Monte Minute‎ > ‎Monte Minute‎ > ‎

Tuberculous Peritonitis - 11/23/2012

posted Nov 23, 2012, 2:20 PM by Monique Tanna   [ updated Dec 27, 2012, 6:46 AM by Purnema Madahar ]
Today's CRS described a 34 year old male originally from Mexico who presented to clinic with 5 days of abominal pain, worst in the epigastrium and RUQ.  Pt also reported nausea, vomiting and decreased appetite for 1-2 weeks, along with several loose stools per day.  He had subjective fevers and chills for 5 days, as well as increased abdominal girth.  He had a history of active EtOH (14-15 beers daily) and cocaine use. 

Exam revealed a temperature of 100.6, HR 120, BP 126/77, a thin male with jaundice, scleral icterus, cervical/supraclavicular LAD, distended abdomen with dullness to percussion, hepatomegaly, tenderness of RUQ and epigastrium.  Labs: Na 130, wbc 6.9, h/h 14.4, plt 168, amylase/lipase 53/67, albumin 3.1, bili's 3.7/1.8, AST/ALT 107/57, alk phos 147, etoh < 10, paracentesis: SAAG 0.9, wbc 960 52% lymphs.  Imaging: new large ascites (compared to 4/2011), R pleural effusion, R base infiltrate. 

The patient was initially treated for SBP & CAP but had persistent fevers, and developed bilateral pleural effusions.  Thoracentesis was done, showing an exudative effusion with 2750 wbc, 92% lymphs.  Repeat paracentesis: SAAG 0.7, wbc 580 with 90% lymphs.  AFB sputum neg, ascites AFB neg.  HIV negative, PPD negative, quantiferon positive.  FNA of L cervical nodes: necrotic tissue with granular debris (nonspecific but can be seen in TB).  Pleural biopsy: reactive changes, acute inflammation and edema.  Pt was empirically treated with RIPE for extrapulmonary TB peritonitis and rapidly improved and defervesced.  Pleural and ascitic fluid cultures eventually grew mycobacterium tuberculosis.

Attached is a review on tuberculous pleural effusions (although our patient had more than just TB pleuritis), as well as paper on the ADA test describing its sensitivity and specificity (courtesy of Zach Rosner!)
Update on Tuberculous Pleural Effusion
Light RW. Update on tuberculous pleural effusion. Respirology. 2010 Apr;15(3):451-8. Epub 2010 Mar 21.

Diagnostic Accuracy of ADA Test
Liang QL, Shi HZ, Wang K, Qin SM, Qin XJ. Diagnostic accuracy of adenosine deaminase in tuberculous pleurisy: a meta-analysis. Respir Med. 2008 May;102(5):744-54. Epub 2008 Jan 28.

Some comments from Dr. LeFrancois: Note meta-analysis paper cutoff definition for positive ADA test in its cited 63 papers varies but in all cases was >27 IU/L (the ADA value from pleural fluid in our CRS case from earlier today)!  In purusing table 1 of article it looks like 30IU/L was lowest cutoff chosen to define positive in many of the studies.   It seems to me that forest plot provided in the meta-analysis reports sens and spec for definition of positive that was used in each paper, and plot does not (unfortunately) give accuracy information for alternative cutoff points (like 27 IU/L). 

However, from glancing through table 1, even those papers that chose a relatively low cutoff (e.g. 30IU/L) there seems to be decent specificity so perhaps a value of 27IU/L gets the PPV up there a significant bit in our patient....although yes, of course, not a complete slam dunk.  I think the meta-analysis more actually pushes the alternative point that a negative ADA does still leaves a 10% FN rate which is pretty high when talking about TB pleurisy and the consequences of missing that dx.
Of course Elena's point (at CRS) that even if positive test with its decent specificity and significant PPV leaves unanswered questions regarding pathogen resistance data (not an insignificant issue given modern era of MDR TB!) and therefore the ADA as a dx test for this reason and others is not a panacea in the world of TB pleurisy!