Transurethral resection of the prostate gland (TURP) is one of the most frequently performed surgical procedures in males >60 yr of age. Since hypertrophied prostate tissue is excised by means of an electrically energized wire loop inserted through a resectoscope, large amounts of electrolyte-free irrigation fluid are needed to distend the urinary bladder, to improve optical conditions, and to remove blood and dissected prostate tissue during surgery. Inevitably, huge amounts of irrigation fluid are absorbed via large opened prostatic venous sinuses. Hence, shifts of water and electrolytes in the body take place, resulting in over-hydration and acute electrolyte imbalance, in particular acute hyponatraemia .
Numerous reports describe particular symptoms, commonly summarized as the so-called TURP syndrome, as a frequent side-effect of the procedure [2,3]. It is characterized by central nervous system complications including visual disturbances, seizures and coma as well as by cardiovascular problems such as bradycardia, hyper- or hypotension and, ultimately, heart failure . However, similar observations with a significantly lower incidence were made after transcervical electrosurgical ablation of the uterine endometrium [4,5] or after colonoscopy . In contrast to the sudden hyponatraemia during transurethral prostate resection (because of rapid volume overload with sodium-free irrigation fluid), chronic hyponatraemia (a slowly developing process) constitutes a medical challenge in malnourished alcohol-depended patients, patients with psychogenic or oxytocin-induced water intoxication, patients with renal failure, and in those treated with the antidepressant agent citalopram [7,8]. Non-osmotic secretion of arginine-vasopressin is thought to be a major pathogenic factor in the latter electrolyte disturbance .
A variety of neurological symptoms, in particular those associated with central pontine myelinolysis (CPM) have been attributed to hyponatraemia and its subsequent correction by sodium substitution. There is some evidence that especially the rapid correction of hyponatraemia rather than hyponatraemia itself is the reason of the observed problems. Case reports [10,11] have described morphological alterations in patient brains that include CPM often seen as subtle demyelinative foci, which are sometimes not apparent in cranial computed tomography and magnetic resonance imaging .
Animal studies revealed significant brain lesions subsequent to rapid correction of hyponatraemia to normonatraemia [12,13], whereas correction to still mildly hyponatraemic concentrations rarely caused demyelinating lesions . In these studies, severe hyponatraemia was induced by arginine-vasopressin injection and subsequent food and water restriction. Shortcomings of this model are that it involves significant interference with the animals' physiology, and that many animals died under these poorly controllable conditions [12,13].
We therefore sought to investigate the effects of mild hyponatraemia induced by intravenous (i.v.) application of Purisole SM®, the most commonly used non-electrolyte irrigation fluid during TURP in Germany, in a rodent model. We decided to infuse irrigation fluid 20 mL kg−1 body weight i.v. in rats. In men this equals an uptake of 1500-2000 mL, a volume which is readily absorbed in clinical practice during a transurethral prostate resection lasting between 30 and 60 min. We also studied the effects of rapid correction of hyponatraemia by hypertonic saline infusion. Since most of the previously published results were obtained from brains studied at the light microscopic level, we investigated the electron microscopic appearance of myelin sheets and glial cell numbers in the striatum and pons of the rat.
Animals, induction of hyponatraemia, treatments and time schedule
The procedures concerning animals reported in this study complied with German legislation for the protection of animals and were approved by the county-government authorities (Bezirksregierung Rheinhessen-Pfalz, Az 177-07/961-30). Twenty-one adult male Sprague-Dawley rats were used. The animals were maintained under constant conditions (light : dark 12 : 12 h, room temperature 21 ± 1°C) with food and water ad libitum.
