The mortality rate for lower gastrointestinal perforation has been reported as 15.5–26.6% [1, 2, 12], similar to the rate in this study. This is probably largely attributable to the presence of numerous Gram-negative bacilli in the large intestine, so bacteremia easily arises following perforation of the lower gastrointestinal tract, and chemical transmitters such as interleukin (IL)–6 are induced, resulting in rapid onset of septic shock. This is considered to lead to multiple organ failure and acute circulatory failure. In abdominal emergencies requiring surgery, the frequency of lower gastrointestinal perforation is not particularly high [13, 14]. Identification of prognostic factors to improve survival rates for this pathology has long been a priority, due to the high mortality rate [15–17]. In addition, the severity of lower gastrointestinal perforation is considered to involve a large number of prognostic factors, and a scoring system is considered important for judging the preoperative condition more comprehensively. Methods for evaluating prognosis in patients with severe disease in general include APACHE II, the Simplified Acute Physiology Score III [18], and the Mortality Prediction Model 0 III [19] as overall evaluations of systemic severity, and multiple organ dysfunction score [20] and SOFA as an evaluation of multiple organ failure. The present study also showed significant differences in APACHE II from univariate analysis, and SOFA was an independent prognostic factor, confirming its usefulness as a severity assessment method. Moreover, Physiological and Operative Severity Score for Mortality and Morbidity (POSSUM) [21, 22] is available as a comprehensive evaluation of various organ functions and the degree of surgical invasion, indicating surgical risk. POSSUM is reported to be an excellent prognostic system even for colorectal peritonitis [23]. However, because these evaluation methods require a large number of items, they are overly complicated for cases of lower digestive tract perforation requiring emergency surgery and cannot be evaluated appropriately at all facilities. Identification of factors that can be evaluated quickly and easily is therefore necessary. In this study, low PMI and high SOFA score were independent prognostic factors in multivariate logistic regression analysis. PMI is a simple but reliable method of prognostic evaluation similar to SOFA. Body composition has recently been reported as a risk factor or prognostic factor for postoperative complications [3, 5, 24, 25]. In particular, skeletal muscle mass is considered important [26–28]. Among the methods of evaluating body composition that include skeletal muscle mass are bioelectrical impedance analysis (BIA), dual-energy X-ray absorptiometry, and CT/magnetic resonance imaging (MRI) using cross-sectional images. BIA and DXA methods show a strong correlationwith each other [29], as do the BIA and CT cross-sectional methods [30]. Each has its advantages and disadvantages. However, in the case of emergency diseases and gastrointestinal cancers, the CT method requires no additional examinations and is simple and quick if images from preoperative examinations are used. Furthermore, with CT methods, one method uses total skeletal muscle area at the L3 level and another measures only the psoas muscle area, but the results are reportedly correlated [30]. In the case of emergency diseases, measuring only psoas may be better in terms of simplicity and speed. The skeletal muscle system accounts for about 40% of the adult body volume, and around 88% of muscle is protein, which represents 50% of total protein in the body. The muscles function as a nutrient storage system, playing the role of distributing amino acids to each organ as a biological defense reaction during invasion. However, in patients with sepsis, active nutrition cannot prevent loss of the body protein compartment despite increases in body fat [31]. In addition, inflammatory cytokines such as IL–6 [32] and tumor necrosis factor α [33] promote proteolysis. Therefore, when extravasation results from emergency surgery under conditions of severe infection, a greater original muscle mass is advantageous for tissue repair as a defense reaction, and organ failure can be avoided. Preoperative skeletal muscle mass is a prognostic indicator in lower gastrointestinal perforation. On the other hand, many reports [34, 35] have identified the CT attenuation value as a useful index of muscle quality in cancer patients, because the increase in non-contractile tissue including fat in muscles decreases the CT attenuation value. In this study, CT attenuation was not a prognostic indicator. Under pathological conditions such as lower gastrointestinal perforation, where damage to the body is largely caused by the release of inflammatory cytokines, muscle mass is considered important regardless of muscle quality. In addition, none of visceral fat mass, subcutaneous fat mass or the ratio of those two values represented prognostic indicators. Muscle mass was considered more important in the acute phase than fat. Survivors with low PMI tended to have high DIC score, high DIC rate, and high postoperative complication rate, and additionally had significantly longer duration of hospitalization. It has been suggested that if skeletal muscle mass is reduced, later treatment will be difficult even if the patient survives. However, no significant difference in treatment methods was seen between the two survival groups.
Postoperative intensive care is important in lower gastrointestinal perforation. In intensive care, blood purification treatments such as CHDF and PMX-DHP are available, along with thrombomodulin alfa (a genetical recombination) as pharmacotherapy. No specific opinion has been obtained regarding prognostic improvements from these treatment modalities, and no clear criteria for the introduction of blood purification therapy have been established. Preoperative medical interventions for skeletal muscle loss are not possible with lower gastrointestinal perforation. To improve the survival rate, early introduction of aggressive blood purification and medication may be necessary in the muscle loss group. In this study, the number of cases was limited because of the single-center design, and the ability to conduct sufficient studies appears limited. To improve the number of cases in the future, multiple-center studies appear necessary. Furthermore, changes in the effects of treatment according to differences in skeletal muscle mass should be accurately examined in prospective studies.