A 73 year old African American male with history of SCCA of the esophagus presented with mild dysphagia. Subsequent endoscopy (EGD) revealed recurrent SCCA of the esophagus. Seven years prior he was diagnosed with a T1-2, N1 SCCA of the left anterior tonsillar pillar and treated with radiation therapy (XRT) (6660 cyG to the primary tumor). Four years later he developed Stage III esophageal SCCA (T4, N0, M0), by chest CT, located at 33 cm from the incisors that was 3 cm in length. This lesion was treated with 5400 cGy XRT and chemotherapy with curative intent. On the third (current) presentation, a moderately differentiated, T2, by endoscopic ultrasound (EUS), SCCA lesion at 24 cm was discovered (Figure 1 and 2). This new lesion was above the prior field of XRT and proximal to the original stage III esophageal SCCA.
The patient was evaluated by the institution's tumor board and deemed neither a surgical nor XRT candidate based on significant cardiac comorbidities and maximum dose XRT with his prior treatments. Palliative treatment options were explored. Based on the success with ablation of esophageal lesions as described above, palliation via CSA was pursued for this patient. At the time of this case, the device was under an investigational device exemption (IDE) issued by the FDA. Therefore, an FDA Humanitarian Use Device exemption was requested and granted for use of the CSA device. The patient signed a consent form approved by the authors IRB and the FDA. The patient's 2 cm long, hemi-circumferential SCCA was treated with two 30 second pulses of CSA under direct endoscopic visualization on an outpatient basis (Figure 3). This particular dosimetry was chosen based on early clinical trials in Barrett's esophagus and swine data. The first CSA of Barrett's esophagus in clinical trials used a dose of 40 seconds (two 20 second applications separated by a 20–30 second thaw). Early animal studies at the author's institution (unpublished) have demonstrated that the depth of injury correlates with duration of freeze. High grade dysplasia has been effectively treated with CSA at 20 seconds times three cycles. Thus in hope of achieving greater depth of injury without excessive necrosis, two cycles of 30 seconds each were applied. CSA was applied in such a way that the entire tumor with margins of 1–2 cm was frozen. This technique resulted in a near circumferential freeze of the esophagus but with a focus on the tumor which endoscopically appeared hemi-circumferential in distribution. Interestingly, freezing of the tumor under direct endoscopic visualization demonstrated that the tumor differentially retracted when frozen relative to the surrounding esophageal mucosa making the margins of the tumor more distinct. Prior to freezing, the margins of the tumor diffusely blended into the esophageal wall.
This patient was treated with lansoprazole 30 mg BID throughout the CSA period. The patient experienced no initial complications. One month latter a follow-up EGD revealed endoscopic resolution of the tumor, but biopsies remained positive for moderately differentiated SCCA (Figure 4). CSA of the area was repeated, treating 40% of the esophageal circumference 4 cm in length spanning the prior SCCA with three, 20 second pulses. This shorter pulsed duration of CSA was chosen because less depth of injury was desired based on the significant response with the initial treatment. Increased numbers of CSA cycles are associated with greater tissue injury. Therefore the intent was a more intense treatment but with less depth of injury. Twenty-four hours later, the patient experienced odynophagia requiring oral narcotics for 3 weeks followed by the development of a stricture at one month post-CSA. This was treated with Savary dilation but has subsequently become a persistent esophageal stricture at that the site. Biopsies at the cancer site 6 weeks after the second cryoablation and bimonthly in the subsequent 12 months of follow-up were completely negative for dysplasia or neoplasia.
The persistent esophageal stricture (Figure 5) has required multiple dilations with various types of dilators (with and without corticosteroid injections) and temporary (3 month) esophageal stent (Polyflex® by Wilson Cook) placement. The patient continues to work and remains cancer free 24 months post-treatment.