For experimentation, the animals were anaesthetized with tribromoethanol (0.3 g kg−1 body weight) intraperitoneally. They were randomly assigned to one of three groups of seven rats each. The right great saphenous vein was exposed and acute hyponatraemia was induced by infusion of 20 mL kg−1 Purisole SM® (27 g sorbitol and 5.4 g mannitol per litre of sterile water, pH 4.5-7.0, theoretical osmolality: 178 mosm L−1) over 30 min (Group 1). Hyponatraemia was subsequently corrected by infusion of 2 mL NaCl 5.85% (in sterile water) over 10-15 min. Group 2 received the same dosage regimen of Purisole SM® but hyponatraemia was 'shamcorrected' by infusion of a balanced salt crystalloid solution (Sterofundin®; Braun, Melsungen, Germany), 2 mL. In both groups, blood was withdrawn via an arterial blood sampler containing 2.8 units lithium-heparin (Ciba-Corning Diagnostics Corp., Medfield, MA, USA). Blood sodium concentrations were analysed using a sodium/potassium analyser (Na-K-Analysator® 614; Ciba-Corning Diagnostics Corporation, Medfield, ME, USA) immediately before and after infusion of the irrigation fluid, as well as 15 min following infusion of the correction fluid. Thereafter, arterial and venous catheters were removed, wounds were sutured, and after recovery from anaesthesia the rats were returned to their cages. Rats of Group 3 served as untreated controls.
Following a 10 day survival, animals were re-anaesthetized in the middle of the light period of their day-night cycle. After obtaining a blood sample for sodium analysis, the rats were perfused transcardially with 100 mL of phosphate-buffered 0.9% saline (to which 15 000 IU heparin L−1 was added) at room temperature followed by 300 mL of ice-cold fixative (1% paraformaldehyde, 1.25% glutaraldehyde in 0.1 M Sorensen's phosphate buffer (PB), pH 7.4) with a constant perfusion rate of 10 mL min−1. The right atrium was opened to enable venous outflow. The brains were removed and cut in the frontal plane into approximately 1 mm thick slices. One block was taken from the central region of the striatum (slice corresponding to interaural 10 mm, cf. Fig. 14 in the stereotaxic atlas of the rat brain ) and another from the ventral region of the pons including the transverse fibre regions (slice corresponding to interaural 2.2 mm, cf. Fig. 44 in the atlas).
The blocks were post-fixed for 7 days in the same solution and then processed for routine electron microscopy as described earlier [15,16]. In brief, the tissues were rinsed in PB containing 5% sucrose, then post-fixed in 2% osmium-tetroxide for 120 min, washed in PB, dehydrated in a graded series of ethanols and flat embedded in Epon. Semi-thin sections were cut from the pons and the striatum, stained with toluidine blue and analysed using a light microscope. Ultra-thin sections were cut from corresponding regions, mounted onto 200 mesh copper grids, stained with uranyl acetate and lead citrate, and scanned on a Zeiss EM 10® electron microscope. Two sections 20 μm apart from each region per animal were analysed for signs of myelinolysis such as invagination of glio-axonal membranes, fibrillary accumulations, vesicular dissolutions or intralamellar splits . In addition, the numbers of astrocytes and oligodendrocytes were determined from 10 randomized grid holes per section. Two sections per region and animal were analysed. According to Hirano , astrocytes were identified by their numerous long, narrow processes and by the presence of intermediate filaments, usually arranged in parallel bundles, in cell bodies and processes. Oligodendrocytes, often arranged in a chain-like manner, were smaller with fewer processes, a dense nucleus surrounded by a narrow rim of cytoplasm without glial filaments.
Determination of cell numbers
Since the grid holes measured 65 × 65 μm each, a total of 84 500 μm2 per animal and region (corresponding to one unit area) was scanned. The investigators performing the electron microscopy were blinded to group/treatment assignment of the material analysed.
Sodium concentrations and cell numbers are presented as mean ± SEM. Statistical analysis comprised analysis of variance (ANOVA), where appropriate, for repeated measures. P < 0.05 was considered as significant. The software package (StatView 4.0®; Abacus Concepts Inc., Berkeley, CA, USA) was used for statistical analysis.
Mean serum sodium concentration in the present study was 145 mmol L−1 before infusion of Purisole SM®. This is in accordance with those found in previous studies in rats (140 mmol L−1; 150 meq L−1). The normal range of serum sodium concentration in rats is between 140 and 150 mmol L−1 thus being higher compared to human beings. Subsequent to infusion of Purisole SM®, serum sodium concentration decreased to 137 mmol L−1 (Group 1) and 132 mmol L (Group 2). Infusion of NaCl 5.85% or Sterofundin® raised serum sodium concentration to 175 mmol L−1 (Group 1) and 139 mmol L−1 (Group 2), respectively, immediately subsequent to correction of hyponatraemia. The sodium blood concentrations are shown in Table 1.
