The safest and most conservative treatment of cirrhotic ascites and edema is the institution of bed rest and a low salt diet. This regimen will produce significant clinical improvement in 20-30% of cirrhotic patients . Bed rest reduces lower extremity venous pooling, shifts some blood from splanchnic to central vessels, and expands the EABV. Although bed rest clearly promotes diuresis and natriuresis, it is only a temporizing maneuver.
Restricting salt intake below the rate of renal excretion will obviously result in negative salt and water balance. However, renal salt excretion may be extraordinarily low in cirrhotic patients, so the required degree of dietary sodium restriction may be difficult or impossible to achieve. Although some investigators suggest that dietary sodium be reduced below 500 mg/day (about 1300 mg of NaCl or 22 mEq of Na+), this is generally only possible when patients are hospitalized. Moreover, diets with this degree of sodium restriction are generally unpalatable and interfere with efforts to improve nutrition. A more realistic level of outpatient salt restriction is about 2 g/day of sodium (about 5 g NaCl or 87 mEq Na+). The fact that some cirrhotic patients excrete only 10-20 mEq of Na+ per day means they will continue to gain weight on such a regimen. When bed rest and dietary salt restriction do not adequately control ascites and edema, diuretics are initiated. However, after initiation of diuretic therapy, a low salt diet remains an important component of the treatment regimen. This will enhance the efficacy of the diuretic and reduce the incidence and severity of diuretic-associated electrolyte complications (see Section on Diuretics, Chapter VA3).
Spironolactone and Other Potassium-Sparing Diuretics
Traditionally, the first diuretic used in patients with cirrhosis is the aldosterone antagonist, spironolactone, at an initial dose of 50-100 mg/day. This competi tive inhibitor of aldosterone is rapidly metabolized to a number of compounds which undergo slower metabolism and excretion. Several of these spironolactone metabolites also have aldosterone blocking and diuretic activity. In fact, the aldosterone metabolite potassium canreneate is available as a diuretic in Europe. As a result of these metabolites, the biologic diuretic activity of spironolactone is prolonged; hepatic dysfunction causes an accumulation of these metabolites and longer periods of activity . Spironolactone and its active metabolites bind to the mineralocorticoid receptors in the cytoplasm of the cortical and medullary collecting tubule cells (and other mineralocorticoid sensitive tissues) and block the mineralocorticoid effects of aldosterone. In general, 7 to 10 days of spironolactone therapy is required to achieve the maximal effect of a given dose in patients with hepatic cirrhosis. Consequently, it is advisable to wait a week or more before deciding whether a given dose is effective. The maximal therapeutic dose of spironolactone is about 400 mg/day .
Many, but not all, cirrhotic patients have a good diuretic response to spironolactone monotherapy . When spironolactone resistance is encountered there are several possible explanations. Spironolactone and its metabolites competitively block the action of aldosterone; therefore, they have little effect when endogenous aldosterone levels are low. Although most patients with decompensated cirrhosis have high aldosterone levels, this is not universally true . Spironolactone will generally be ineffective in this low aldosterone subgroup. Conversely, when aldosterone levels are extremely high, they may not be adequately blocked by a competitive inhibitor. Another potential cause of spironolactone resistance is low distal tubule delivery of sodium salts. Aldosterone can only increase reabsorption of sodium which is delivered to the aldoste-rone-sensitive distal sites. Avid sodium reabsorption by the proximal tubule, thick ascending limb of Henle (TALH), and early distal tubule will reduce markedly sodium delivery to the late distal and cortical collecting tubules. Under such circumstances even very high aldosterone levels do not produce much sodium reabsorption and blocking aldosterone activity with spironolactone in these patients can only produce a modest diuretic effect. Addition of other diuretics with more proximal effects will deliver sodium to more distal sites where spironolactone acts.
Many spironolactone side-effects result from its potent antiandrogenic activity. Indeed, this side-effect has been effectively employed to treat patients with hyperandrogenic hirsutism and/or acne. This effect can cause painful gynecomastia in many patients receiving this drug. The electrolyte derangements produced by spironolactone include hyperkalemia and hyperchloremic metabolic acidosis. They occur with increased frequency in patients with a reduced renal function.
