Spontaneous rupture of the esophagus : report of two cases

Objective To study the diagnosis, prognosis and management of spontaneous rupture of the esophagus. Design: This is a retrospective study through the analysis of two cases with delayed diagnosis and subsequent treatment at the Track Surgery Service. Locale: The study was performed at the Thoracic Surgery Unit of the Hospital do Servidor Público Estadual Francisco Morato de Oliveira in the city of São Paulo. This is a specialized service. Participants: The two patients reported on had suffered spontaneous rupture of the esophagus. They were transferred to the Thoracic Surgery Unit because of the worsening of their condition in the previous institution which they had been admitted into. Measurement: The two patients with esophagus pleural fistula received similar treatment, initially advocated by Kanashin in Russia and Hauer-Santos in the United States, which consists of washing the fistula and using continuous pleural aspiration. Results: Although both patients had to spend a long period of time in hospital, their evolution was satisfactory with the treatment adopted, and the fistula closed. Conclusion: The authors conclude that the method of lavage of the mediastinum and continuous pleural aspiration, in patients who after spontaneous rupture of the esophagus developed a pleural esophagus fistula due to belated diagnosis, is an alternative and satisfactory therapy. Furthermore, in order to have the best outcome, an early diagnosis is recommended and thoracostomy as the surgical procedure, with primary suture.


INTRODUCTION S
pontaneous rupture of the esophagus was first reported in 1784 by Hermann Boerhaave.I This professor of medicine at Leyden was called to attend to Baron Johann Van Wassenaer -Great AdmiraI of the Dutch Fleet -who presented a severe gastric disorder after having taken part in a banquet.
To ease the Baron's malaise Boerhaave gave him an emetic.The Baron, after forceful vomiting, had severe pains in the chest.His cli~ical condition quickly worsened and he himself stated that the upper part of his stomach had ruptured and th~t, given the intensity of pain his demise was unavoidable.The Baron's prognosis was correct: he Address for correspondence: R. Pedro de Toledo, 1800 10º Andar -Ala Central V.Clementino São Paulo/SP -Brasil-CEP 04039-001 died 18 hours later.The autopsy, performed by Boerhaave himself, disclosed an ulceration on the distaI third of the esophagus and food residues in the pleural cavity.
Although the Boerhaave Syndrome is quite rare, it cannot be forgotten in the presence of a sev~re pain in the chest after a sudden increase of intraesophageal pressure, as is the case in the vomiting mechanism.
To delay diagnosisand treatment willlead to a h~gh rate of morbidity' and mortality.
The two cases managed recently at our Thoracic Surgery Unit describe the severity of this syndrome.

CASE 1
A.T.S.F., male, CaucasiaIi, 52 years of age, was admitted into the Emergency Ward of the hospital at 9.00 am on October, 15, 1994, coming from the southem coastal region of the State.
He reported that at 2:00 p.m. of the previous day, after having ingested a large amount of food and alcoholic beverages, he presented intense vomiting and, afterwards a severe pain in the chest.
He was taken to the local hospital where, in spite of analgesics administered, pain was not suppressed, and upon worsening of the clinicaI picture he was transferred to São Paulo.
At the admissional examination, the patient presented tachypnea and tachycardia (f 24 pm, FC 96 pm, AP 11 x 7).He had an emphysema in the left supraclavicular compartment.A reduced vesicular murmur in the left hemithorax was noted at auscultation.
A gasometry was taken and indicated normal.The.blood test measured Hb 14 g/%; Htc 44; GB: 3,500 with a deflection to the left.
Simple X-ray ofthe chest in anteroposterior position showed a hazy shadow on the left hemithorax.
One of the hypotheses was that of esophagic rupture, and an endoscopy was immediately requested which confmned a laceration in the left distaI third with presence of food residues.
The patient was immediately forwarded to the Surgery Room and submitted to a left posterolateral thoracostomy.Inventory of the cavity disclosed 300 ml of yellowish fluid with debris of fibrin.However, the mediastinum pleura presented intact, with outpouching, swollen and covered by fibrin.It was opened and let out a large quantity of food residues, that once removed disclosed a posterolateral esophagic laceration, extending for about 6 cm from the cardia.
After cleaning of the site, the lesion was closed with four simple cotton 4.0 sutures on a single plane, essentially intended to orient cicatrization along the suture line, as the wound was potentially infected.The pleural cavity was washed and a multiperforated drain was attached to the rib-vertebrae gutter, next to the lesion.After this, the patient was sent to the ICU.
On the fifth P.O. an esophagogram was performed, using lodate contrast, which disclosed an esophageal fistula in the distaI esophagus.
Although the patient was not unstable, at this moment he presented pleural-esophageal fistula and bilateral empyema (Fig. 2), bearing evidence of severe mediastinal infection.
The chosen treatment was that proposed by Hauer-Santos (2) which consists of an enteral probe, with its distaI end close to the fistula, infusing physiological saline solution to which was added a chloranphenicol solution.
The pleural drainage tube was kept under continuous suction, allowing for constant removal of the washout of the whole infected area.
Simultaneously a jejunostomy was performed to maintain' the nutritional condition.
The patient presented 'progressive improvement.On November 22, he was submitted to another esophagogram which disclosed closure of the fistula (Fig. 4).He was released on December, 9, 1994, in good physical condition.His clinicaI and radiological pulmonary condition was practically"normal (Fig. 5).

