Complications of Fluid Overload from Resectoscopic Surgery

Abstract

Excess absorption of liquid distending media is one of the most frequent complications of operative hysteroscopy. Although most women recover uneventfully, we are seeing cases of permanent morbidity or death resulting from this complications.

(J Am Assoc Gynecol Laparosc 5(1):63-67, 1998)

Distending media for hysteroscopy include carbon dioxide, dextran 70(Hyskon) and various low viscosity crystalloid, such as normal saline, Ringer’s lactate solution, glycine, sorbitol, and mannitol solutions. With the development of the dual-channel solutions, flow hysteroscope sheath for the resectoscope, crystalloids have largely replaced other types of media.

Solutions used for intrauterine electrosurgical procedures must be nonconductive so the electrical current is not dissipated, and therefore those containing electrolytes cannot be used with the resectoscope. (At time of this writing, resectoscope designs that allow the use of electrolyte containing solutions are under development). Hyskon is sticky and crystallizes on instruments; it is therefore rarely used in conjunction with the resectoscope. When crystalloids are used for diagnostic hysteroscopy or for operative hysteroscopic procedures in which mechanical instruments are employed, normal saline and lactated Ringer’s solution are the media of choice.

Urologists originally used sterile water during transurethral surgery with the resectoscope. Because absorption of water causes hemolysis, solutions were developed that contain osmotically active particles to increase the osmolality of the fluid to a point at which hemolysis will not result should intravascular absorption occur.

Although commonly used liquid distending media do not cause hemolysis, their excess absorption can result in hyponatremia and hypoosmolality. This in turn can cause what is described as the TURP (transurethral resection of prostate) syndrome. It is characterized by hyponatremia, hypoosmolality, nausea, vomiting, and neurologic symptoms including muscular twitching, grand mal seizures, and coma.¹ When recognized promptly and treated properly, most patients with TURP syndrome recover without sequelae. On the other hand, convulsions, respiratory arrest, and permanent brain damage were associated with an average delay of 16 hours before therapy was given. Delayed recognition appears to be an important factor in morbid.

Gynecologists have assumed that pathophysiologic events observed by urologists in transurethral prostate surgery also apply to hysteroscopic surgery. However, the response of older men to hyponatremia and hypoosmolality differs from that of younger women. Whereas hyponatremia occurs with equal frequency in men and women.²The immediate cause of death and morbidity may be more a function of decreased serum osmolality than of hyponatremia.³

Computerized axial tomography showed cerebral edema occurring as a result of fluid absorption during hysteroscopic resection of the endometrium.⁴ In this study 1.5% glycine was used for uterine distention at a pressure of 100 mm Hg. Postoperative decrease in serum sodium correlated with glycine deficit. Ten patients had glycine deficits of less that 500 ml and a mean serum sodium decrease of 2.5 mmol/L during hysteroscopic surgery. Cerebral edema was diagnosed in 1 (10%) of these women. Two patients had glycine deficits of 500 to 1000 ml. Both had nausea, and one experienced cerebral edema. Eight patients had a deficit of 1000 ml or more and all had nausea and cerebral edema.

Irrigant absorption was measured ruing hysteroscopic surgery, and serum sodium fell to 120 mmol/L when 1000 ml of irrigant was absorbed.⁵ In contrast to the previous study, absorption of 1000 ml was associated with a decrease in serum sodium levels of 4 to 8 mmol/L.

Absorption of hypotonic fluid causes an osmotic imbalance between extracellular fluid and brain cells.⁷ Water moves into brain cells, causing cerebral edema, which can lead to pressure necrosis and progress to brain stem herniation and death. The brain’s compensatory mechanism includes pumping osmotically active cations out of the cells, thus reducing swelling. Laboratory studies suggest that this pump can be adversely affected by sex hormones, thus accounting for the vastly increased morbidity of hyponatremia in premenopausal women.⁸ This problem may be compounded or a similar one cause by elevated postoperative levels of vasopressin. ^2,9

^{Prevention of Fluid Overload}

The prevention of complications from distending media revolves around three principles:

• Avoiding excess fluid absorption
• Prompt recognition and proper treatment of overload should it occur.
• Selecting the distending medium least likely, to cause serious complications should it be absorbed in excess.

Fluid is delivered to the resectoscope by means of gravity or by pump. The lowest intrauterine pressure necessary to provide a clear field of view should be used to decrease the rate of absorption of the medium. Intrauterine pressure is controlled by a number of variables, such as inflow pressure, resistance of the tubing, inflow port/stopcock of the resectoscope, amount of suction (if any) applied to the outflow port, and existence of leaks around instruments.

Fluid absorption increases significantly when intrauterine pressure exceeds mean arterial pressure (MAP).¹⁰ The MAP appears to be an important concern and requires further study.

Since the specific gravity of mercury is approximately 13 times that of the distending medium, the pressure in inches of water can be approximated by the formula:

Pressure (inches of H₂O) =

Pressure (mm Hg) x 12/25 mm/inch.

Using this formula, 70 to 100 mg Hg pressure converted to inches of water would be 36 to 52 inches. If intrauterine pressure exceeds this value, the risk of rapid absorption of fluid increases.

