidneys for bilateral preservation. For this reason, we decided to set strong dose constraints for each kidney, i.e. a V10 of 0 Gy and a mean dose < 5 Gy with a successful achievement. In the comparative planning study of VMAT and IMRT conducted by Eppinga et al. in a series of 10 patients with locally advanced pancreatic cancer, doses to the left kidney were significantly reduced with VMAT (V15 of 7.2% and V20 of 2.4%), whereas right kidney sparing was similar with both techniques (V15 of 20.4% and V20 of 9.9% with VMAT). The mean doses delivered to the left and right kidneys were 8.8 and 10.9 Gy, respectively [21].

Our PTV's volumes were smaller. This difference may explain for a part the improvement of OAR sparing. In our present study, the delineation of CTV/PTV corresponds to the new guidelines (manuscript accepted for publication in the Int J Radiat Oncol Biol Phys).

It is unclear if any of these dose and volume parameters could be sufficient to preserve renal function, and the lack of clinical outcome data makes challenging the interpretation of any results. This is the reason why the kidney function of patients requiring abdominal radiotherapy and treated by RA in our institution is now routinely evaluated at baseline. A dimercaptosuccinic acid (DMSA) scintigraphy is performed before and at the end of treatment, and blood tests are carried out during the follow-up period. Since early changes in renal flow are correlated with an increased toxicity, this management offers the possibility to accurately evaluate radiation-associated kidney injury.

Our study demonstrated that kidney sparing can be achieved without compromising the bowel function. Improvements in decreasing dose to the bowel are therefore essential because GI toxicity is the most common acute side effect. Devisetty et al. showed a significant higher acute GI toxicity for V30 > 450 cm³ than for V30 ≤ 450 cm³ (33% vs. 8%, p = 0.003, respectively) [25]. Recommendations for V45 < 195 cm³ when peritoneal cavity was delineated were reported by Kavanagh et al. [26, 27]. We delineated the bowel as the peritoneal cavity, and obtained a V30 < 200 cm³ and a V40 < 55 cm³ with RA, which is very promising for reducing acute digestive toxicity. Dosimetric data with respect to the liver complied with the Quantec recommendations (i.e. mean dose < 30-32 Gy) for both techniques with reported mean doses of 11.0 Gy and 9.7 Gy with 3D-CRT and RA, respectively [27].

Despite these positive results, the IMRT/RA techniques raise concerns about the increased number of MU required, and the low-dose radiation delivered to healthy tissue compared with 3D-RT treatment leading to a potential increase of the risk of secondary malignancies [28, 29]. In this study, we noted superiority of 3D-CRT over RA for doses inferior to 15 Gy, but for "middle doses" (> 15Gy) healthy tissue doses were decreased by RA. The real impact and the dose level potentially incriminated in the risk of developing a second cancer are still unclear.

The RA technique results in better OAR sparing than 3D-CRT for pancreatic and bile duct malignancies, especially for small bowel and kidney preservation, while covering target volume, making the rationale for use of radiotherapy in these diseases more solid and less controversial. These improvements have led us to implement rapidly this technique into the clinic as a standard of care for pancreatic and bile duct cancers and after 4 to 6 cycles of chemotherapy.

The authors declare that they have no competing interests.