Gastrointestinal Brachytherapy

Alexandra J. Stewart, Nikolaos Tselis, Michele Albert, Nitika Thawani, and Arthur Sun Myint

Brachytherapy in gastrointestinal cancer is generally underutilized but its uptake is increasing as more phase III evidence is becoming available for its use and mechanisms of administration are becoming easier and more reproducible. This chapter covers brachytherapy from esophagus to anus including 50 kV contact brachytherapy and injectable microspheres.

ESOPHAGEAL CANCER

Esophageal cancer is the eighth most common cancer worldwide, comprising 4% of the total cancer incidence (1). Eighty-three percent of these cases were diagnosed in the developing world. Predisposing factors include tobacco and alcohol use (predominantly squamous cell carcinoma) and reflux and obesity (predominantly adenocarcinoma). Esophageal cancer is the sixth most common cause of cancer death, with mortality following geographical patterns for incidence.

There will be an estimated 17,000 patients with esophageal cancer in the United States in 2015. In the United States, the incidence of adenocarcinoma of the esophagus is increasing, whereas that of squamous cell carcinoma is decreasing slightly. Despite the accessibility of the organ, only 6% of esophageal cancers receive brachytherapy, according to the Patterns of Care Study (2).

Figure 11.1 Single-channel esophageal HDR brachytherapy applicator (Elekta, Stockholm, ­Sweden). HDR, high dose rate.

In the setting of a high-grade obstruction, a more complex approach is indicated. The goal is to have the esophageal applicator placed through the nose. If the patient has a gastrostomy tube, the procedure can be done quite easily. The applicator can be passed through the nose and led out of the mouth. Then the endoscope is passed through the mouth and advanced, until the gastrostomy tube is visualized. A long suture or snare should be inserted into the gastrostomy tube and grasped by the endoscope. The endoscope should be withdrawn, pulling the suture out through the patient’s mouth, while care must be taken to have the other end of the suture secure outside the gastrostomy tube. The suture (or snare) should be attached to the weighted end of the esophageal catheter and slowly pulled from the gastrostomy tube while advancing the catheter into the patient’s mouth. If the endoscope cannot pass the obstruction, an additional step of passing a thin biopsy catheter past the lesion may work, thereby grabbing the snare, which, in turn, will be used to draw the applicator safely down. Care must be taken so as not to disrupt the mucosa or to create a false lumen. These procedures require a high level of endoscopic skill and carry significant potential risk of bleeding, mediastinitis, and infection, and belong in tertiary referral centers with significant case loads.

Figure 11.2 Anteroposterior conventional simulator radiograph demonstrating brachytherapy catheter in situ with clinical target volume and offset marked in red.

Figure 11.3 Sagittal CT image of the thorax demonstrating the isodose curves for an esophageal brachytherapy plan prescribed at 1 cm. This was a customized plan taking catheter curvature into account.

The best role for the addition of brachytherapy for esophageal cancer would be in Stage I tumors. Superficial tumors, described as being only mucosal or submucosal, have shown responses and low toxicity (9−11). However, the use of brachytherapy as a sole radical treatment modality is controversial as The Japanese Society for Therapeutic Radiation Oncology released a contradictory study that indicated that, for superficial lesions, brachytherapy had no benefit over external beam radiotherapy (EBRT) (12). However, most studies show a benefit with low toxicity, especially for the mucosal lesions.

Brachytherapy can be used to deliver a highly localized esophageal boost following radical EBRT. The tumor response in radical schedules has been shown in a small radiobiological analysis to be related to dose with a brachytherapy biological effective dose (BED) higher than 28 Gy in addition to 44.2 Gy EBRT yielding a significant increase in survival (13). The α/β ratio used for BED calculation was not given but the reported BEDs would be consistent with a higher α/β ratio representing acute toxicity or tumor control. Brunner et al examined chemoradiotherapy with a concomitant boost delivering up to 64.8 Gy EBRT plus up to 18 Gy in three fractions of brachytherapy (14). Survival was 61% at 1 year and 18% at 3 years. The fistula rate was 12% but it was tumor related in 8%. Aggarwal et al used hypofractionated EBRT in combination with 10 to 15 Gy single-fraction esophageal brachytherapy in patients unfit for surgical excision or chemoradiotherapy and demonstrated 51% 1-year survival (15). There were no fistulae and 8% of patients developed benign strictures. Calais et al added Mitomycin-C with cisplatin and 5-fluorouracil (5-FU) to 60 Gy EBRT and 10 Gy in two fractions of brachytherapy (16). The 3-year local control was 57% with 11% long-term toxicity.