When gravity feed is used, bags of liquid distention media should be at the lowest height that adequately distends the uterus, and at a level lower than MAP. Caution should be exercised when inflow tubing incorporates a drip chamber. When an air-fluid level is present in the chamber, its height determines infusion pressure. If the chamber fills, however, pressure will be determined by the height of the bag or bottle of fluid. This can instantly result in a calculated increase of 10 to 20 mm Hg. The bag would have to be lowered accordingly.

Pumps are being developed to monitor continuously and limit approximate intrauterine pressure, and they may decrease fluid absorption. Mechanical pumps, however, have not been entirely safe.¹¹These systems use measurements at the input or output port of the hysteroscope. True intrauterine pressure can be significantly higher than pressure measured at resectoscope ports, and is affected by the design of the resectoscope, rate of fluid flow though the resectoscope, and whether or not suction is applied to the outflow part. Additional research is required to clarify these measurements. Resectoscopes that incorporate a direct measurement of intrauterine pressure are being developed, and can be expected to provide more precise control.

Accurate measurement of intake and output remains the mainstay of preventing excess absorption. Such measurement, however, is not without potential problems. Commercially available containers of fluids may contain 5% to 10% more fluid than is specified.¹² An error of 5% can lead to an additional 1 L absorbed if a total of 20 L is administered. Another dilemma is that not all "returned" fluid may be recovered, some is commonly lost in drapes on the operating room floor. Several devices provide continuous monitoring of fluid deficit by simultaneously weighting inflow containers and outflow canisters. They provide an instantaneous readout of fluid status and decrease chances of error.

Another method being investigated to detect fluid absorption involves adding ethanol to the distending medium.^5,6,12 Measurements of blood or respiratory alcohol levels are used to calculate fluid absorption. Continuous monitoring of alcohol levels would provide an early indication of excessive fluid absorption. When alcohol levels are used to monitor fluid absorption in men undergoing prostate surgery, it was recommended that the surgeon be notified when 1000 ml of irrigant was absorbed and that the procedure be terminated when 2000 ml was absorbed.¹³ Although older men undergoing prostate surgery may tolerate this amount of absorption, the safety of such large amounts in women has not been demonstrated.

^{Choice of Distending Fluid}

Saline and lactated Ringer’s solution are crystalloids of choice for diagnostic hysteroscopy and hysteroscopic procedures using mechanical instruments. Fluids used with conventional resectoscopes must be nonconductive. Osmotically active particles must be added to the solution to prevent complications from absorption of hypotonic fluid. The osmolality of commonly used distention media is shown in Table 1.

Glycine 2.2% and 5% sorbitol are also isotonic, but are less readily available than 5% mannitol. Glycine has the disadvantage of being metabolized into ammonia by the liver. Although absorption of isotonic fluids lowers serum sodium, it is less likely than absorption of hypotonic media to cause cerebral edema.

Sorbitol and glycine are more commonly used for distention than mannitol and thus the knowledge base is wider for them. Sorbitol has been associated with hemodilution and hyponatremia. Complications of glycine include production of ammonia as a by-product, with secondary vision disturbance and muscle aches.

^{Treatment of Hyponatremia}

The treatment of hyponatremia in large part depends on the nature of the fluid that was absorbed. Excess absorption of 5% mannitol or other isotonic solution combinations, such as 5% sorbitol, and 4% sorbitol with 2% mannitol, may be less dangerous than absorption of hypotonic solutions. ^{10, 11, 16} All patients with severe or symptomatic hyponatremia should be provided supplemental oxygen until their condition has stabilized, as hypoxia may be a factor in causing neurologic sequelae. Hypocalcemia can accompany hyponatremia, ¹⁷and calcium should be administered if indicated.

If fluid overload was caused by 5% mannitol, no specific treatment may be necessary if mannitol-induced diuresis occurs. Loop diuretics may be contraindicated, as hypotension was observed in men when 5% mannitol solution irrigating fluid was combined with a postoperative intravenous diuretic. Serum potassium and calcium should be monitored, and replacement provided as necessary. Although the patient will be hyponatremia, she is unlikely to have reduced serum osmolality. She is therefore at low risk for cerebral edema from hypoosmolality. It is usually not necessary to administer hypertonic saline in the absence of abnormal neuromuscular or cardiac function. As hyponatremia usually coexists with hypoosmolality, the level of sodium that requires urgent treatment when osmolality is normal has not been determined. It may not be necessary to administer hypertonic saline in the absence of abnormal neuromuscular or cardiac function (A.I. Arieff, personal communication).

Treatment of overload from hypotonic agents is more complex. Although most women recover, seizures, permanent brain damage, and death were reported even with serum sodium levels of 116 ± 2mmol/L.¹⁷ Any patient with a serum sodium less then 120 mmol/L resulting from absorption of glycine, sorbitol, or other hypotonic fluid should be considered at risk for these complications, monitored and treated in a critical care setting. If the patient is symptomatic or serum sodium is less than 120 mmol/L, she should be treated with 3% sodium chloride at a rate to increase serum sodium by about 1 mmol/L hour. If 3% saline is not readily available it is possible to substitute two ampules (50 mEq each) of sodium bicarbonate for initial sodium replacement. Serum potassium and calcium levels and osmolality should be monitored frequently. No attempt should be made to raise the serum sodium level above 130 mmol/L.