In the pons and striatum, no differences in myelin appearance were observed between experimental and control animals. In cross sections, the myelin revealed the typical lamellated appearance and surrounded the axon (Fig. 1). Signs of pathological alterations as reviewed  were not found in either group.
The average glial cell numbers are given in Table 2. No significant differences between groups were found in the striatum. In the pons, oligodendrocytes (Fig. 1d) were significantly reduced in Group 1 (corrected hyponatraemia) compared to controls. In addition, mean numbers of Group 2 (uncorrected hyponatraemia) were reduced but this difference lacked statistical significance. Astrocyte (Fig. 1b) numbers were significantly increased in Group 2 (uncorrected hyponatraemia) compared to Group 1 (corrected hyponatraemia) and to Group 3 (controls). In some cases, astrocytes exhibited signs of hypertrophy such as cytoplasmic swelling and increased glial filaments. Occasionally, oligodendrocytes showed membranous inclusions, which are thought to represent pathological change . Otherwise, there was no indication that the ultrastructure of glial cells was significantly different between the groups.
In the present study, we induced acute moderate hyponatraemia in rats by i.v. infusion of Purisole SM® and rapidly (within 15 min) corrected this in one experimental group by hypertonic sodium infusion (NaCl 5.85%), while hyponatraemia in the second experimental group was 'sham-corrected' by infusion of a balanced salt crystalloid solution (Sterofundin®). The comparison of pontine and striatal ultrastructure from controls and experimental animals 10 days after treatment revealed that myelin lamellae in these structures remained unaltered, while in the pons (but not in striatum) distinct alterations of glial cell numbers were found in experimental animals compared to controls.
Oligodendrocytes were numerically reduced in the pons of treated animals in the present study, reaching statistical significance only in Group 1 (corrected hyponatraemia). The functions of this glial cell type include the support of neuronal metabolism and the formation of myelin sheets. It is known from previous studies that, in lesions of CPM, destruction of oligodendrocytes is present . Our observation of significantly reduced oligodendrocyte numbers in the corrected hyponatraemia group agrees with another rat study in which hyponatraemia, induced by subcutaneous vasopressin injections, produced only minor effects on brain morphology while the rapid correction to normonatraemic levels resulted in significant lesions . In addition, it is thought that the neurological symptoms observed in human patients may not be due to hyponatraemia but rather to its rapid correction [5,19,20].
However, in view of the reduced number of oligodendrocytes our finding of apparently unchanged myelin upon either corrected or uncorrected hyponatraemia was somewhat surprising. Data in the literature provide evidence that pontine myelinolysis may be caused by the rapid correction of hyponatraemia. Kleinschmidt-DeMasters and Norenberg  reported that subcutaneous injections of vasopressin in rats induced hyponatraemia and that its rapid correction can lead to demyelinative lesions which may be the cause of CPM in male. This view is supported by Weissman and Weissman  who described a case of TURP-related corrected hyponatraemia in which the patient developed severe acute neurological symptoms associated with pontine lesions later detected by magnetic resonance imaging. Since these studies did not involve electron microscopy, it is debatable how comparable they are to the present study. However, it cannot be excluded that the extent of hyponatraemia in the present study (132-137 mmol L−1) was not sufficiently severe to induce visible signs of myelin degeneration after rapid restoration of normonatraemia. Moreover, it has to be kept in mind that the normal range of serum sodium concentration in rats (140-150 mmol L−1) is higher than in male.
In order to keep the experimental setting comparable to clinical situations, the purpose of our study was to examine the effects of rapid correction of moderate hyponatraemia. To guarantee patient safety, the therapy of hyponatraemia during transurethral prostate resection should commence before catastrophic deterioration from severe hyponatraemia, which occurs with serum sodium concentration <120 mmol L−1. Hence, ultrastructural cerebral changes associated with the treatment of moderate hyponatraemia are of much more relevance in everyday clinical practice than changes during severe hyponatraemia.