Two randomized trials have examined the role of a brachytherapy boost in radical treatment. Yin studied 200 patients (17). The dose delivered by external beam radiation alone was 70 Gy in 35 fractions compared with 50 Gy external beam in 25 fractions plus three to four applications of medium dose rate (MDR) brachytherapy for a dose of 19.6 to 26.16 Gy. The 5-year local control was improved in the brachytherapy arm, 61.3% versus 43%, and 5-year overall survival was also improved, 17% versus 10% (P < .05). The toxicity was reported to be the same, 12.6% of perforation or hemorrhage in either arm. A second randomized study by Sur et al had similar results for 50 patients (18). The dose prescribed was 55 Gy EBRT versus 35 Gy EBRT and 12 Gy HDR boost in two fractions 1 week apart. The complication rate was higher for strictures in the brachytherapy arm, 8% versus 4%.

Table 11.1  Table of toxicity for esophageal endoluminal brachytherapy

A Cochrane meta-analysis in 2014 showed that brachytherapy has a longer duration of palliation than stenting with significantly better quality of life and lower complications (33% vs 47% [25]) and may even provide a survival advantage. Brachytherapy plus EBRT shows a trend toward better quality of life and improved dysphagia. Brachytherapy was seen to be comparable to laser for palliation and may improve the dysphagia-free interval when EBRT is added to the combination of laser and brachytherapy. The quality of life also appears to be improved in this setting (26).

An International Atomic Energy Agency (IAEA) randomized study showed that for patients with obstructive squamous cell carcinoma, the addition of EBRT 30 Gy in 10 fractions to 16 Gy in two fractions of brachytherapy resulted in significantly better symptom relief than brachytherapy alone with dysphagia relief of 83% versus 67% at 100 days (27). This benefit was maintained at the 300 day follow-up point. There was no difference in overall survival between the two groups. Whether the EBRT course could further be shortened is currently under investigation.

Table 11.3  American Brachytherapy Society–suggested schema for esophageal brachytherapy

Pancreatic cancer comprises 2% of the worldwide cancer incidence and yet accounts for 4% of worldwide mortality from cancer (1). Patients often present with inoperable disease at an advanced age and curative treatment is rare with high associated morbidity. Brachytherapy is rarely used in pancreatic cancer with the mainstays of treatment being surgery for cure or chemotherapy for palliation. One reason for this is the fear of pancreatic fistula or pancreatitis during or after brachytherapy treatment.

A series of 12 patients receiving EBRT and ¹²⁵I seed implants, placed operatively under direct vision, showed favorable local control with only one patient recurring locally and good symptom palliation (32). There were two cases of gastric hemorrhage and two pancreatic fistulae, which resolved spontaneously. An enhanced series from the same institution describes 98 patients who received operative ¹²⁵I implants, 27 of whom received EBRT, and 27 also received chemotherapy (33). Postoperative complications occurred in 19 patients, predominantly related to the surgical procedure. Median survival was 7 months overall but in stage T1 patients an improved survival of 18.5 months was seen. Palladium-103 (¹⁰³P) intraoperative implants were also used with similar outcomes and possibly lower toxicities (34,35). However, this form of brachytherapy was an invasive, technically specialized technique that has not had widespread uptake. It is likely with the continued development of stereotactic body radiation therapy that the need for pancreatic brachytherapy will further diminish.

The use of wires loaded with radium-224 (²²⁴Ra) has been investigated in vitro and in vivo in murine pancreatic cancer models and shows promising results for local control (36). The use of stents loaded with ¹²⁵I seeds placed into pigs’ normal pancreatic ducts was investigated. No complications were seen as a result of stent placement, meriting further investigation of this as a potential treatment modality (37).

BILIARY TRACT CARCINOMA

Cholangiocarcinoma is rare but still accounts for 1% of worldwide cancer deaths (1). Treatment is often limited by the location of the tumor or advanced patient age, with an average age of diagnosis being 65 years. Surgical excision carries a high mortality and a 5-year survival rate of only 11% to 44% depending on the site of lesion within the biliary tree (38). Therefore, strategies that provide radical treatment where surgery is not anatomically possible or palliation for patients who are not fit for surgery are desirable.

Figure 11.4 (A) X-ray image showing radiopaque biliary stents with a brachytherapy catheter in situ. A marker wire is placed inside the brachytherapy catheter to define its position. (B and C) The same axial CT image of the upper abdomen showing the brachytherapy catheter pictured in (A) within the liver. The clinical target volume is delineated in bold red. The brachytherapy isodoses (B) and IMRT isodoses (C) are pictured with the 100% isodose line in narrow red. Organs at risk have been delineated. IMRT, intensity modulated radiation therapy.

