Correspondence *General Surgery, Northwestern Memorial Hospital, 251 E Huron, Chicago, IL 60611, USA

Email r-maciver@md.northwestern.edu

The case

A 62-year-old female underwent endoscopic injection of botulinum toxin type A (Botox^®; Allergan, Irvine, CA) for esophageal achalasia. The patient had shown symptoms of dysphagia for over a year but had received no previous treatment for the disorder. The patient's BMI was 43.6 kg/m² and 1.5 years before receiving Botox^® she had been diagnosed as having diabetes, which was controlled with glibenclamide and insulin. The patient had developed retinopathy as a result of chronic diabetes. An esophagram taken during the month before Botox^® was administered to the patient demonstrated that she had a dilated esophagus (diameter 16.1 cm) with delayed emptying. Before receiving the Botox^® injections the patient had been instructed not to take any anticoagulants (e.g. aminosalicylic acid, NSAIDs, etc.), and she had been compliant with this request. The patient was not taking warfarin preoperatively.

Under conscious sedation (achieved with 50 mg pethedine and 3.5 mg midazolam), the patient received four equally sized injections of 5 ml Botox^® solution (5 U/ml) in four equally spaced quadrants of her lower esophageal sphincter (LES), just proximal to the squamocolumnar junction (total dose 100 U); topical anesthetic was used at all injection sites. At the time of treatment the patient's esophagus was empty and contained no liquid or food residuals, and no Candida albicans infection was noted. The patient tolerated the injections well with no retching, coughing or belching during the procedure. Immediately following the treatment, the patient complained of a mild sore throat. Her throat pain continued at home and she was treated with throat lozenges, pantoprazole (40 mg twice daily) and clonazepam by mouth (1 mg daily) for suspected gastroesophageal reflux and esophageal spasm.

The patient presented to the emergency department with chest pain 4 days after receiving Botox^® treatment. Chest radiography and Gastrografin^® (Bracco Diagnostics Inc, Princeton, New Jersey) followed with barium swallow study both revealed no abnormalities. The patient was sent home only to return the next day with continued chest pain and newly developed fever. After a cardiac event had been ruled out by electrocardiography and measurement of cardiac enzyme levels, a CT scan was performed and revealed minimal inflammation of the esophagus around the area of the LES but no evidence of proximal extension, mediastinitis or abscess. An esophagoduodenoscopy was also performed and revealed dysmotility and bluish discoloration of the distal esophagus, but no mucosal disruption. The patient's fever was attributed to a urinary tract infection after the results of a urinalysis were found to be positive (leukocyte esterase, 2+; bacteria, 3+; white blood cells, 50-10010⁹/l [50-10010³/l]; nitrite, negative) and she was prescribed levofloxacin (500 mg daily for 5 days) and discharged to her home the same day. The patient presented again to her gastroenterologist on day 9 after Botox^® administration, with continuing chest pain and fever.

A repeat CT scan revealed increased inflammation of the patient's esophagus around and extending a few centimeters above the LES, consistent with mediastinitis, but no evidence of perforation was observed (Figure 1). The patient was admitted to hospital owing to continued fever and was started on total parenteral nutrition and treatment with intravenous piperacillin–tazobactam (3.375 g every 6 h), intravenous ciprofloxacin (400 mg every 8 h) and intravenous fluconazole (800 mg daily). After 24 h of antibiotic therapy the patient continued to be febrile. A repeat esophagoduodenoscopy on day 11 after the procedure revealed a lateral 2 cm ulceration located approximately 36 cm from the incisors, in the area of the Botox^® injections.

Figure 1 Chest CT scan taken 9 days after the patient received Botox^® (Allergan, Irvine, CA) injections for achalasia.

Thickening of the esophagus and of the esophageal lumen around and extending a few centimeters above the lower esophageal sphincter can be seen (arrow), which is indicative of inflammation consistent with mediastinitis. No evidence of perforation is present.

Full figure and legend (15K)Figures & Tables indexDownload Power Point slide (62K)

Esophageal ultrasonography had been scheduled to assess whether the patient had an intramural abscess but was cancelled owing to concerns of esophageal perforation. A left thoracotomy through the seventh intercostal space was performed on the patient on day 13 after the procedure to investigate possible abscess, mediastinitis, perforation or a combination of these complications. Approximately 400 ml of turbid pleural fluid was encountered that surrounded an edematous inferior pulmonary ligament. The ligament was mobilized and revealed a proximally dilated esophagus that was surrounded by necrotic tissue above the level of the gastroesophageal junction. No obvious perforation was noted.

