Acetaminophen Poisoning

Diagnosis 

Prompt recognition of APAP toxicity is essential to preventing morbidity and mortality. This recognition is made difficult by the nonspecific clinical findings early in the course of APAP toxicity. The first 24 hours are considered the first phase of APAP poisoning and are characterized by nonspecific findings, such as nausea, vomiting, anorexia, pallor, and lethargy. In this phase, however, the patient may appear normal. In the second phase of APAP poisoning, the patient begins to develop clinical and laboratory evidence of hepatotoxicity. Classically, it is taught that APAP-induced liver dysfunction occurs only after a latent period of 24 to 48 hours. However, Singer and colleagues [3] demonstrated that liver enzymes often become elevated in the initial 24 hours. In the third phase, the patient progresses to FHF with all its associated complications. Phase four, usually occurring 72 to 96 hours after ingestion, is the resolution of liver function and complete recovery if the patient survives the initial insult.

APAP toxicity should be suspected in all intentional overdoses because of the ubiquitous nature of APAP. The early symptoms of APAP ingestion are often vague or absent, leading to concern that potentially toxic ingestions could be missed owing to the poor histories often encountered in the overdose patient. To address this concern, several studies have attempted to evaluate the role of screening all intentional ingestions with serum APAP levels. Ashbourne and colleagues [4] found that, in patients without history of APAP ingestion, approximately 1 in 70 had detectable APAP levels and approximately 1 in 500 had potentially hepatotoxic APAP levels. A large retrospective study by Sporer and colleagues [5] found similar results: the rate of potentially hepatotoxic APAP levels in patients without history of ingestion was 0.3%.

Based on the wide availability of APAP, the low cost of quantitative APAP levels, and the potentially disastrous consequences of unrecognized APAP ingestion, all intentional overdoses should be screened for APAP toxicity. Since all overdose patients should be screened for APAP, it is equally important that all acute care facilities have the ability to rapidly run an APAP assay.

It should be noted the false positive APAP levels have been seen in patients with hyperbilirubinemia. The actual threshold for bilirubin's interference with the APAP assay is unclear. Because of the treatment and prognostic decisions that could follow a misdiagnosis of APAP induced hepatitis, caution is warranted and the clinician should be aware of this lab phenomenon [6], [7].

APAP and histopathology 

Predicatable histopathologic changes occur following acetaminophen poisoning. From autopsy and liver biopsy specimens, a characteristic pattern is observed in the liver [8], [9], [10]. Microscopic examination shows necrosis of the centrilobular hepatocytes. Associated findings include passive congestion, scattered mononucleated and polynucleated leukocytes and no fatty infiltration. Reticulin stain reveals maintenace of the hepatic architecture in areas of hepatocyte necrosis. In one case, a liver biopsy during the acute phase of an acetaminophen poisoning demonstrated the above changes. A wedge biopsy was obtained 4 years after that showed normal liver architecture and no fibrosis or cirrhosis [8]. However, additional fibrotic changes have been observed months after acute poisoning in association with chronic alcoholism [11]. The kidneys show diffuse acute tubular necrosis with tubular epithelial swelling and fatty vacuolization [9], [10]. Autopsies have also demonstrated diffuse acute pancreatitis and diffuse cerebral edema with brainstem herniation [9], [10].

Risk factors 

Several populations of patients are at increased risk for hepatotoxicity from APAP. An understanding of the pathophysiology is necessary to evaluate the need for treatment in those who may be at increased risk.

