Esophageal Cancer: Radiation Therapy Planning



Fig. 4.1
The esophagus is divided into four regions by the AJCC. The cervical esophagus extends from the upper esophageal sphincter to the thoracic inlet. The upper thoracic esophagus extends from the thoracic inlet to the inferior border of the azygous vein. The middle thoracic esophagus extends from the inferior border of the upper thoracic region to the inferior border of the inferior pulmonary vein. The lower thoracic esophagus extends from the mid-thoracic esophagus to the GEJ. The distance of each from the incisors is indicated on the left (Reproduced with permission from: Saltzman and Gibson [44] Copyright © 2016 UpToDate, Inc. For more information visit www.​uptodate.​com.)





 

  • (b)


    Similar to other gastrointestinal sites, the esophageal wall has well-defined layers.


    1. (i)


      The epithelium is separated from the lamina propria by a basement membrane.

       

    2. (ii)


      Sequential layers deep to the lamina propria include the muscularis mucosae, submucosa, and muscularis propria.

       

     

  • (c)


    With respect to esophageal cancer treatment planning, longitudinal lymphatic channels can extend into the submucosa and lamina propria, potentially allowing for lymphatic spread by even relatively superficial tumors along the entire length of the esophagus.

     






      4.2 Patterns of Spread





      1. (a)


        Given the extensive lymphatic channels of the esophagus, nodal spread of disease, both proximally and distally, is common.


        1. (i)


          For patients with squamous cell carcinoma of the esophagus, lymph node metastases can be found in 70 % at autopsy [2].

           

        2. (ii)


          In a review of 1,077 patients treated with esophagectomy for squamous cell carcinoma, 6 % had “skip” lymph node metastases [3].


          1. 1.


            Specifically, in patients with upper thoracic tumors, 55.6 % of patients had involved lymph nodes that were above the level of the carina, and 22.3 % had lymph nodes inferior to the carina, including 5.6 % with subdiaphragmatic involvement.

             

          2. 2.


            Similarly 4.5 % of patients with lower thoracic lesions had either upper thoracic and/or cervical lymph node involvement.

             

          3. 3.


            Finally, 63 % of all involved lymph nodes had microscopic deposits, underscoring the importance of appreciating nodal levels at risk, even in absence of positive imaging findings.

             

           

         

      2. (b)


        Similar rates of lymph node involvement have been observed for lower thoracic and GEJ adenocarcinomas, where approximately 70 % of patients have nodal metastases at presentation.


        1. (i)


          Based on the Siewert classification of GEJ adenocarcinomas [4], type I tumors (defined as arising from 5 to 1 cm proximal to the Z line) spread to both superior and inferior lymphatics, with greater than 50 % involvement of the paraesophageal and left gastric lymph nodes [5].

           

        2. (ii)


          Type II tumors (arising within 1 cm proximal and 2 cm distal to the Z line) have similar rates of left gastric lymph node involvement but less than 10 % involvement of the more superior paraesophageal nodes.

           

        3. (iii)


          Type III tumors (arising 2–5 cm inferior to the Z line) typically spread toward the celiac axis.

           

         

      3. (c)


        Lymph node involvement increases with depth of invasion of the primary tumor with 45 %, 85 %, and 100 % of adenocarcinoma patients with T2, T3, and T4 lesions, respectively, harboring nodal involvement in one series [5].

         


      4.3 Imaging





      1. (a)


        Multiple imaging modalities are used to determine local and distant disease extent and guide the treatment planning process.


        1. (i)


          EGD facilitates defining the location of the primary tumor within the esophagus in relation to the distance from the incisors and with respect to the GEJ.


          1. 1.


            The carina and GEJ are typically located at 25 and 40 cm, respectively, from the incisors.

             

          2. 2.


            EGD also facilitates biopsy of sites concerning for satellite lesions or submucosal spread.

             

           

        2. (ii)


          Depth of tumor invasion as well as paraesophageal and perigastric lymph node involvement is assessed by endoscopic ultrasound (EUS).


          1. 1.


            As compared to surgical pathology, EUS provides 85–90 % and 75–80 % accuracy rate for depth of invasion and nodal involvement, respectively [6–9]. This is generally better than CT alone which has an approximately 70 % and 50-70 % accuracy for tumor invasion and nodal involvement, respectively.

             

           

        3. (iii)


          The addition of PET imaging can improve detection of stage III or stage IV disease by approximately 20 % in both scenarios, but is not as accurate as EUS for determining locoregional nodal involvement [10, 11].

           

         


      4.4 Simulation





      1. (a)


        Patients should undergo CT simulation for treatment planning.


        1. (i)


          Patients are typically supine with a customized immobilization device such as a wing board for 3D planning or an alpha cradle, body fix, or Vaclock on indexed wing board for IMRT planning.

           

        2. (ii)


          For more proximal lesions involving the upper thoracic or cervical esophagus, immobilization with a mask may be indicated.

           

        3. (iii)


          Esophageal contrast delivery can aid in delineation of the primary tumor, although it can artificially dilate the proximal esophagus, while small bowel contrast can help identify any bowel in field.

