Oral Feeding in Necrotizing Pancreatitis

34

G. Pupelis et al.

Acta Chir Belg, 2014, 114, 34-39

Oral Feeding in Necrotizing Pancreatitis G. Pupelis1, H. Plaudis1, K. Zeiza1, N. Drozdova1, M. Mukans1, V. Boka1 1

Department of General and Emergency Surgery, Riga East University Hospital, Riga, Latvia.

Abstract. Background : Restoration of gastrointestinal function is a crucial determinant of favorable outcome in severe acute pancreatitis (SAP). The purpose of this study was to retrospectively review our experience with early oral feeding in patients with the necrotizing form of SAP. Over the last 10 years, we have routinely gradually increased oral feeds in order to restore gastrointestinal function. Methods : Early low volume oral (ELVO) feeds containing 248-330 kcal/daily were routinely provided for all patients to help stimulate gastrointestinal function. Patients who received ELVO feeding within 72 hours of admission were ­allocated to Group I ; those who received ELVO feeds after 72 hours were allocated to Group II. The volume and ­calories of the feed, magnitude of systemic inflammation, levels of C-reactive protein (CRP) and lipase, incidence of organ dysfunction, main outcomes, and complications were analyzed. Results : In total, 129 patients received ELVO feedings. The mean CRP level on day 7 was 160 ± 77.6 mg/l in Group I compared to 200.2 ± 103.2 mg/l in Group II, p = 0.043. Normalization of CRP below 100 mg/l was observed on day 14 in both groups. The rate of infection and the need for surgical intervention (46.8% vs. 26%) were significantly higher in Group II (p = 0.026). Group II also had longer ICU/ hospital stays (p = 0.039/p = 0.002). Overall mortality was 10%. Conclusions : ELVO feeding provides physiologic stimulation and promotes recovery of bowel function, preparing the gastrointestinal tract for low-fat hospital food in patients with necrotizing SAP. The majority of patients required no ­additional nutritional support.

List of abbreviations APV, artificial pulmonary ventilation CECT, contrast enhanced computed tomography CRP, C-reactive protein ELVO, early low volume oral ERCP, endoscopic retrograde cholangiopancreatography ICU, intensive care unit IQR, interquartile range MODS, multiple organ dysfunction syndrome SAP, severe acute pancreatitis SD, standard deviation SIRS, systemic inflammatory response syndrome SOFA, sequential organ failure assessment

The clinical course and prognosis of SAP largely depends on the host’s ability to localize tissue destruction and balance the systemic inflammatory response. Restoration of gastrointestinal function is required in order to control the inflammatory process, and enteral nutrition plays an important role in the conservative treatment of acute pancreatitis (1, 2). The positive effect of 7-day complete enteral nutrition on acute phase responses in patients with acute pancreatitis has already been demonstrated. This treatment was not associated with significant changes in repeated contrast-enhanced computed tomography (CECT) scans, thus documenting the safety of enteral nutrition (3). A non-functioning gastrointestinal tract is the main obstacle in providing goal feeding, in

5057-pupelis-.indd 34

which calculated volumes of enteral formula are given daily and subsequently increased each day. Unfortunately, some degree of gastrointestinal dysfunction is common during the early course of acute pancreatitis. Several guidelines recommend enteral nutrition (4, 5, 6) ; however, poor compliance with the recommended principles in intensive care unit (ICU) settings is also reported (7, 8, 9). The aim of this study was to retrospectively review our 10 years of experience with the routine application of early oral feeding for restoration of gastrointestinal function in patients with necrotizing SAP. Material and Methods The current study was a continuation of the pilot study carried out to assess if early oral feeding with a whole protein feeding formula and early supplementation with low-fat low caloric hospital food can be an alternative to tube feeding in patients with acute pancreatitis when ­gastro-enteric transit is not severely impaired (10). Of the 309 patients with necrotizing SAP treated in our institution between September 2001 and December 2011, we selected 129 patients admitted with the first or a new ­episode within a 72-hour period from the onset of the disease. These patients received treatment according to our protocol. The main purpose of conservative therapy in the early phase of severe acute pancreatitis was provision of organ support, active recovery of tissue perfusion,