Esophagoduodenoscopy was performed during the thoracotomy, which enabled the previously identified intramural ulcer to be located (Figure 2). The ulcer communicated into the muscle layer, but did not freely communicate with the mediastinum. Insufflation with immersion of the esophagus revealed no perforation. Before the patients' thorax was closed a nasogastric tube and a chest tube were inserted and a Blake drain was placed around the esophagus. The chest tube enabled 290 ml of pleural fluid to be drained before it was removed after 2 days, at which point the patient was still afebrile; the Blake drain enabled removal of a further 1,736 ml of fluid over 9 days.

Figure 2 Esophagoduodenoscopy image of the case patient's esophagus performed 13 days after Botox^® (Allergan, Irvine, CA) injections for achalasia.

A lateral 2 cm intramural ulcer is visible, located approximately 36 cm from the incisors and at the area of the Botox^® injections. The ulcer communicated into the muscle layer, but did not freely communicate with the mediastinum.

Full figure and legend (14K)Figures & Tables indexDownload Power Point slide (61K)

The patient's condition improved postoperatively over the next week. The results of laboratory tests on blood and wound cultures from samples taken during surgery were negative for growth of bacterial organisms after 5 days of culture. The results of all final anaerobic, aerobic and fungal cultures were negative for growth. Of note, the patient had received more than 5 days' therapy with broad-spectrum antibiotic treatment at the time of her surgery. After a postoperative upper gastrointestinal swallow study proved negative for esophageal leak the patient was started on a solid oral diet. The paraesophageal Blake drain was removed from the patient after it had been verified that output of pleural fluid did not increase following oral food intake. A repeat CT scan performed approximately 2 weeks after thoracotomy revealed a marked reduction in paraesophageal inflammation.

Two years on from these procedures the patient continues to have dysphagia yet has so far decided against any further treatment. She tolerates solids and liquids and there has been no notable change in her BMI. If her dysphagia worsens it has been recommended that she undergoes a Heller myotomy.

Discussion of diagnosis

Chest pain, most probably caused by reflux and/or esophageal spasm, is the most frequently reported complication of Botox^® injection for achalasia.¹ In the case patient, the chest pain was caused by mediastinitis, but other reports of mediastinitis resulting from Botox^® injection or any other transient chemical therapy are lacking. Mediastinitis has, however, been reported to occur after use of chemical substances that have permanent effects on the esophagus,² and has also been described in relation to other endoscopic treatments, such as esophageal sclerotherapy.³

Two case studies have reported ulceration of the mucosa with sinus tract formation after Botox^® injections.⁴ The progression from simple mucosal irritation to ulceration with mediastinitis is, however, rare and in the present case was probably a consequence of the patient's diabetes-related achalasia.

Differential diagnosis

Tests should be performed to immediately rule out esophageal perforation in any patient presenting with pain and fever after esophageal instrumentation because it is associated with a high risk of morbidity and mortality; chest pain is frequently reported after endoscopic Botox^® injection, but fever is not. Perforation can result from mechanical tears caused by the endoscope, overdistension by air insufflation, or from the performance of procedures that could damage the esophagus wall (e.g. biopsy or the removal of a polyp). Investigations to rule out perforation should include plain films (to assess for free air), CT scan (to assess for free air and possible abscess formation) and a swallow study using Gastrografin^®, barium or both (to assess for esophageal leak). An esophagram enables prompt diagnosis of a perforation. Upper gastrointestinal endoscopy can provide the advantage of identifying areas at risk of perforation that can be followed up with repeat endoscopy, if actual perforation has already been ruled out by an esophagram. In addition, endoscopic ultrasonography can identify areas of submucosal necrosis or abscess. In this case it was felt that further instrumentation such as ultrasonography could have caused a perforation that would have further soiled the mediastinum of the patient.

In patients in whom perforation and suspicious lesions have been ruled out, other differentials that also need to be excluded are gastroesophageal reflux and traditional sources of fever. Gastroesophageal reflux is the most likely cause of chest pain. In the case patient chest pain was empirically treated with antacid therapy and antispasmotics, with no symptom relief. When excluding the traditional sources of fever, special attention should be paid to ruling out aspiration pneumonia, as regurgitated food particles can be aspirated.

Treatment and management

Primary achalasia is an esophageal motility disorder that is characterized by impaired relaxation of the LES because of unopposed excitation of acetylcholine. Traditionally, this disorder has been treated by mechanically disrupting the narrowed lower esophageal mechanism, which is currently achieved by surgical or endoscopic disruption of the muscle fibers. The choice of treatment is generally directed by patient preference and the risks associated with surgery.⁵ The risk of esophageal rupture and the risks related to use of general anesthetic have led practitioners to investigate less invasive options, such as Botox^® injections.