A further interesting observation in our study was that uncorrected (Group 2) but not corrected hyponatraemia (Group 1) resulted in a significant numerical augmentation of astrocytes. These cells are constituents of the blood-brain barrier and support liquid transport in the brain. This increase of astrocyte numbers in the pons of uncorrected hyponatraemia animals and their retention in the corrected group, although difficult to interpret, reveals that the effects of hyponatraemia, or its correction, or both, may be multifarious in animal studies as well as in human patients in whom astroglial alterations may represent a pathogenic factor in CPM .
In conclusion, in this animal model of moderate hyponatraemia we found rapid correction of serum sodium to result in a reduction in oligodendrocytes and an increase in astrocytes in the pons. No changes were seen in the striatum nor any changes in myelin in either site.
We thank Ciba-Corning, Germany, for providing the sodium analyser and the lithium-heparinized arterial blood samplers.
1. Jensen V. The TURP syndrome. Can J Anesth
2. Sunderrajan S, Bauer JH, Vopat RL, Wanner-Barjenbruch P, Hayes A. Posttransurethral prostatic resection hyponatremic syndrome: case report and review of the literature. Amer J Kidney Dis
3. Radziwill AJ, Vuadens P, Borruat FX, Bogousslavsky J. Visual disturbances and transurethral resection of the prostate: the TURP syndrome. Eur Neurol
4. Arieff AI, Ayus JC. Endometrial ablation complicated by fatal hyponatremic encephalopathy. JAMA
5. Rosenberg MK. Hyponatremic encephalopathy after rollerball endometrial ablation. Anesth Analg
6. Cohen CD, Keuneke C, Schlemann U, et al.
Hyponatraemia as a complication of colonoscopy. Lancet
7. Odeh M, Beny A, Oliven A. Severe symptomatic hyponatremia during citalopram therapy. Amer J Med Sci
8. Wang JY, Lin SH, Lin YF, Chen HY, Yu FC, Fu MT. An unforgotten cause of acute hyponatremia: Water intoxication due to oxytocin administration in a pregnant woman. Nephron
9. Anderson RJ, Chung HM, Kluge R, Schrier RW. Hyponatremia: a prospective analysis of its epidemiology and the pathogenetic role of vasopressin. Ann Int Med
10. Weissman JD, Weissman BM. Pontine myelinolysis and delayed encephalopathy following the rapid correction of acute hyponatremia. Arch Neurol
11. Clifford DB, Gado MH, Levy BK. Osmotic demyelination syndrome. Lack of pathologic and radiologic imaging correlation. Arch Neurol
12. Kleinschmidt-DeMasters BK, Norenberg MD. Rapid correction of hyponatremia causes demyelination: relation to central pontine myelinolysis. Science
13. Ayus JC, Krothapalli RK, Armstrong DL. Rapid correction of severe hyponatremia in the rat: histopathological changes in the brain. Amer J Physiol
14. Paxinos G, Watson C. The Rat Brain in Stereotaxic Coordinates,
2nd edn. San Diego: Academic Press, 1986.
15. Reuss S. Pineal 'synaptic' ribbons in sympathectomized rats
. Acta Anat
16. Reuss S, Mattern E, Spessert R, Riemann R, Weber A, Vollrath L. Lack of effect of oxytocin on the numbers of synaptic ribbons, cyclic guanosine monophosphate and serotonin N-acetyltransferase activity in organ-cultured pineals of three strains of rats
. Cell Tissue Res
17. Hirano A. A Guide to Neuropathology.
Tokyo, Japan: Igaku-Shoin, 1981.
18. Gocht A, Löhler J. Changes in glial cell markers in recent and old demyelinated lesions in central pontine myelinolysis. Acta Neuropathol
19. Gozal Y, Schwarzenberger J. Rapid correction of severe hyponatremia to normal values is dangerous. Anesth Analg
20. Sterns RH. Severe hyponatremia: the case for conservative management. Crit Care Med