CASE 2
H.B. 68 years old, male, Adventist, was admitted on June 10, 1994, rnentioning that on the day before, after an abundant rneal he had forceful vomiting, followed by severe retrosternal pain spreading out to e back.He went to an ernergency day-care ward where an angina pectoris or a rnyocardial infarction were suspected, although not confirmed, after which the patient was dismissed.
Because of the persistent and increasingly severe pains, he carne to the Ernergency Attendance of the Hospital.A chest x-ray disclosed an effusion in the pleura on the right side with characteristics of posterior sepsis.An esophageal rupture was suspected: an esophagoscopy was requested, which confirmed a tear of about 4 cm at the distaI third.The patient was then submitted to surgery, pharyngostomy, block of distaI transit with a Foley probe, pleural drainage on the right side and jejunostomy for parenteral nutrition.
This treatment was continued for 20 days.There was, however, a progressive worsening of the overall condition with an intense toxemic picture, and it was decided to perform a thoracostomy, debridernent of the pleural space and drainage of the rnediastinurn.A wide-bore pleural drainage tube was introduced near to the esophageal fistula under continuous   aspiration.The margins of the scar were brought together by four stitches of non absorbable sutures to orient cicatrization.
The Foley probe was introduced again, roughly over the position of the fistula, for infusion of the physiological saline solution.
As after 10 days the washing out liquid was still purulent, chloranphenicol was added to the physiological saline solution ( 19 / 1,000 ml) dripping into the vein at 120mllhour.
After another ten days, the clinicaI picture had improved with a significant reduction of output deposits in the collecting bottle, showing a preferential flow to the stomach.An endoscopy undertaken at that time, evinced that the fistula had been reduced from 4 to 1.2 cm.
Treatment was continued for another 18 days and a new endoscopic examination disclosed an even greater reduction of the fistula's size, now down to 0.2 cm.On the 30th day of treatment the fistula was fully closed.