30/01/14 10:48

35

Oral Feeding in Necrotizing Pancreatitis and reduction of fluid sequestration in the third space. The treatment consisted of isovolaemic haemodilution with early adequate colloid transfusion and early continuous veno-venous haemofiltration when indicated. In all cases, the necrotizing forms were diagnosed by CECT scans at the end of the first week of treatment to avoid false positive results associated with early CECT. All patients treated with the conservative therapy received ELVO feeding with whole protein enteral formula or low-fat natural yogurt providing the median of 300350 ml/248-330 kcal daily as part of the treatment protocol since their gastro-enteric transit was not severely impaired. After the test bolus consisting of 20 ml of standard feeding formula, the feeding was continued at a rate of 20 ml every two hours, increasing the feeding rate according to individual tolerance. After the first 12-hour period of successful introduction of the feeding, low-fat natural yogurt and oatmeal soup were the next feed administered as a supplement. At the end of the week and according to individual tolerance, patients were transferred to low-fat hospital food consisting of mashed potatoes, mashed vegetable soup, mashed chicken breast, and white fish. A subgroup of patients was started directly on low-fat natural yogurt as the test bolus, continuing the feeding according to the same principles as above. Patients did not receive any other form of nutritional support, including parenteral nutrition. All patients were divided into two groups to assess whether the time of commencement of the feeding influenced the outcome. Patients in the first group (Group I) received oral feeding within 72-hours from admission ; patients in the second group (Group II) received oral feeding later than 72-hours from admission. The main variables characterizing the nutritional support included mean caloric intake and feed volume provided. We analyzed evidence of the systemic inflammatory response (SIRS), organ dysfunction according to sequential organ failure assessment (SOFA) scores and incidence of multiple organ dysfunction syndrome (MODS). The magnitude of SIRS was assessed evaluating the dynamics of CRP level and lipase activity in the blood. Among the inclusion criteria was maximal CRP ≥ 150 mg/L. The primary endpoints of this study were mortality and complication rate. The secondary endpoints were infection rate, incidence of surgical interventions, length of stay in the ICU and length of hospital stay. Two types of peripancreatic infections were defined : 1. Primary infection – when patients received only conservative treatment and it failed to prevent sepsis ; 2. Secondary infection (drain related infection) –contamination of the necrotic tissue and fluid collections was a consequence of the early operation or percutaneous drainage of noninfected collections. The authorization for the study was obtained from the Ethical Committee of Riga Stradins University. Data distribution was analyzed using the Kolomogorov-Smirnov

5057-pupelis-.indd 35

test. Continuous data were presented in mean values with standard deviation (SD) or median values with interquartile ranges (IQR). Statistical comparison was performed with a non-parametric method, using the Mann-Whitney U Test, and parametric data comparison was done using the Independent samples t-Test. Categorical data were analyzed with chi-square and Fisher’s exact test. Statistical significance was considered at the p value level of ≤ 0.05, with a confidence interval of 95%. Multivariate analysis was provided to identify the relationships between variables and their relevance to the oral feeding and main outcomes. Statistical analysis was done on SPSS version 17.0. Results ELVO feeding was provided to 129 patients who suffered necrotizing SAP and were admitted with the first or a new episode of the disease within a 72-hour period from the onset. Male patients were the majority in both groups, and the most common etiologic factor was alcohol. The main demographic data are displayed in Table 1. In 50 patients, ELVO feeding was started within a 72hour period after admission (Group I) ; in 79 patients, it was started more than 72 hours after admission (Group II). Severity assessment of the early phase of the disease revealed that the incidence of systemic inflammatory response and organ dysfunction was similar in both groups ; however, on day 7, the SOFA score was significantly lower in patients who received early oral nutrition, p = 0.023. At the same time, the need for renal replacement therapy in the form of continuous veno-venous haemofiltration was observed significantly more often in Group II, p = 0.017. More than half of the patients from both groups presented with pleural exudate. That corresponded to a higher need for mechanical ventilatory support in Group II, although not reaching a significant difference, Table I. There were no statistically significant differences in the incidence of renal (23 vs. 47, p = 0.150), pulmonary (24 vs. 32, p = 0.205), liver (20 vs. 25, p = 0.349) and metabolic (8 vs. 22, p = 0.139) dysfunctions in the early phase of disease. ELVO feeding was started notably earlier in Group I, at a mean of 1.8 ± 1 days from admission. The mean oral feeding volume provided on day 1 was similar in Group I and Group II (190.2 ± 124.2 ml vs. 235 ± 164 ml, respectively). Patients in both groups tolerated the feeding well, and on day 9 received a mean of 551.7 ± 234.7 ml vs. 479 ± 281 ml accordingly. The median daily orally provided calorie supply during these nine days was 248 (116.5-495) kcal in Group I and 330 (150-500) kcal in Group II, Table 2. This mode of treatment was sufficient enough to achieve balanced normalization of systemic inflammation in both groups ; however, it was more ­stable in patients who started receiving the feeding

