Terminally ill patients suffering from intractable cancer pain are tre
ated in our hospital on an outpatient basis with a percutaneous intrat
hecal (i.t.) catheter and a portable pump delivering morphine continuo
usly. In a patient showing an increased demand of morphine the dose wa
s raised from 1.5 to 2 mg/h, but pain intensity did not decrease. Subs
equently a 1.5 ml dose of 5% lidocaine was administered; however, no m
otor or sensory block was observed. After controlling the catheter pos
ition and passage through the catheter, a sample of cerebrospinal flui
d (CSF) was taken and the pH was measured. It was found to be outside
the physiological range of 7.19 (normal range: 7.27-7.37), possibly ex
plaining the decreased activity of the local anesthetic. The purpose o
f this study was to determine the influence of morphine, with or witho
ut sodium metabisulfite, on pH in vitro, using artificial CSF (ACSF) a
nd on pH in vivo during i.t. administration of morphine. An in vitro m
odel was used to measure pH changes by adding a morphine solution (con
centrations of 0.5, 2, 5 and 10 mg/ml) with and without sodium metabis
ulfite to ACSF solutions (Elliott B). Fourteen patients were consecuti
vely selected for continuous administration of morphine. An i.t. cathe
ter was inserted, tunnelled and connected with an external pump (Provi
der 5500, Abbott, Chicago, IL). CSF was aspirated and pH was measured
with a blood-gas system (Ciba-Corning 288, Medfield, USA). In vitro, m
orphine solutions with or without sodium metabisulfite added to an Ell
iott B solution (pH = 7.47, 37 degrees C) caused a concentration-relat
ed decrease in pH. In vivo, in 9 patients, the daily morphine dose was
less than 24 mg and the range of the CSF pH was between 7.25 and 7.44
. In 5 patients, the daily morphine dose was higher than 24 mg and the
CSF pH ranged from 6.92 to 7.24. In 4 patients myoclonic seizures and
an increase in pain were observed. Our study suggests that long-term
i.t. administration of high-dose morphine solutions with or without so
dium metabisulfite may decrease the pH of CSF outside the normal range
of 7.27-7.37. This may explain the diminished effectiveness of i.t. a
dministered local anesthetics and morphine due to the increase of the
ionized part of the available drug. Buffer capacity of CSF is insuffic
ient to maintain normal pH values in patients receiving high doses of
morphine solutions. A more neutral morphine formulation may be necessa
ry for i.t. use.