DISCUSSION
Spontaneous rupture of the esophagus consists of the full rupture of a segment of its wall, as a result of a sudden increase of the endo-Iuminal pressure.The other causes of rupture, such as those provoked by surgical instrumentation or perfora,ting wounds on the chest would be excluded by definition.
Post-prandial vomiting has been most frequently held responsible for this rupture, but nevertheless, the literature also mentions other ruling factors for the rapid increase of abdominal pressure such as coughing, labor and defecation.3 ,4.5 Currently, it is believed that the most likely mechanism involved is a malfunction of the overall vomiting reflexo The coordinated relaxation of the cricopharyngeal muscles will not take place, bringing about a high pressure chamber inside its lumen, mainly in the distaI third where rupture usually takes place.
Mostly, the rupture is located in the left posterolateral compartment of the esophagus, next to.the cardia, and can reach up to 6 cm in length.3 , 4 Occurrence of laceration in this site has been attributed, among other anatomical causes, to the entrance of vessels and nerves, to the angulation of the organ and to the absence of supportive tissue.4 According to Faber,6 the sudden increase of pressure would be an even more significant factor than the rate of increase itself.
Of the damages caused to the esophagus by such mechanisms, complete rupture orBoerhaave Syndrome is the most dramatic, although some of lesser severity have been described.
Mallory & Weiss, in 1929,7 reported a gastrointestinal bleeding associated with vomiting.The anatomopathological substrate consisted of shreds of the mucosa in the proximity of the cardia.
Williams in 1957,8 reported intramural dissection of post -emetic cause that did not lead to total rupture of the , esophagean wall, forming a hematoma within the muscle bundles (Fig. 6).
Until the mid-19th Century, the Boerhaave Syndrome was exclusively a finding at autopsies.In 1858, Mayer 9  would have been the first to reach a diagnosis prior to death.
Derbes & Mitche1l 4 attribute the first surgical treatment to Overholt in the United States, in 1943.However, it was Barrett in England, in 1947,10 who first published a paper on the therapeutical success of a surgical approach.
Boerhaave Syndrome has a higher incidence in the mal e gender, at an average of 50 years of age.II Its predisposing factors would be alcoholism, a neurological disease, a peptic ulcer, a hiatal hemia or arterial hypertension.12 However, many sick individuaIs do not present any' of these disorders.The classic clinicaI triad includes: vomiting, pain in the chest and subcutaneous emphysema.13 Other abnormalities exhibiting pain in the chest or the upper abdomen and vomiting may be confounded with the syndrome; among them, we can consider myocardiac ischemia, perforated peptic ulcer and acute pancreatitis, dissecting aortal aneurysm and aspiration pneumonia.
X-rays are an essential to 01 for correct diagnosis.Chest x-rays may show widening and emphysema of the mediastinum, hidrothorax, pneumothorax and lung condensations.Naclerio l4 described the "V" sign which consists of an irregular increase in pulmonary density on the left side of the cardiac contour, stressing chemical pneumonia next to the esophageal area.To ascertain the presence of a rupture, swallowing Iodate contrast will show the effusion.In case of doubt an endoscopic examination will confirm the hypothesis.Barrett 12, 10after his experience with the first case of surgical treatment by primary suture of the tear, claimed that only surgery would guarantee the patient's survival.
However, it has been.observedthat, to obtain positive results in this type of treatment, an early diagnosis is required.Its delay dramatically increases mortality rates.According to some authors, 15, 16after the first 24 hours, this rate would range from 20% to 50%.Tissue friability, as a result of necrotizing enzymes and infectious agents, would hamper cicatrization.However, in medicaI literature there has been no agreement regarding the time span for the performance of the primary suture.According to Samson 17the maximum delay for a rhaphy could not be for more than 15 hours.
In view of the above statements, various more complex procedures have been described intending to .reduce the possibility of dehiscence: repairs of sutures with synthetic flaps3 or with live tissues, such as the intercostal pedicle 18 of the pericardium,19 of the pleura,2oas well as protection of the ga~tric stump.21 Johnson, Schenegran & Kimby15recommended the functional exclusion of the esophagus through a cervicostomy and ligature of the distaI sphincter.More drastic surgeries, including esophagostomy22 were also recommended.
Nevertheless, so far no conclusion has been reached about how beneficiaI for the severely impaired patient those major surgical procedures indeed are.
The essential management for rupture of the esophagus consists of reversal of the chain of events triggered by the entrance of bacteria and active chemical matter inside the mediastinum, that may lead to an uncontrollable sepsis.23 As infection is not controllable, the laceration will not close.
A therapeutic approach, with continuous lavage of the rupture site, initially described in Russian literature 24 -26was accepted by other authors 2 ,27as a rather efficient therapy for the closure of the fistula.The ongoing removal of the necrotic matter and of the bacterial constituent using physiological saline solution and continued suction, evinced positive results in ruptures at alI esophageallevels.In our service, we have therefore systematized the procedures for the spontaneous rupture of the esophagus: thoracotomy, cleansing of the site, suture of the laceration (even many hours after the rupture), drainage of the cavity, and parenteral nutrition by jejunostomy.Should cicatrization not be successful, the pleuralesophagic fistula will be treated according to the work of Hauer Santos 2 : I. Placing. of a nasogastric probe (Dobbhoft) a little above the fistula; 2. Transpleural drainage of the mediastinum by a wide gauge drain, multiperforated and attached riext to the fistula; 3. Continuous dripping of physiologic serum with chloranphenicol (lg/IOOOml) at a 120ml/hour rate; 4. Continuous aspiration of the pleural cavity at a pressure of 25mmlh20 (Fig. 7).This procedure aims to avoid that the contaminated matter of the oropharynx, the active enzymes and the necrotic tissue retain mediastinal infection, which would deter closing of the fistula.
The final objective ofthese two reports is to enhance surgeons' awareness of the need for immediate repair of the lesion and, should an early.diagnosis not have been reached, introduce a simple altemative treatment with a positive outcome.

Figure 1 -
Figure 1 -Esophagram of patient A.T.S.F on 6th RO.Note the extravasation in the mediastinum contrast and in the pleural cavity due to suture dehiscence.

Figure 2 -
Figure 2 -Computerized tomography of the thorax of patient A:r.S.F on the 14th P.O.Note the bilateral pleural etfusion (emphyema) as a consequence of the mediastinitis.

Figure 3 -
Figure 3 -Patient A.T.S.F. on the 15th RO.Thoracic drainage with continuous aspiration, enteral probe forwashing out of fistula and jejunotomy.

Figure 4 -
Figure 4 -Esophagram of patient A.T.S.F.on the 28th.RO.There is no contrast overflowing.Fistula is closed.

Figure 5 -
Figure 5 -Plain X-ray of thorax of patient A.T.S.F.on the 30th RO. after fistula had closed.Almost fuI! resolvability of the bilateral emphyema.

Figure 7 -
Figure7-Treatment of esophagus rUpture by washing out of fistula.An enteral probe is located a little above the site of rupture, continually instilling physiological saline solution with antibiotics.A multiopening pleural drainage tube is fixed next to the fistula with constant aspiration.