Postanesthetic Polyuria Attributable to Central
Fumio Otsuka, MD*, Satoshi Mizobuchi, MDt, Kiyoshi Morita, MDt, Hideo Matsumoto, MD*,
Toshio Ogura, MDS, Masahisa Hirakawa, MDt, and Hirofumi Makino, MD*
Departments of *Medicine III, tAnesthesiology, and $Surgery II, Okayama University Medical School; and §Health and
Medical Center, Okayama University, Okayama, Japan
A fter general anesthesia, the clinical manifesta- anesthesia was as follows: lactated Ringer’s solution
tions of polyuria have occasionally observed. 1100 mL, acetated Ringer’s solution 1000 mL, 5% plasma
However, because of the usual abrupt onset and protein 250 mL, blood loss 230 g, and urine output 1100 mL.
The resected tumor was localized to the left upper seg-
rapid spontaneous remission, detailed investigations
ment, in the S (1+2) area, and measured 1.7 x 1.5 x 1.5 cm.
of polyuria have been limited. We present a case of It did not involve the pleura or the regional lymph nodes.
severe polyuria after lung lobectomy and examine the The pathologic findings revealed a highly differentiated
etiology of the polyuria both endocrinologically and lung adenocarcinoma. On Day 1 postoperatively, the pa-
radiologically. We discuss the attributes of this poly- tient’s daily urine volume was 7500 mL. She complained of
uric manifestation. polydipsia and severe general fatigue. Because the hourly
urine volume was more than 500 mL and central venous
pressure was decreased 30-40 mm H,O, rapid fluid replace-
ment was administered for the prevention of dehydration.
Laboratory values, which were mildly controlled by the
Case Report fluid replacement, were as follows: red blood cell count
A 60-yr-old woman weighing 41 kg was discovered to 3.46 X 106/pL (normal 3.7-4.9 X 106/pL), white blood cell
have a coin lesion in the left upper lobe of the lung on chest count 9.2 X 103/pL (normal 3.0-9.4 X 103/pL), hematocrit
roentgenogram (Fig. 1). After transbronchial lung biopsy, 31.4% (normal 35%~44%), hemoglobin 10.9 g/dL (normal
she was scheduled for left upper lobectomy of the lung. The 11.5-14.5 g/dL), serum sodium 143 mmol/L, serum potas-
preoperative examination of the patient revealed no abnor- sium 4.2 mmol/L, serum chloride 105 mmol/L, blood urea
malities except for the lung tumor. Of those examined pre- nitrogen 4.3 mg/dL (normal 8.1-22 mg/dL), serum creati-
operatively, electrolytes were as follows: serum sodium nine 0.43 mg/dL (normal 0.44-1.04 mg/dL), fasting blood
140 mmol/L (normal 136-144 mmol/L), potassium glucose 105 mg/dL (normal 65-105 mg/dL); urine N-acetyl-
4.4 mmol/L (normal 3.7-4.9 mmol/L), chloride 103 mmol/L P-glucosaminidase 1.8 II-I/L (normal 1.1-20.3 III/L), urine
(normal 102-110 mmol/L). The patient was premeditated &-microglobulin 0.07 mg/dL (normal ~0.2 mg/dL). Liver
with hydroxydine 50 mg subcutaneously 1 h before arriving function tests were within normal limits. The endocrinologic
in the operating room, after which an epidural catheter was data under 6-h water deprivation are shown in Table 1. The
introduced at T5-6 and 0.125% bupivacaine was adminis- serum osmolality increased to 300 mOsm/kg. The urine
tered intraoperatively at a rate of 4.0 to 5.0 mL/h via the osmolality decreased to 140 mOsm/kg, and the plasma ar-
catheter. The patient was tracheally intubated with 200 mg ginine vasopressin (AVP) concentration was 1.1 pg/mL. The
thiopental and 6 mg vecuronium, followed by OS%-2.0% serum cortisol concentration was increased at 26.0 pg/dL,