           

        4. (iv)


          IV contrast can aid in distinguishing both involved lymph nodes and arterial structures, including the celiac axis and its branches, to guide elective nodal coverage.

           

        5. (v)


          CT scans are obtained with approximately 2–3 mm slices from the upper neck through the mid-abdomen to provide sufficient visualization of the target tissues and surrounding structures.

           

         

      2. (b)


        Target motion due to respiration, the cardiac cycle, and peristalsis can be evaluated with a 4D CT scan and treatment with respiratory management implemented as necessary. Abdominal compression or breath holding techniques can limit respiratory motion in select cases.

         

      3. (c)


        For tumors that involve the lower esophagus and stomach, fasting prior to simulation and treatment each day may allow for greater reproducibility of treatment. In contrast, simulation without fasting and delivery of oral contrast may allow for planning in a “worst-case” scenario when the patient is later instructed to fast prior to treatment delivery.

         


      4.5 Target Delineation





      1. (a)


        Gross tumor volume (GTV) delineation should take into account findings from all pretreatment diagnostic tests including EGD, EUS, barium swallow, CT, and PET imaging.


        1. (i)


          The primary tumor can often be seen as thickening of the esophageal wall on diagnostic and planning CTs. This should be correlated with findings from EGD which provide useful information as to the tumor location in relation to other anatomical landmarks and length in centimeters.

           

        2. (ii)


          The results of EUS can help determine extension of the primary tumor outside of the esophageal wall and into any nearby tissues as well as the extent of nodal involvement.

           

        3. (iii)


          Any overt lymphadenopathy should also be included in the GTV.


          1. 1.


            PET imaging can be helpful in identifying nodal disease although the sensitivity of this modality for detecting nodal metastases has been reported to be as low as 67 % [12].

             

          2. 2.


            The results from PET, however, have been shown to correlate with those from endoscopic ultrasound [8], and the addition of PET reduces variability in GTV contouring.

             

           

         

      2. (b)


        The clinical target volume (CTV) includes subclinical regions at risk for spread of disease.


        1. (i)


          As potentially involved sites of subclinical disease are not readily apparent on exam or by imaging, CTV design for esophageal cancer relies on known patterns of spread based on pathologic findings in studies following resection or autopsy specimens.

           

        2. (ii)


          In terms of longitudinal spread within the esophagus itself, a pathologic analysis of 66 resections from patients with squamous cell carcinoma of the esophagus demonstrated that a 3 cm margin both proximally and distally on the primary tumor would cover microscopic disease in 94 % of patients [13].

           

        3. (iii)


          Similarly, in cancers involving the GEJ, a 3 cm proximal margin and a 5 cm distal margin have been shown to cover subclinical disease in 100 % and 94 % of patients, respectively [13].

           

        4. (iv)


          In the CROSS trial comparing preoperative chemoradiation to surgery alone, a 4 cm margin superior and inferior from the GTV to planning target volume (PTV) was employed except with tumor extension into the stomach when a 3 cm distal margin was used [14]. With these margins, locoregional recurrence occurred in 5 % within the target volume, in 2 % in the margins, and in 6 % outside the radiation target volume. Only 1 % had an isolated infield recurrence after CRT plus surgery [15].

           

        5. (v)


          The current RTOG 1010 trial investigating the addition of trastuzumab to neoadjuvant chemoradiation recommends a 4 cm margin superiorly and inferiorly and 1–1.5 cm radially from the GTV to CTV expansion, with an additional 0.5–1 cm expansion from the CTV to PTV [16].

           

         

      3. (c)


        Elective nodal coverage incorporated into the CTV varies based on the primary tumor’s location in the esophagus.


        1. (i)


          A recent consensus contouring guideline panel recommended that the CTV should extend 1 cm superior to the most proximal-involved periesophageal node [17]

           

        2. (ii)


          Lesions of the cervical and upper thoracic esophagus generally require coverage of the supraclavicular and superior mediastinal lymph nodes. Some authors have recommended coverage of the supraclavicular and superior mediastinal lymph nodes with any primary disease at or above the level of the carina [17].

           

        3. (iii)


          Lesions of the middle esophagus should include paraesophageal lymph nodes within the mediastinum but also may consider subdiaphragmatic nodal basins as the involvement of these has been shown to approach 20 % [3]. Coverage of both regions may, however, produce a significantly large treatment volume, and consideration should be given to the potential associated toxicity.

           

        4. (iv)


          For distal esophageal adenocarcinomas, the periesophageal lymph nodes and celiac nodal basins should be included for most, if not all, patients.


          1. 1.


            A recent consensus contouring guideline panel recommended that CTV should extend to the celiac axis and approximate the lateral border of T12 on the right and extend 0.5–1 cm lateral to the aorta on the left [17].

             

           

        5. (v)


          In a pathologic analysis by Meier and colleagues, lymphovascular invasion, greater depth of invasion by the primary tumor, and higher grade were all predictive of lymph node involvement, suggesting that the CTV should be extended to include additional at-risk regions in these circumstances [18].


          1. 1.


            Specifically, for tumors involving the GEJ with lymphovascular invasion, coverage of the left and right gastroepiploic, greater curvature, celiac trunk, and splenic hilum lymph nodes is indicated.