Amiloride and triamterene are other potassium-sparing diuretics which in hibit sodium reabsorption in the cortical collecting tubule (CCT). These are relatively weak diuretics with a 12- to 24-hr duration of action. Unlike spironolactone, amiloride and triamterene are not competitive inhibitors of aldosterone; their diuretic activity is independent of the aldosterone level. The effects of amiloride have been better characterized. The drug acts on the luminal membrane of the CCT to block sodium reabsorption through the sodium channels in this tubule segment. Triamterene has similar transport inhibiting effects, but its mechanism of action may be somewhat different. Like spironolactone, these two drugs can generate a meaningful diuresis only when a significant quantity of sodium is delivered to the CCT.
Both drugs reduce the negative electrical potential difference which sodium reabsorption usually produces in the CCT. This negative charge is critical for distal tubule potassium and proton secretion. Consequently, as with spironolactone, serum electrolytes must be monitored to detect hyperkalemia or hy-perchloremic metabolic acidosis.
Triamterene also has several unique side-effects. The drug can precipitate within the distal renal tubules and has been associated with acute renal insufficiency, especially when it is used together with nonsteroidal anti-inflammatory drugs. Clinically significant kidney stones, composed principally of triamterene, have also been reported.
When potassium-sparing distal tubule diuretics do not generate an adequate diuresis, loop diuretics may be substituted or added. Loop diuretics are often required when the GFR is reduced . In general, furosemide pharmacokinetics are similar in patients with cirrhosis and those seen with normal liver function with comparable levels of renal function . However, extrarenal clearance of bumetanide and torsemide are reduced in cirrhotic patients. Therefore, the blood levels and renal delivery of these two loop diuretics are increased in cirrhotic patients [12, 17]. When cirrhotic patients also develop advanced renal insufficiency, high peak plasma levels of loop diuretics are necessary to produce adequate intratubule drug concentrations. Under such circumstances, severe hypoalbuminemia may reduce diuretic efficacy. Less diuretic bound to albumin will increase the drug's volume of distribution and thereby decrease peak diuretic plasma levels (see Section on Diuretics, Chapter VA3 . However, this effect is probably only clinically significant when severe (<2 g/100 ml) hypoalbuminemia and renal failure coexist.
Even when adequate concentrations of a loop diuretic are delivered to the active site of action in the TALH, some cirrhotic patients may remain diuretic resistant. This resistance represents a pharmacodynamic phenomenon of un certain mechanism but is probably due to increased sodium reabsorption in the proximal tubules and the distal tubule segments beyond the TALH. Thiazides can be used to overcome this form of resistance (see below).
A reasonable initial oral dose of furosemide in a cirrhotic patient with relatively normal renal function is 40-80 mg. Equivalent doses of bumetanide are 1-2 mg and of torsemide 25-50 mg. If renal function is reduced the initial dose may be doubled or tripled. The two more recently introduced loop diuretics are more completely and reliably absorbed from the GI tract than furosemide. Torsemide also has a longer half-life than furosemide (about 3 hr compared with 1 hr in normal subjects) and its duration of action is further prolonged in patients with cirrhosis.
Once an effective dose of loop diuretic is established, further increases of each dose have minimal additional effect (i.e., the dose-response curve becomes flat once an effective dose is reached). If each dose of loop diuretic is effective but the daily magnitude of diuresis remains insufficient, then periods of avid sodium retention between doses may explain such resistance. The impact of this phenomenon can be reduced by decreasing salt intake, switching to diuretics with a longer half-life (such as torsemide) or increasing the frequency of administration of the diuretic.
Thiazide diuretics rarely generate an adequate diuresis when used alone in cirrhotic patients. However, this class of diuretics has synergistic effects when combined with loop agents . Furosemide is a potent inhibitor of sodium absorption in the TALH. Thiazides have a weak proximal tubule carbonic an-hydrase inhibiting effect and more potent effects in the early distal cortical tubule. Combining a thiazide and loop diuretic is often synergistic because the loop diuretic inhibits sodium reabsorption in the TALH and markedly increases sodium delivery to the thiazide-sensitive distal cortical tubule. Combining a long-acting thiazide or thiazide-type drug, such as metolazone, with furosemide is often a successful strategy. If parenteral diuretics become necessary, intravenous chlorothiazide, at doses of 500-1000 mg/day, may be combined with an intravenous loop diuretic. These very potent diuretic combinations can rapidly produce hypovolemia and extreme electrolyte abnormalities, especially hypokalemia. Therefore these patients must be carefully monitored clinically and biochemically.
