There is an overall high complication rate after therapy of oesophageal perforation, both after surgical and after conservative therapy.22, 24, 25, 26, 27 A correlation between an inhomogeneity of therapies and the complication rate can be assumed.
Most studies show case reports or small cohorts of about 10 patients at most, because esophageal perforation occurs rarely. One study described oesophageal perforations of 29 patients, and there is a meta-analysis of 75 studies, but their comparison proves difficult because there was no grading of perforation. Heterogeneous survival rates are evident.2, 25
As described, there is a significant correlation between the age of injury, the clinical type of severity and the lethality. Thus, a prediction of mortality at diagnosis is possible.
The different manifestations of esophageal perforations can be classified into severity grades by listing the concomitant symptoms and anatomical presentation of the perforation as described above. There is clearly a significant correlation between the severity of perforation and lethality. The higher the degree of oesophageal perforation, the lower the patient's probability of survival.
The increased homogeneity of the patient collective of this study allows transferability of the results to other cases and clinical application of the presented grading of oesophageal perforation e.g. for further investigations concerning lethality of esophageal perforations.
The general decision shown in literature for or against conservative or surgical therapy was surgeon- and case-dependent. Therapeutic conservative options are gastric tube, intestinal stenting and endovac therapy in combination with systemic antibiotics.22 Surgical options are suturing the perforation, a plastic cover, e.g. with pleura, pericardium, gastric fundus, omentum majus or covering with a pedicled muscle flap.25, 27
The review of the previous literature shows a very heterogeneous distribution of therapeutic strategies due to the lack of both a grading of the perforation and a proposal as a guiding structure for the therapeutic decision. 3, 4, 7, 9, 10, 11, 13, 15, 17, 18, 19, 21, 22, 26
There was no grade classification in any of the studies. Comparability of the studies is therefore not given.
The in Table 4 presented classification of oesophageal perforation now allows a comparison of patients`collectives according to the degree of perforation and eventually lead to a therapy recommendation.
Table 4
Classification and therapy.
Type | Endoscopic-radiological | Therapy |
I | covered perforation with air in the mediastinum, no evidence of mediastinitis or sepsis | conservative therapy, gastric tube, if necessary control with gastrografin swallow after 8h |
II | fresh perforation with radiological leakage or retention without evidence of mediastinitis or sepsis | intestinal stent, endovac if necessary, antibiotic therapy, control with gastrografin swallow after 8h |
IIIa | cervical perforation, possibly mediastinitis | surgical reconstruction with suturing, internal and external drainage, antibiotic therapy, endovac if necessary |
IIIb | thoracic perforation or pleural empyema or mediastinitis | surgical sanitation, internal and external drainage, antibiotic therapy, endovac if necessary |
IIIc | perforation of the cardia or mediastinitis or peritonitis | surgical suturing and fundoplication, internal and external drainage, antibiotic therapy, endovac if necessary |
IV | older perforations with persistent fistula, mediastinitis, possibly oesophageal necrosis, possibly long-stretch rupture or possibly sepsis | oesophagectomy, two-stage reconnection, internal and external drainage, antibiotic therapy |
Type I: The oesophageal perforation is covered. There is air in the mediastinum, but no evidence of mediastinitis or sepsis. In our patient collective we indicated a conservative approach. A gastric tube was inserted immediately after diagnosis. The inflammatory parameters have always been checked regularly during the course of the procedure, and a gastrographin swallow was performed eight hours after diagnosis in order to detect a leakage.
Type II: The perforation is fresh with radiologically confirmed leakage or mediastinal retention. There is no evidence of mediastinitis or sepsis at this time. Grade II perforations were treated with an intestinal stent; endovac therapy was sometimes necessary. Furthermore, a calculated antibiotic therapy was started immediately after diagnosis. If necessary, the antibiotic therapy was changed after the antibiogram was available. X-ray control check with a water-soluble contrast agent was done after 8 hours.
Type IIIa: There is cervical oesophageal perforation. Mediastinitis may be present. In the case of a grade IIIa perforation, cervical suturing was performed, if necessary with a sternocleidomastoid flap. An internal and an external drainage in some cases an endovac was placed. Antibiotic therapy was calculated immediately from the time of diagnosis, as soon as possible according to antibiogram.
Type IIIb: There is a thoracic oesophageal perforation. Pleural empyema or mediastinitis may be present. Surgical repair was performed for a grade IIIb perforation. A thoracoscopic suturing was covered with pericardium, an intercostal or latissimus flap. An internal and an external drainage/ endovac was placed. Pleural empyema was cleared if present. Antibiotic therapy was indicated from the time of diagnosis.
Type IIIc: The perforation appears in the most distal part, at the level of the abdominal oesophagus and cardia. Peritonitis may be present. Because grade IIIc perforation occurs very distally, the perforation was sutured primarily and covered with a fundoplication. An internal and external drainage/ endovac was placed. Antibiotic therapy began immediately calculated from the time of diagnosis, and was adjusted if necessary after an antibiogram.
Type IV: There is an older perforation with persistent fistula. Mediastinitis is also present. Sepsis, long-stretch rupture or esophageal necrosis may also be present. Oesophagectomy with two-stage reconnection is recommended for grade IV perforation. Antibiotic therapy was calculated immediately, in the following according to antibiogram.
For a long time, esophageal perforation was not classified, and there were no standard treatment guidelines. Therapy was characterized by a lack of homogeneity. Through this study, a uniform classification is evident, as homogeneity in the rare clinical picture has now been created. Lethality can hence be predicted at the time of diagnosis.
For further evaluation and as an outlook of research, it would be of interest to classify the known studies and case reports into the classification of oesophageal perforation and conduct a prospective multicenter study in order to evaluate the feasibility and impact of this classification on the daily routine.