|
|
Robin McKenzie, M. D.
09-29-2009
- No incidence/prevalence data from Africa, but would expect declining incidence as ddI and d4T are being phased out in Zambia.
- In the referenced paper from Uganda, of ~1000 pts with CD4 < 250 started on ART and followed for ~18 months, < 0.5% developed acute pancreatitis.
- TB may be a common etiology in Zambia.
- Clinical presentation and elevated enzyme levels are key to Dx in Zambia, since imaging not widely available.
- Initial therapy (as below) is same in Zambia.
Zambia Information Author: David Riedel, M.D.
- Increased incidence in HIV+ pts, esp. with ddI treatment, previous pancreatitis, or CD4<200.
- Medications are most frequent cause in HIV+ pts:esp. ddI and to lesser extent other "d-drugs" which may cause mitochondrial toxicity w/ lactic acidosis, hepatic steatosis +/- pancreatitis.
- Increased risk with ddI+ certain other drugs: d4T, TDF, ribavirin, hydroxyurea (HU), pentamidine. (Do not use ddI+ d4Tin pregnant women, and avoid if possible in all pts. Use reduced dose of ddIwith TDF.)
- Less common: 3TC (children), IV/aerosolized pentamidine, TMP-SMX, INH, rifampin, corticosteroids, erythromycin, valproic acid.
- Other causes: hypertriglyceridemia (fasting TG >1000 assoc. with LPV/rand other PIs), alcohol, gallstones, post-ERCP, HIV (acute HIV infection), OIs (CMV, toxoplasmosis, MAC, TB, PCP, cryptosporidiosis).
- Hx: Acute, constant, upper abdominal pain that may radiate to back; nausea +/- vomiting. Presentation varies from no symptoms to shock/coma.
- PE: Fever, tachycardia, epigastric tenderness. Hemorrhagic discoloration of flanks (Grey-Turner's sign) or periumbilical area (Cullen's sign) in ~1% cases.
- Increased amylase nonspecific. Common in asymptomatic HIV infection and other conditions: renal insufficiency, intestinal or fallopian tube disease, macroamylasemia, acidosis, various medications. Usually >3x normal in pancreatitis, but may be normal.
- Increased lipase (>3x normal) more specific. Mildly increased lipase also common with asymptomatic HIV, renal failure, bowel disease, DKA, macrolipasemia, medications. Lipase remains elevated longer than amylase.
- Contrast-enhanced CT useful to R/O other Dx and stage severity of pancreatitis. Not indicated for mild disease unless Dx uncertain. Necrosis (lack of enhancement) best identified several days after presentation; % glandular necrosiscorrelates with mortality rate. MRI more sensitive for mild pancreatitis and more specific for categorizing fluid collections as necrosis, abscess, hemorrhage or pseudocyst; but less clinical correlation available with MRI.
- If cause not obvious (medications, EtOH) consider U/S, MRCP, or endoscopic U/S to detect gallstones or biliary tract disease. ERCP may be needed to detect/remove gallstones and obtain fluid to stain and Cx for OIs (more likely if CD4 low).
- CT- or U/S-guided FNA indicated if infected necrosis suspected.
- Stop ddI or other causative medications
- Give IV fluids. (Avoid hemoconcentration)
- Control pain. (Often requires narcotics.)
- Infection probably due to translocation of gut flora. Infection rate proportional to extent of necrosis.
- No prophylactic ABx for mild pancreatitis: may cause superinfection with resistant bacteria and Candida.
- For acute, severe pancreatitis (>30% necrosis by CT), previous studies supported use of imipenem 500mg IV q6h as prophylaxis.More recent trials do not support this recommendation. Use of prophylactic AB xfor necrotizing pancreatitis not recommended(see refs). Instead, if infection suspected, obtain aspirate for gram stain and Cx, begin empiric ABx, and continue ABX only if Cxpositive.
- Infected necrosis usually presents in 2nd or 3rd week. Both sterile and infected necrosis cause abd pain, fever, leukocytosis. CT- or US-guided aspiration needed for Dx.
- Choice of ABx for infected necrosis ideally based on Cx results. GI flora (aerobes and anaerobes) often present. ABx with good penetration: carbapenems (imipenem, meropenem, ertapenem, doripenem), fluoroquinolones (ciprofloxacin,levofloxacin, mosifloxacin), ceftazidime, cefepime, metronidazole, clindamycin, fluconazole. For empiric Rx consider carbapenem alone or fluoroquinolone + either metronidazoleor clindamycin.
- If OI suspected or Dx'd, see appropriate module for management.
- Mild pancreatitis: start clear liquids early and advance diet as tolerated.
- Severe pancreatitis: begin enteral nutrition as soon as possible. NJ feeding usually used, but NG as safe in recent smallstudy (Eatock) .
- Give parenteral nutrition only if adequate enteral feeding not tolerated.
- Calculate APACHE II score on admission: >8 in HIV- pt and >9-14 in HIV+ pt predicts severe course.
- Ranson criteria (table) less predictive in HIV+ pts: >3 in HIV-negative pt and >4 in HIV+ pt suggests severe course.
- Obtain contrast-enhanced CT unless pancreatitis is mild. Necrosis best detected 2-3 days after admission. Necrosis >30% predicts mortality >20%.
- Measure CRP: >21 mg/dL on day 2-4 and >12 end of first week predict severe pancreatitis.
- Early deaths (first 2 wks) primarily due to multisystem organ failure.
- Later deaths due to local and systemic infection.
- Infected necrosis must be Dx'd (CT-guided aspirate) and drained (percutaneously or surgically).If blood Cxs negative and no evidence of sepsis, drainage may be delayed to allow organization of necrotic area.
- Early ERCP indicated for severe gallstone pancreatitis. Cholecystectomy usually performed before discharge in mild cases or within a few months in complicated cases.
- Monitor triglycerides in pts taking PIs.
- Dellinger EP, Tellado JM, Soto NE, et al.;
Early antibiotic treatment for severe acute necrotizing pancreatitis: a randomized, double-blind, placebo-controlled study.;
Ann Surg;
2007; Vol.
245; pp.
674-83;
ISSN:
0003-4932;
PUBMED: 17457158
Rating:
Basis for recommendation
Comments:This is second randomized, double-blind, placebo-controlled trial of prophylactic antibiotics for severe acute pancreatitis. In this trial 50 patients received meropenem, and 50 received placebo. As in previous blinded, placebo-controlled study, there was no difference in rate of infection or death.
|
|