             

          2. 2.


            T3 or T4 tumors require treatment of the left and right gastroepiploic, greater curvature, celiac trunk, splenic hilum, splenic artery, and common hepatic artery nodes.

             

          3. 3.


            High-grade tumors also correlated with an increased risk of involvement of the left gastroepiploic, greater curvature, and celiac trunk nodes.

             

          4. 4.


            In addition, tumors extending greater than 1.5 cm proximal to the Z line have an increased risk of involvement of periesophageal nodes in the middle esophagus, and consideration should be given to covering these areas electively.

             

           

        6. (vi)


          Elective nodal irradiation for esophageal squamous cell carcinoma patients is an area of controversy.


          1. 1.


            In a prospective study of 53 patients with T1-4 N0-1 disease treated with 3D-CRT alone to 68.4 Gy in 41 fractions, only 3, or 8 %, had an isolated, out-of-field nodal recurrence [19].

             

          2. 2.


            A more recent study of patients with T4 squamous cell carcinoma demonstrated a 1.8 % rate of elective nodal failure out of 56 patients treated with concurrent chemotherapy and radiation [20]. Elective nodal coverage in patients with squamous cell carcinoma remains a topic of study.

             

           

        7. (vii)


          As before, enlarging target volumes must be balanced with increasing risk of associated toxicity.

           

         

      4. (d)


        Planning target volume (PTV) design takes into consideration both inter- and intrafraction variability including organ motion.


        1. (i)


          For the esophagus, this can include motion associated with respiration, the cardiac cycle, and peristalsis.

           

        2. (ii)


          Reports of esophageal motion range from 0.1 to 0.4 mm in the anterior-posterior direction, 0.3–4.2 mm laterally, and 3.7–10 mm superior to inferior [21]. One study of interfraction motion suggested margins of 12 mm left to right, 10 mm posteriorly, and 9 mm anteriorly to account for this [22].

           

        3. (iii)


          In general, margins used to expand from the GTV to PTV include 5 cm proximally and distally and 2–2.5 cm radially [14, 17].

           

         


      4.6 Field Design





      1. (a)


        Cervical and upper thoracic tumors


        1. (i)


          Treatment of proximal esophageal lesions can require fields spanning from the inferior aspect of the larynx superiorly to the level of the carina inferiorly for the primary tumor and periesophageal/mediastinal lymph nodes as well as the supraclavicular nodal basins.

           

        2. (ii)


          Using older methods, this was achieved with lateral parallel opposed or oblique portals to the primary tumor and a single anterior field for the supraclavicular and superior mediastinal nodes.

           

        3. (iii)


          Other techniques treated lesions in this region by means of a four-field box approach, using a wax bolus to compensate for the lack of tissue above the shoulders, arc rotations, anterior wedged pairs, and three- or four-field methods using posterior oblique portals combined with a single anterior portal or anteroposterior-posteroanterior (AP/PA) fields.

           

        4. (iv)


          Using three-dimensional (3D) approaches, varying techniques have been implemented.


          1. 1.


            Historically, one approach entailed treating the primary tumor and lymph nodes using an AP/PA approach to 39.6–41.4 Gy at 1.8 Gy per fraction, followed by a left or right opposed oblique pair to bring the total dose to 50.4 Gy, thereby limiting the spinal cord dose.

             

          2. 2.


            This technique will generally exclude the supraclavicular fossa, and a separate electron field is often added, treating to a depth of 2–3 cm, depending upon individual anatomy.

             

           

        5. (v)


          More recently, intensity-modulated radiation therapy (IMRT) has been implemented for cervical lesions, similar to tumors of the anatomically related head and neck.


          1. 1.


            Compared to 3D planning, IMRT has been shown to have improved conformality to the target with decreased dose to normal tissues [23].

             

          2. 2.


            An IMRT plan, as shown in Fig. 4.2, allows for a single plan that spares normal tissues including the spinal cord and lungs and is our preferred method for treating these tumors.

            A331976_1_En_4_Fig2_HTML.gif


            Fig. 4.2
            (a–c) GTV (blue) and PTV (red) contours on axial slices of a patient with lymph node positive cervical esophageal cancer treated with IMRT. (d) 3D rendering of GTV and PTV volumes in relation to the lungs (yellow), heart (cyan), and spinal cord (purple). (e) Representative isodose curves for the same patient

            Only gold members can continue reading. Log In or Register to continue

            Related posts:

            Default ThumbnailBasics of Chemotherapy Default ThumbnailRectal and Colon Cancer: Background and Clinical Evidence Default ThumbnailPrimary Liver Cancer: Background and Clinical Evidence Pancreatic Cancer: Background and Clinical Evidence Rectal and Colon Cancer: Radiation Therapy Planning Gastric Cancer: Radiation Therapy Planning

            Stay updated, free articles. Join our Telegram channel
      1. Tags:
      Oct 18, 2017 | Posted by in GASTROENTEROLOGY | Comments Off on Esophageal Cancer: Radiation Therapy Planning

      Full access? Get Clinical Tree
      Get Clinical Tree app for offline access