A PRELIMINARY-REPORT ON URINARY BT-PABA PAS EXCRETION INDEX, SERUM PANCREATIC ISOAMYLASE AND FECAL CHYMOTRYPSIN TESTS OF PANCREATIC DYSFUNCTION IN SOWETAN AFRICANS/
C. Chaloner et al., A PRELIMINARY-REPORT ON URINARY BT-PABA PAS EXCRETION INDEX, SERUM PANCREATIC ISOAMYLASE AND FECAL CHYMOTRYPSIN TESTS OF PANCREATIC DYSFUNCTION IN SOWETAN AFRICANS/, Clinica chimica acta, 233(1-2), 1995, pp. 89-99
The steady increase in chronic pancreatitis among black Africans at So
weto, RSA, in the past 40 years necessitates an objective and non-inva
sive test to detect the disease at an early stage. Given the biphasic
nature of the disease - secretory hyperfunction with periodic active i
nflammatory episodes followed by steady exocrine impairment - we asses
sed three potential aids. Urinary BT-PABA/PAS excretion index (PEI), s
erum pancreatic isoamylase (PIA) and faecal chymotrypsin activity (FCA
) were measured in the following groups: 16 outwardly healthy hospital
workers, 16 consecutive patients with calcifying chronic pancreatitis
and 19 with abdominal pain ascribed to other conditions (disease cont
rols). (1) Healthy controls had lower PEI than those at Manchester, UK
, or Madras, India, from subclinical acinar loss as shown by lower PAP
A recovery whereas intestinal absorptive capacity was maintained, as s
hown by recovery of PAS, (2) Using the popular cut-off for PEI (0.75)
only 9 of 14 patients with chronic pancreatitis were identified (sensi
tivity 64%, 2 tests unsatisfactory), while a value of less than 0.54,
the mean -2 S.D, in local controls, yielded sensitivity of 50%. (3) If
PEI of less than 0.75 or PIA outside the reference range was taken to
indicate the disease, 5 of 9 disease controls would have been classed
as chronic pancreatitis (among those with both tests satisfactory): r
etrospective ultrasound scans did not identify these. (4) Although FCA
was less than the preselected cut-off, 5 units/g, in every patient wi
th chronic pancreatitis (100% sensitivity) its poor predictive value w
as indicated by low specificity: subnormal levels in 4 of 14 and 6 of
16 healthy controls or disease controls, respectively, most of whom ha
d near-normal values of PEI, PIA or both. (5) Collectively, these resu
lts suggest a high frequency of subclinical chronic pancreatitis at So
weto, but also that the combination of tests required to identify it m
ay prove impractical - whether in field surveys or hospital practice.