30/01/14 10:48

36

G. Pupelis et al. Table 1. — Patient characteristics and dynamics of SOFA score. All patients n = 129

Time of commencement of oral feeding from admission

Male / Female = 96 / 33

Group I within 72 hours n = 50

Group II later than 72 hours n = 79

p-value

21/42 13/26 1/2 15/30

37/46.8 18/22.7 7/8.8 17/21.5

0.717 0.678 0.150 0.301

Etiology: Alcohol, nr./% Gallstones, nr./% ERCP*, nr./% Other, nr./%

SIRS, nr.,/% MODS, nr.,/% CVVH, nr.,/% APV, nr.,/% Pleural exudate, nr.,/% SOFA day 1, points (IQR) SOFA day 3, points (IQR) SOFA day 7, points (IQR)

49/98 45/90 24/48 6/12 34/68 2 (1-3) 3 (2-5) 1 (0-2)

76/96.2 69/87.3 55/69.6 18/23.4 57/72.2 2 (0-3) 3 (2-5) 2 (1-4)

0.999 0.781 0.017 0.165 0.999 0.784 0.687 0.023

*ERCP – endoscopic retrograde cholangiopancreatogaphy.

Table 2. — Provided volume and calories.

Mean start of feeding, days ± SD Mean start of feeding, hours ± SD Median volume per day, ml (IQR) Mean volume day 1, ml ± SD Mean volume day 3, ml ± SD Mean volume day 5, ml ± SD Mean volume day 7, ml ± SD Mean volume day 9, ml ± SD Median kcal per day (IQR) Mean kcal day 1 ± SD Mean kcal day 3 ± SD Mean kcal day 5 ± SD Mean kcal day 7 ± SD Mean kcal day 9 ± SD

Group I n = 50 1.8 ± 1 44.1 ± 23.2 300 (192-525) 190.2 ± 124.2 378.4 ± 267.7 502.1 ± 276.9 451 ± 336.9 551.7 ± 234.7 248 (116.5-495) 152.4 ± 114.8 286.4 ± 213.3 372.1 ± 242.9 392.2 ± 311 516.6 ± 250.7

e­ arlier. Mean CRP on day 7 was 160 ± 77.6 mg/l in Group I compared to 200.2 ± 103.2 mg/l in Group II, p = 0.043. Normalization of CRP below 100 mg/l was observed in both groups on day 14, Table 3, Figure 1. The lipase ­level, an indirect marker of the inflammatory response and the direct marker of the response to feeding was ­stable within one week, Table 3. The late phase of the disease was characterized by ­infection and the need for surgical treatment, including debridement and drainage of the necrotic area and the infected fluid collections. The overall infection rate was significantly higher in Group II (46.8%, n = 37) compared to Group I (26.0%, n = 13 patients), p = 0.026. Those who were fed early had less primary infection and did not require early invasive manipulations or early surgical interventions. The number of patients infected sec-

5057-pupelis-.indd 36

Group II n = 79 5.7 ± 2.1 135.7 ± 48.9 350 (172.5-500) 235 ± 164 336 ± 219 531 ± 334 464 ± 259 479 ± 281 330 (150-500) 220 ± 158 324 ± 217 522 ± 343 453 ± 266 480 ± 403

p-value