sevoflurane in 6 L/min 0, using a semiclosed circuit and the presumably because of the postoperative stressful state. The
epidural infusion. Although arterial blood pressure de- plasma renin activity and the aldosterone concentration de-
creased to 90/40 mm Hg 15 min after the first epidural creased because of the hyperosmolar state. Thyroid hor-
administration, it recovered immediately with a 4-mg mones, catecholamines, and other pituitary hormones were
ephedrine injection. Blood pressure (BP) and heart rate (HR) normal. The patient was treated with desmopressin
were stable (BP 100/60 to 120/75 mm Hg, HR 60-75 bpm), (DDAVP), and sequential hourly urine volume diminished
and the saturated oxygen (Spo2) was kept at 100%. The total significantly (before: 600 mL, 4 h after: 40 mL), and the
operative time was 4 h and 40 min. The patient inhaled hourly urine osomolality increased (before: 128 mOsm/kg,
sevoflurane for 4 h and 5 min (4.2 minimum alveolar anes- 4 h after: 724 mOsm/kg) (Fig. 2). A loading test of 5%
thetic concentration [MAC] hours). Fluid balance during sodium chloride was given, which resulted in a smaller
increase in AVP levels in response to the high serum osmo-
Accepted for publication July 16, 1997. lality (Fig. 3a). Cranial magnetic resonance imaging (MRI)
Address correspondence and reprint requests to Fumio Otsuka, exhibited a slightly decreased signal in the high-intensity
MD, Department of Medicine III, Okayama University Medical posterior pituitary region on the Tl-weighted images (Fig.
School, 2-5-l Shikata-cho, Okayama City, Okayama 700, Japan. 4A) and showed mild swelling of the pituitary gland and
01997 by the International Anesthesia Research Society
940 Anesth Analg 1997;85:940-3 0003.2999/97/$5.00
ANESTH ANALG CASE REPORTS 941
DDA\P 10 u g
1 2 3 4
Figure 2. Desmopressin (DDAVP) loading test. Intranasal desmo-
pressin effectively changed the urine volume and urine osmolality.
Hourly urine volume was decreased from 600 mL to 40 mL for 4 h.
Urine osmolality was elevated from 128 mOsm/kg to 724
mOsm/kg after 4 h.
Figure 1. Chest roentgenogram. In the left upper lobe, a small coin
lesion was found (arrow). Histologically, the resected tumor was
diagnosed as a well differentiated adenocarcinoma.
Table 1. Endocrinologic Data at the Onset of Polyuria
Serum osmolality 300 mOsm/kg (285-295)
Urine osmolality 140 mOsm/kg (50-1400)
Plasma arginine 1.1 pg/mL (0.8-6.3)
Human atria1 natriuretic 26.8 pg/mL (<43)
Plasma renin activity CO.1 ng/mL/h (0.3-2.9) 220 2io’40’G3o’&’ 0
Plasma aldosterone 27.4 pg/mL (56.9-150.3) SerumOsmolality (mOsm/kg)
concentration Figure 3. 5% sodium chloride loading test. Upon the manifestation
Serum cortisol 26.0 &dL (5-21) of polyuria, the arginine vasopressin (AVP) secretive response to 5%
Median ranges are shown in parentheses
sodium chloride was lowered (a); 1 mo (b), 2 mo (c), and 10 mo after
(d), the AVP response was gradually improved. The dotted area
shows the normal range for the ratio of AVP concentration to serum
pituitary stalk on the gadolinium enhanced image (Fig. 4B). osmolality (8).
The patient’s daily urine volume was maintained at approx-
imately 4000-5000 ml, with DDAVP at doses up to 30 PLg for
14 days postoperatively, and then gradual improvement of normal urine volume and no complaints 10 mo after the
urine volume and osmolality was spontaneously shown operation without DDAVP treatment.
