The Natural History of Untreated Penetrating Diaphragm Injury: An Animal Model

Abstract

Surgical repair of diaphragm injuries is rather straightforward, but diagnosis can be difficult. The natural history of diaphragm injury is quite controversial. Undoubtedly, many diaphragmatic injuries are missed at the initial evaluation. Some theories state that diaphragm injuries do not heal, and that all eventually lead to herniation. Current theories regarding uniform herniation of all diaphragm injuries are not supported by animal models. The authors developed a penetrating diaphragm injury model to study the natural history of this injury. This study used 48 male Sprague-Dawley rats weighing 300-425 g. During the study, 24 of the rats received a 16-gauge needle puncture of the left diaphragm mimicking a small penetrating wound, whereas 24 of the rats received an injury produced with a 2.7-mm aortic punch, which created a defect comparable to a much larger penetrating wound. Half of the animals in each group were euthanized at 1 month, and the other half at 10 months. This allowed short- and long-term follow-up of the injuries. Gross inspection of the left diaphragm was performed after the animals were sacrificed via a thoracoabdominal incision. All diaphragms then were removed for examination. All the rats experienced perioperative recovery. None of the 24 rats with a 16-gauge needle injury had an injury at 1 month (n = 12) or 10 months (n = 12). No patent injury was noted in the aortic punch injury group (n = 12) sacrificed at 1 month. At 10 months, 1 of 12 animals had a small hepatic herniation through the aortic punch injury. All the injuries displayed adhesions to the underlying left hepatic lobe. The authors developed a penetrating diaphragm injury model to understand better the natural history of this injury. Spontaneous healing occurred in 98% of the animals. In this animal model, because the left lobe of the liver is present beneath the left diaphragm, healing without herniation usually occurs. A role may exist for nonoperative treatment of human right diaphragm injuries in clinical practice. This animal model may prove useful in further defining future management for these injuries.