The role of adjuvant brachytherapy following surgical excision is unclear with several large series showing a significant survival advantage using radiotherapy (usually using a combination of EBRT and brachytherapy) particularly in the setting of R1 resection (48). However, in a series of 196 patients with perihilar tumors who were surgically operable, adjuvant brachytherapy alone was administered to 45% of patients but was not seen to affect survival on multivariate analysis (38).

LIVER TUMORS

According to World Health Organization (WHO) estimations, liver cancer is the fifth most common cancer type among men (7.5% of all male malignancies) and the seventh among women (3.4% of all female malignancies), whereas it accounted worldwide for approximately 746,000 deaths in 2012 (primary and secondary liver cancer), being the second cause of all tumor-associated deaths (50). Primary hepatocellular carcinoma (HCC) is increasing in incidence, and it is now the sixth most common cancer worldwide (1), with particularly high prevalence in Southeast Asia. It is associated with hepatitis B and C virus infection (85% of international incidence [51]), alcohol intake (52), and foods contaminated with aflatoxin (53). Because of its high fatality rates, the incidence and mortality rates are almost equal (54). HCC is the third most common cause of death from cancer worldwide; hence, improved treatment strategies are widely sought after. Complete surgical resection remains the gold standard for the cure of primary hepatic cancer (55); however, a large subgroup of patients is not suitable for an open surgical approach and the recurrence rate after partial hepatectomy is more than 50% (56). Therefore, strategies to improve surgical cure rates are under investigation (57).

In patients with unresectable disease, hyperthermal techniques such as radiofrequency ablation (RFA) and laser-induced thermotherapy have been reported as alternative local treatment choices (58−60). However, one of the major challenges with these approaches is incomplete treatment of large tumors due to their proximity to large volume vascular structures resulting in residual cancer tissue. Factors impairing the therapeutic ratio are tumor size, with an accepted upper size limit of 3 to 4 cm for optimal treatment, and the “heat sink” effect of nearby vasculature, stopping effective cytoreduction in perivascular lesions (61). To overcome these limitations, hyperthermal ablative therapies have been combined with other treatment modalities such as transarterial chemoembolization (TACE [62]), percutaneous ethanol injection (PEI [63]), or radiotherapy. Lin et al combined RFA with MRI-guided percutaneous implantation of ¹²⁵I seeds to give a complete response rate of 96% in lesions that would otherwise not be fully covered by RFA (64). In addition to use in the adjuvant setting (65), LDR brachytherapy has also been employed as the sole treatment for unresectable primary liver malignancies. Lin et al treated 65 inoperable hepatocellular carcinoma lesions using MRI-guided percutaneous implantation of ¹²⁵I seeds (66). The minimum peripheral dose delivered was 144 Gy and the treatment yielded an overall response rate of 84.5% with no clinically relevant complications.

Figure 11.6 (A) A 55-year-old woman with a diagnosis of multifocal hepatocellular carcinoma (histologically proven) with no overt cirrhosis. MRI images with gadoxetic acid contrast demonstrating the appearances of the tumor at baseline (i, ii) and follow-up at 3 months (iii, iv) and 18 months (v, vi) after ⁹⁰Y radioembolization. Figures i, iii, and v display the arterial phase, and ii, iv, and vi the hepatobiliary phase. The image series displays at baseline a multifocal bilobar disease with a partial response after 3 months and a complete remission of the disease (according to perfusion criteria) after 18 months. The hyperintense elliptical mass at the right lateral liver capsule resembles a persisting hematoma, which was related to a liver biopsy. (B) A 69-year-old woman with synchronous hepatic metastatic colorectal cancer, first diagnosed 2 years prior to presentation for a ⁹⁰Y radioembolization. In that period, the patient received all standard palliative chemotherapy including antibodies. However, the disease progressed on standard chemotherapy and further therapy was warranted. MRIs with gadoxeticacid contrast are demonstrated at baseline (i–iii), at 3 months (iv–vi), and at 12 months (vii–ix) after ⁹⁰Y radioembolization. Series i–iii displays hepatobiliary phase imaging, series iv–ix display portal venous phase imaging. The image series display at baseline a multifocal bilobar disease with a beginning partial response after 3 months and a further remission of the disease after 12 months. No concomitant chemotherapy was administered during this follow-up period. A relapse of the hepatic disease occurred 18 months after radioembolization. Salvage chemotherapy was recommenced and the patient died 34 months after radioembolization.