(Fig. 5). The 5% sodium chloride loading test was repeated 1,
2, and 10 mo (Fig. 3, b-d) after the operation, and the AVP
secretory response to high serum osmolality gradually in- Discussion
creased, particularly 10 mo after the operation. Cranial MRI
2 mo after the operation exhibited diminished volume of the After general anesthesia, the manifestation of polyuria
pituitary gland with a slender pituitary stalk in comparison is sometimes observed. The onset is abrupt and within
with previous MRI findings (Fig. 4C). The patient had a several days after the operation. In cases of brain
942 CASE REPORTS ANESTH ANALG
In this patient, sevoflurane was used for general
anesthesia. Sevoflurane is one of the inhaled fluori-
nated anesthetics that rarely causes polyuria (2). How-
ever, the pathogenesis of polyuria has not been clari-
fied, and sevoflurane has not been shown to produce
renal impairment, despite its administration to ap-
proximately 2 million patients in Japan. Biotransfor-
mation of the fluorinated volatile anesthetics results in
the production of fluoride metabolites, which may be
associated with hepatic or renal toxicity (3,4). Sevoflu-
rane undergoes hepatic biotransformation in humans,
Figure 4. The sequential cranial magnetic resonance imaging find- producing peak plasma inorganic fluoride concentra-
ings. At the onset of polyuria, the pituitary gland exhibited a de-
creased high-intensity signal (arrow) on the Tl-weighted image (A) tions similar to those associated with renal impair-
and slight swelling of the pituitary gland (small arrow) and stalk ment after methoxyflurane anesthesia (5,6). In the case
(large arrow) on the gadolinium-enhanced image (B). Two months of methoxyflurane, biotransformation by the renal cy-
later, the pituitary gland (small arrow) was diminished with a
slender stalk (large arrow) on the gadolinium-enhanced image (C). tochrome I’450 is believed to be closely associated
with renal damage (4-6). In the case of enflurane,
which is also a fluorinated anesthetic, the mechanism
of polyuric renal impairment is controversial (7). Pre-
vious investigations suggest that renal tubular dam-
age is due to a direct toxic effect of these drugs rather
than to the inorganic fluoride ion produced when they
The patient’s renal function tests (serum creatinine,
blood urine nitrogen, and urinary N-acetyl-P-
glucosaminidase and P,-microglobulin both) at the
onset of polyuria and at two months showed no ab-
normalities. The polyuria was diagnosed as partial
Figure 5. The clinical course. One day after the surgery, the patient central DI because of the blunted AVP-secretory re-
abruptly manifested polyuria, 7500 ml/day. She received fluid
replacement to prevent dehydration. Desmopressin (DDAVP) was sponse to hyperosmolality (8) and to the adequate
administered intranasally for the control of urine volume. Approx- renal concentrating response to exogenous vasopres-
imately 2 wk after the onset of polyuria, her urine volume was sin. In a study by Murakawa et al. (9) of 24 patients
decreased without desmopressin, and 1 mo after that, it was com-
pletely normalized. Daily urine volume (UV) is shown as closed who underwent sevoflurane anesthesia, no significant
column, and daily fluid replacement volume (DIV) is shown as open pre- or postoperative changes in plasma AVP or hu-
column. man atria1 natriuretic peptide levels were reported.