Patients who remain diuretic-resistant despite use of thiazide/loop diuretic combinations occasionally will respond to the addition of mannitol which ex pands the extracellular compartment, increases the GFR, and increases solute delivery out of the proximal tubule. An intravenous bolus of 12.5 to 25 g of mannitol may be administered q 6-8 or 250-500 ml of a 20% mannitol solution may be infused. Although the potent carbonic anhydrase inhibitor aceta-zolamide will also increase delivery of filtrate out of the proximal tubule, this drug should probably not be utilized in cirrhotic patients because it frequently produces hypokalemia and simultaneously alkalinizes the urine. The potential pernicious effects of these biochemical derangements are discussed under Diuretic Complications, below.
Another therapeutic option which may potentiate diuretic efficacy is expansion of the intravascular space with infusions of salt-poor albumin. Albumin infusions are very expensive and beneficial effects are relatively transient. Nonetheless, markedly hypoalbuminemic patients who are resistant to diuretic combinations may respond favorably to short-term courses of intravenous albumin. Although nonprotein colloids such as hydroxyethyl starch or polymerized gelatin solutions are less expensive, the albumin substitutes may exacerbate bleeding problems in these patients and can also be immunogenic .
Low dose dopamine infusions dilate renal arterioles, inhibit renal epithelial sodium transport, and produce a brisk diuresis in normal subjects. On occasion, the addition of dopamine or dopamine agonists will generate a diuresis in otherwise refractory patients with cirrhosis . In general, this is reserved for patients with near end-stage hepatic decompensation.
The goal of diuretic therapy in patients with cirrhosis is a reduction of the volume of ascites and degree of edema to clinically tolerable levels. Overdi-uresis will result in orthostatic hypotension, organ ischemia, metabolic derangements, and progressive cardiovascular collapse. Aggressive diuresis has also been implicated in the development of fatal hepatorenal syndrome. Patients treated with diuretics require careful monitoring and the diuresis must be slowed or stopped when complications develop. Clinical parameters including the weight and the supine and upright pulse rate and blood pressure should be monitored regularly. The BUN, creatinine, and electrolyte concentrations must also be measured at regular intervals.
The BUN of cirrhotic patients may be reduced as a result of their low dietary protein intake and decreased hepatic urea synthesis. Their creatinine concentration is also often reduced because these patients frequently have muscle wasting. Consequently, both the BUN and creatinine concentrations may overestimate the GFR. Acute changes in the BUN and creatinine concentrations and the BUN/creatinine ratio can be very helpful markers of renal perfusion and ischemia (see Section on End Points for Diuresis, Chapter VA2).
Cirrhotic patients with ascites can be divided into two groups on the basis of the presence or absence of significant peripheral edema. Shear and co-workers demonstrated that edema fluid can be mobilized into the vascular compartment much more rapidly than ascitic fluid . This permits a diuresis to more rapidly remove edema than ascites. Peripheral edema can be mobilized at rates up to one liter/day, while ascites returns to the vascular space at only about half this rate. Consequently, cirrhotic patients with generalized edema and ascites can tolerate aggressive diuresis with fewer adverse hemodynamic and biochemical derangements than those without edema. The maximal rate of weight reduction in the cirrhotic patient with ascites and generalized edema should be no more than 1-1.5 kg/day, while it is only about 0.3-0.5 kg/day in those without edema [21, 26].
