Utility of the Ratio between Lactate Dehydrogenase (LDH) Activity and Total Nucleated Cell Counts in Effusions (LDH/TNCC Ratio) for the Diagnosis of Feline Infectious Peritonitis (FIP)

Abstract

Background: We tested the hypothesis that the ratio between lactate dehydrogenase activity (LDH) and total nucleated cell counts (TNCC) in effusions may be useful to diagnose feline infectious peritonitis (FIP).

Methods: LDH/TNCC ratio was retrospectively evaluated in 648 effusions grouped based on cytology and physicochemical analysis (step 1), on the probability of FIP estimated by additional tests on fluids (step 2) or on other biological samples (step 3, n = 471). Results of different steps were statistically compared. Receiver Operating Characteristic (ROC) curves were designed to assess whether the ratio identify the samples with FIP "probable/almost confirmed". The cut-offs with the highest positive likelihood ratio (LR+) or Youden Index (YI) or with equal sensitivity and specificity were determined.

Results: A high median LDH/TNCC ratio was found in FIP effusions (step1: 2.01) and with probable or almost confirmed FIP (step 2: 1.99; 2.20 respectively; step 3: 1.26; 2.30 respectively). The optimal cut-offs were 7.54 (LR+ 6.58), 0.62 (IY 0.67, sensitivity: 89.1%; specificity 77.7%), 0.72 (sensitivity and specificity: 79.2%) in step 2 and 2.27 (LR+ 10.39), 0.62 (IY 0.65, sensitivity: 82.1%; specificity 83.0%), 0.54 (sensitivity: 82.1%; specificity 81.9%) in step 3.

Conclusions: a high LDH/TNCC ratio support a FIP diagnosis.

Keywords: cat; clinical chemistry; diagnostic accuracy; effusion cytology; feline coronavirus (FCoV).

Figure 1

Flowchart summarizing the selection of the caseload according to the exclusion criteria.

Figure 2

LDH/TNCC ratio recorded in effusions classified based on cytological and physicochemical classification ((A), step 1), in terms of probability of FIP based on the subjective interpretation of cytological and physicochemical analysis of the effusions ((B), step 2) or on the cytological and physicochemical analysis of the effusions associated with additional laboratory data ((C), step 3). The boxes indicate the first to third interquartile range (IQR), the horizontal lines indicate the median value, whiskers extend to further observation within the first quartile minus 1.5 × IQR or to further observation within the third quartile plus 1.5 × IQR. Black dots indicate values not classified as outliers; grey dots indicate the near outliers (values exceeding the third quartile ± (1.5 × IQR)) and white dots the far outliers (values exceeding the third quartile ± (3.0 × IQR). In order to expand the boxes and whiskers, the scale has been limited to a LDH/TNCC ratio of 21, and far outliers that exceeded the limit of the scale have been reported on the top of each graph, with the corresponding point value.

Figure 3

Receiver operating characteristic curve built using the values of LDH/TNCC ratio (in red), LDH (in green) and TNCC (in light blue) calculated based on the results of step 2 ((A), probability of FIP based only on cytological and chemico-physical analysis of the effusion) or of step 3 ((B), probability of FIP based on cytological and chemico-physical analysis of the effusion and on additional signalment or laboratory data). The central grey line represents the no-discrimination line.