Although the present case also showed a normal-
surgery involving the pituitary gland or hypothala- ranged AVP concentration at the onset of polyuria, a
mus, the postoperative central diabetes insipidus (DI) precise examination of the correlation with osmolality
is easily recognizable; however, anesthetic and oper- using the sodium loading test proved an impaired
atic procedures, except for cranial surgery, rarely af- AVP-secretory capacity. On the other hand, the re-
fect water balance. In this procedure, among the fac- markable improvement of AVP response on sodium
tors affecting water balance were blood and fluid loss, loading tests in 10 months obscures a presumable
excessive fluid transfusion, anesthetics, unexpected preoperative impairment of AVP secretion. A pitu-
brain damage, and previously undetected DI. In the itary MRI finding at the onset of polyuria showed a
present case, as in the preceding description, blood slight decrease of high intensity in the posterior lobe
loss, excessive transfusion, and brain metastasis on Tl-weighted images, which indicates the dimin-
within one year after the operation were negative, but ished AVP-neurosecretory granules (10). It also
we could not exclude the manifestation of an unex- showed a mild swelling of the pituitary gland and
pected DI. Although operative stress and damage, stalk on the gadolinium-enhanced images. Moreover,
particularly in thoracic surgery, affects the barorecep- two months of postoperative images exhibited dimin-
tor and AVP inhibitory vagal nerve in thorax (l), the ished volume of the pituitary gland and stalk. The
symptomatic feature is usually a syndrome of inap- sequential change in the pituitary MRI findings strongly
propriate secretion of antidiuretic hormone rather suggests that the central DI in this case was closely
than polyuria. This is also in agreement with cases associated with “infundibulo-neurohypophysitis,”
with hormone producing neoplasm, e.g., antidiuretic which is a known cause of idiopathic (11) or
hormone secreting oat cell carcinoma in lung. autoimmune-associated central DI (12).
ANESTH ANALG CASE REPORTS 943
Although postoperative polyuria tends to be under- 3. Frimk EJ, Maran TP, Isner RJ, et al. Renal concentrating function
with prolonged sevoflurane or enflurane anesthesia in volun-
estimated because of its reversibility and the lack of
teers. Anesthesiology 1994;80:1019-25.
apparent renal damage, further examination of both 4. Kharasch ED, Thummel KE. Identification of cytochrome P450
pituitary function and AVP responsiveness is neces- 2El as the predominant enzyme catalyzing human liver micro-
sary to clarify the pathogenesis. Fortunately, we could somal defluorination of sevoflurane, isofurane, and methoxy-
examine a patient with postanesthetic polyuria, which flurane. Anesthesiology 1994;79:795-807.
5. Cousins MJ, Mazze R, Kosek JC, et al. The etiology of methoxy-
was diagnosed as attributable to partial central DI;
flurane nephrotoxicity. J Pharmacol Exp Ther 1974;190:530-41.
however, we cannot completely determine what 6. Cousins MJ, Mazze RI. Methoxyflurane nephrotoxicity: a study
caused the central DI. To exclude the unexpected fac- of dose response in man. JAMA 1973;225:1611-6.
tors, inducing water imbalance, other than sevoflu- 7. Mazze RI, Calverly RK, Smith NT. Inorganic fluoride
rane anesthesia, and to determine this pathogenesis, nephrotoxicity: prolonged enflurane and halothane anesthesia
in volunteers. Anesthesiology 1977;46:265-71.
further studies including the AVP-secretory response 8. Robertson GL. Vasopressin in osmotic regulation in man. Annu
toward high osmolality or water deprivation on cases Rev Med 1974;25:315-22.
of postoperative and postanesthetic polyuria are 9. Murakawa T, Kudo T, Kudo M, et al. Effects of surgical inter-
necessary. vention on plasma level of antidiuretic hormone and o-human
atria1 natriuretic polypeptide under sevoflurane anesthesia [in
Japanese]. Jpn J Anesthesiol 1989;38:1195-200.
10. Fujisawa I, Asato R, Kawata M, et al. Hyperintense signal of the
References posterior pituitary on Tl-weighted MR images: an experimental
1. Otsuka F, Hayashi Y, Ogura T, et al. Syndrome of inappropriate study. J Comput Assist Tomogr 1989;13:371-7.
secretion of antidiuretic hormone following intra-thoracic cis- 11. Imura H, Nakao K, Shimatsu A, et al. Lymphocytic infundibulo-
platin. Internal Med 1996;35:290-4. neurohypophysitis as a cause of central diabetes insipidus.
2. Horinouchi S, Iwatsuki N, Ono K, et al. A case of polyuria N Engl J Med 1993;329:683-9.
probably induced by sevoflurane [abstract, in Japanese]. Jpn J 12. Otsuka F, Amano T, Ogura T, et al. Diabetes insipidus with
Anesthesiol 1987;36:1471. Behcet’s disease. Lancet 1995;346:1494-5.