Renal underperfusion combined with persistently high ADH levels produces continuous concentration of the urine. Therefore, water loads cannot be excreted and hyponatremia is common in cirrhotic patients. The development of hyponatremia indicates that the neurohormonal forces which reduce renal dilation capacity are activated and suggests that the liver disease is advanced. Consequently, spontaneous hyponatremia in cirrhotic patients is associated with a poor prognosis (this is also true in patients with CHF) [ 1 ]. The tendency of these patients to develop hyponatremia is exacerbated by aggressive diuresis, especially when thiazides are utilized. The thiazides reduce renal diluting capacity but do not decrease renal concentrating function. Consequently, patients who receive thiazides are still capable of excreting very concentrated urine in response to renal hypoperfusion, increased proximal tubule reabsorption and high ADH levels. Loop diuretics impair both diluting and concentrating mechanisms and are less likely to exacerbate hyponatremia. Water restriction should be initiated when the sodium concentration falls below 130 mEq/liter. In the near future drugs which specifically block the renal effects of ADH should become available for use in patients with persistent hyponatremia.
Thiazide and/or loop diuretics increase delivery of sodium-rich fluid to the aldosterone-sensitive distal nephron. This markedly increases distal renal tubule potassium and hydrogen secretion. Increased distal salt and water delivery will dilute the tubule fluid potassium concentration and improve the gradient for potassium secretion. In addition, distal delivery of sodium salts to aldoste-rone-stimulated tubules increases distal sodium reabsorption and the secretion of potassium and hydrogen. Hypokalemia and metabolic alkalosis develop. The alkalosis is then be maintained as a result of hypokalemia, decreased EABV, and high angiotensin II levels. High ADH levels will also stimulate distal tubule potassium secretion. A prospective study of furosemide therapy in cirrhotic patients reported clinically significant hypokalemia developed in 16% [ 18]. The combination of hypokalemia and metabolic alkalosis has a number of important adverse effects.
The electrolyte abnormalities produced by diuretic therapy can precipitate or exacerbate hepatic encephalopathy. Hepatic encephalopathy is partly due to the accumulation of ammonia and other ionized nitrogenous compounds within the brain. Hypokalemia is a potent stimulus for renal ammonia generation. In part, this is due to movement of potassium out of renal tubule cells and the movement of protons into these cells. This causes relative intracellular acidosis. A low intracellular pH in the renal proximal tubule cells stimulates am-moniagenesis. Ammonia synthesized by renal tubule cells exits the kidney either by entering the urine or via the renal vein. Many factors affect the ammonia partition ratio between urine and renal vein, but one of the most important is the pH of renal distal tubule fluid and urine. An acid pH in the distal tubule fluid shifts tubule fluid ammonia from NH3 to NH4+ as a result of the reaction NH . + H+ —> NH4+. NH3 is much more permeable across cell membranes than ionized NH4+. The low pH within the lumen reduces the NH3 concentration and this increases the gradient for NH3 to diffuse from renal tubule cells and interstitium into the tubule lumen. Thus, acidification of distal tubule fluid "traps" ammonia in urine and increases the fraction of total ammonia which is excreted rather than transported to the systemic circulation via the renal vein (see Fig. 3). Conversely, an alkaline urine produces the opposite effect, reducing the fraction of ammonia trapped in the urine and this increases the fraction of total synthesized ammonia which enters the renal vein. As a result hypokalemic metabolic alkalosis can increase plasma ammonia levels markedly by (i) increasing total renal production and (ii) favoring its transfer to the systemic circulation.
Hypokalemic alkalosis has additional adverse systemic ammonia effects. The principles of nonionic diffusion trapping discussed above in reference to the kidney and renal tubules also apply at the ECF/intracellular fluid interface. An alkaline blood pH increases the relative concentration of blood NH3 by shifting the reaction NH3 + H ^ NH4+ to the left. A higher NH3 concentration increases the entry of NH3 into brain cells. In addition, as discussed, hypokalemia can produce relative intracellular acidosis (potassium moves out of cells into the ECF and protons move in the opposite direction). The lower intracellular pH shifts the above reaction to the right and decreases the intracellular NH3 concentration. The net effect is a gradient which favors movement of ammonia into cells (see Fig. 3).
Consequently, hypokalemic alkalosis increases the total renal ammonia synthetic rate and elevates the proportion of renal ammonia production which en-
Brain Cell Blood Plasma Kidney
Was this article helpful?
A Hard Hitting, Powerhouse E-book That Is Guaranteed To Change The Way You Look At Your Health And Wellness... Forever. Everything You Know About Health And Wellness Is Going To Change, Discover How You Can Enjoy Great Health Without Going Through Extreme Workouts Or Horrendous Diets.