Clostridium botulinum, a soil-based, spore-forming bacteria, was originally identified as a pathogen that causes diffuse paralysis. A derivative form of the botulinum toxin with no systemic effects—Botox^®—was developed and used initially to treat strabismus.⁶ The subsequent use of Botox^® in muscular contraction disorders, for both cosmetic and functional reasons, has since become popular. In patients with achalasia, endoscopic injection of Botox^® inhibits acetylcholine release and thus reduces cholinergic excitation at the LES.⁷ Botox^® treatment is thought to have no long-term effects on the muscle fibers. The short-term efficacy of Botox^® injections (single sets of injections repeated as necessary) for the treatment of achalasia in children and adults has been reported in the range 77–83%.^{8, 9} The efficacy of long-term Botox^® treatment in patients who respond to initial treatment has been reported in the range 30–70%.^{9, 10, 11, 12}

The increased risk of mediastinitis associated with diabetes as seen in the case patient presents a problem for the treatment of high-risk patients with achalasia. Botox^® injections are indicated for high-risk patients (e.g. those suffering from diabetes, renal failure or coronary artery disease) because these patients are not considered candidates for surgery. Although mediastinitis after Botox^® treatment for achalasia is a rare occurrence, complications after injection of Botox^® into the esophagus must be considered as part of the treatment algorithm for esophageal disease, especially as the risk of complications is additive when multiple treatments are necessary. The high-risk subset of patients with achalasia might be better off facing the risk of surgery up front rather than facing the latent risk of complications after Botox^® injection, a risk that increases with each subsequent treatment.

Conclusions

The treatment of achalasia by Botox^® injection is a generally viable alternative to surgical or endoscopic management of the high-risk patient. Selection of patients continues to be the greatest obstacle when finding a place for Botox^® in the achalasia treatment algorithm. To find the best treatment option for each patient, a clinician must assess the extent to which the disease is affecting the patient's wellbeing, the type of recovery the patient would want, and finally, and most importantly, which procedure he or she can tolerate. Close observation of high-risk patients and a low threshold for hospital admission are essential for optimal patient care and the avoidance of potentially lethal complications. In addition, further study is required to determine which patients are truly high-risk and, therefore, require Botox^® treatment rather than the definitive treatment with either a laparoscopic Heller myotomy or endoscopic esophageal dilation.

References

1. Eaker EY et al. (1997) Untoward effects of esophageal botulinum toxin injection in the treatment of achalasia. Dig Dis Sci 42: 724–727 | Article | ChemPort |
2. Tintillier M et al. (2004) Esophageal abscess complicating endoscopic treatment of refractory gastroesophageal reflux disease by Enteryx injection: a first case report. Am J Gastroenterol 99: 1856–1858 | Article |
3. Schuman BM et al. (1987) Complications of endoscopic injection sclerotherapy: a review. Am J Gastroenterol 82: 823–830
4. Fitzgerald JF et al. (2002) Clinical quiz. Sinus tract between esophagus and fundus. J Pediatr Gastroenterol Nutr 35: 38, 98 | Article |
5. Andrews CN et al. (1999) Laparoscopic Heller's myotomy or botulinum toxin injection for management of esophageal achalasia. Patient choice and treatment outcomes. Surg Endosc 13: 742–746 | Article |
6. Jankovic J and Brin MF (1991) Therapeutic uses of botulinum toxin. N Engl J Med 324: 1186–1194 | ChemPort |
7. Pasricha PJ et al. (1995) Intrasphincteric botulinum toxin for the treatment of achalasia. N Engl J Med 332: 774–778 | Article | PubMed | ChemPort |
8. Hurwitz M et al. (2000) Evaluation of the use of botulinum toxin in children with achalasia. J Pediatr Gastroenterol Nutr 30: 509–514 | Article | ChemPort |
9. Kolbasnik J et al. (1999) Long-term efficacy of Botulinum toxin in classical achalasia: a prospective study. Am J Gastroenterol 94: 3434–3439 | Article | PubMed | ChemPort |
10. D'Onofrio V et al. (2002) Long-term follow-up of achalasia patients treated with botulinum toxin. Dig Liver Dis 34: 105–110 | Article | PubMed | ChemPort |
11. Neubrand M et al. (2002) Long-term results and prognostic factors in the treatment of achalasia with botulinum toxin. Endoscopy 34: 519–523 | Article | ChemPort |
12. Zaninotto G et al. (2004) Botulinum toxin injection versus laparoscopic myotomy for the treatment of esophageal achalasia: economic analysis of a randomized trial. Surg Endosc 18: 691–695 | Article | ChemPort |

Contact the journal about this article

Subject areas under which this article appears: Upper gastrointestinal tract | Endoscopy