^{Guidelines for Resectoscopic Surgery}

Surgery with the resectoscope should be done only by surgeons with specific training in the procedures. The anesthesiologist and all operating room staff should be aware of the potential for fluid overload and monitor patients accordingly.

Isotonic solutions may increase the margin of safety and should be considered for uterine distention. Mannitol 5% may be the medium of choice for use with the resectoscope because of its diuretic properties. Other isotonic fluids, such as 5% sorbitol, 4% sorbitol with 2% mannitol and 2.2% glycine, and also available. When resectoscopes that are designed to function in the presence of electrolyte containing solution are available, normal saline or Ringer’s lactate solution will eliminate risk of hypoosmolality and hyponatremia but not of pulmonary edema.

A Foley catheter should be placed in the bladder to monitor urine output in any case considered at risk for excess fluid absorption, or if excess absorption is suspected. One person in the operating room should be assigned the responsibility of measuring intake and output and informing the surgeon and anesthesiologist of the measurements. Intake and output should be reported after each container of fluid is used, and at a minimum of every 1- minutes. Consideration should be given to use of an instrument that continuously monitors fluid balance.

If 500 to 1000 ml (less if the patient is medically compromised) is believed to be absorbed, the following should be undertaken: (1) the procedure should be suspended until fluid status is ascertained, (2) a Foley catheter should be placed, if not already in place; (3) consideration should be given to sending serum for immediate measurement of electrolytes; and (4) consideration should be given to rapid conclusion of the operative procedure as appropriate, as once fluid absorption is started in can progress rapidly.

If greater than 1500 ml is absorbed or serum sodium is less than 125 mmol/L, the procedure should be terminated as rapidly as reasonable. Patients with a serum sodium level below 120 mmol/L should be considered for treatment in a critical care setting, especially if a hypotonic distending medium was used.

TABLE 1. Osmolality of Commonly Used Nonconductive Distending Media

Medium Osmolality (mOsmol/L)

290

200

178

280

 Reprinted with the permission of The Journal of the America Association of Gynecologic Laparoscopists

References:

References:

1. Arieff AI: Hyponatremia associated with permanent brain damage. Adv. Intern Med 32:325-344. 1987.

2. Ayus JC, Wheeler JM, Arieff Al: Postoperative hyponatremic encephalopathy in menstruant women. Am Intern Med 117:891-897, 1992

3. Baggish MS, Brill AI, Rosensweig B. et al: Fatal acute glycine and sorbitol toxicity during operative hysteroscopy. J Gynecol Surg 9,137-143, 1993

4. Istre O, Bjoennes J, Naess R, et al: Postoperative cerebral oedema after transcervical endometrial resection and uterine irrigation with glycine. Lancet 344:1187-1189, 1994

5. Duffy S, Cruise M, Reilly C, et al: Ethanol labeling: Detection of early fluid absorption in endometrial resection. Obstet Gynecol 79:300-304, 1992

6. Hahn RG, Ekengren JC: Patterns of irrigating fluid absorption during transurethral resection of the prostate as indicated by ethanol. J Urol 149:502-506, 1993

7. Arieff AI: Treatment of symptomatic hyponatremia: Neither haste nor wast (editorial). Crit Care Med 19:748-751, 1991

8. Fraser CI, Arieff AI: Fatal central diabetes mellitus and insipidus resulting from untreated hyponatremia: A new syndrome. Ann Intern Med 112:113-119, 1990

9. Perry CP: Syndrome of inappropriate antidiuretic hormone after laparoscopic-assisted vaginal hysterectomy. J Am Assoc Gynecol Laparosc 1(3):173-175, 1994

10. Garry R, Hasham F, Kokri MS, et al: The effect of pressure on fluid absorption during endometrial ablation. J Gynecol Surg 8:1-10, 1992

11. Shirk GJ, Gimpelson RJ: Control of intrauterine fluid pressure during operative hysteroscopy. J Am Assoc Gynecol Laparosc 1:229-233, 1994

12. Vulgaropulos SP, Haley LC, Hulka JF: Intrauterine pressure and fluid absorption during continuous flow hysteroscopy. Am J Obstet Gynecol 167(2):386-390. 1992

13. Hahn, RG: Monitoring of TURP with ethanol [letter]. Lancet 338:1602, 1991

14. Jackson S, Lampe G: Operative hysteroscopy intravascular absorption (OHIA) syndrome. West J Med, In Press

15. Matouschek E, Loening S: Urologic Endoscopic Surgery. Philadelphia, BC Decker, 1989, pp 35-38

16. Krohn JS: Dilutional hypocalcemia in association with dilutional hyponatremia. Anesthesiology 79:1136-1138, 1993

17. Arieff AI: Management of hyponatremia. BMJ 307:305-308, 1993