With therapeutic dosing, 90% of APAP is conjugated with glucuronide or sulfate to form nontoxic metabolites. Approximately 5% of APAP is metabolized by the hepatic cytochrome p450 mixed-function oxidase enzyme to a toxic metabolite, N-acetyl-p-benzoquinoneimine (NAPQI). In normal dosing, NAPQI is rapidly detoxified by glutathione (GSH) to nontoxic metabolites. Acetaminophen overdoses overwhelm conjugation pathways, resulting in increased use of the cytochrome p450 pathway and increased formation of NAPQI, increased depletion of GSH, and, ultimately, hepatic injury. Because APAP is metabolized to its toxic metabolite, NAPQI, by the p450 system, any agent that induces this system theoretically increases the risk for APAP hepatotoxicity. However, few interactions have been reported or formally studied [12], [13], [14], [15], [16]. Ethanol is the best-studied inducer, with chronic alcohol ingestion inducing the p450 enzyme system and in turn making the liver capable of metabolizing more APAP to the toxic metabolite NAPQI. This interaction has been demonstrated in animal models, but its clinical relevance is unknown. Currently, chronic alcoholics who acutely ingest APAP should be treated in the same manner as other patients, by obtaining a 4-hour APAP level and plotting it on the nomogram [16], [17], [18], [19], [20].

Chronic alcoholics have been shown to have significantly lower levels of plasma GSH than controls: this may be a risk factor for APAP toxicity [20]. To investigate the influence of acute and chronic alcohol intake on the clinical course and outcome, one study looked at 209 patients who had an acute APAP overdose. This study showed that, after one has corrected for all other factors, chronic alcohol intake is associated with a fivefold increase in the relative risk of developing hepatic encephalopathy. The study was unable, however, to demonstrate a significant change in mortality [21].

Because alcohol and APAP are metabolized by the same enzymes system, some have speculated that they may competitively inhibit each other's metabolism. Acute alcohol intake does not independently influence the clinical course of APAP overdose in humans [21]. By contrast, some animal studies have shown a marginal benefit from acute alcohol ingestion in APAP overdoses [12], [22], [23]. The clinical significance of acute alcohol ingestion is probably minimal and should not affect treatment decisions.

Specific drugs metabolized by the p450 system may theoretically play a role in the formation of NAPQI. For example, isoniazid is often implicated as an inducer of the p450 system and thereby may play a role in APAP toxicity [24], [25], [26], [27], [28]. APAP is metabolized through a specific subset of enzymes (CYP2E) in the p450 system, and only drugs actually inducing this subset of enzymes affect the formation of NAPQI [16].

GSH is essential for the detoxification of NAPQI. Patients who are at increased risk for APAP hepatotoxicity also include those with decreased GSH stores. The most common reason for decreased GSH stores is malnutrition—for example, anorexia nervosa [29] and chronic alcoholism [19]. Reports also exist of APAP-induced hepatotoxicity associated with fasting from febrile illnesses and chronic disease [30], [31]. These risk factors do not appear to play a major role in acute ingestions but may play an important one in nonacute ingestions.

Treatment 

The Rumack-Mathew treatment nomogram is the primary tool used to guide treatment after acute ingestion of APAP. A firm understanding of the nomogram's use and its limitations is essential in the management of APAP exposure. The nomogram was first studied retrospectively in 64 cases of acute ingestion of APAP in an attempt to correlate APAP serum levels with hepatotoxicity. Hepatotoxicity was arbitrarily defined as an aspartate aminotransferase (AST) level of 1000 IU/L. Serum APAP levels at or above a line connecting 200 μg/mL at 4 hours postingestion and 6.25 μg/mL at 24 hours postingestion consistently predicted hepatotoxicity [32]. This line is referred to as the probable toxicity line. In Europe, a level above this line is an indication for NAC therapy. When the nomogram was introduced in the United States, the US Food and Drug Administration (FDA) insisted on a 25% reduction of this treatment threshold. A line connecting 150 μg/mL at 4 hours and 4.7 μg/mL at 24 hours, which is considered the possible toxicity line, is used for a treatment threshold [16]. The nomogram was later validated in a large trial using the 72-hour NAC protocol [33].

Acute poisoning, which has not been well defined in the literature, is generally regarded as a single ingestion of APAP over less than a 4-hour time period. The Rumack-Mathew nomogram can only be used to determine the need for NAC treatment after an acute ingestion [16]. All other ingestions are not applicable to the nomogram. Typically, the potentially toxic dose is 150 mg/kg, or 15 g in the healthy adult [34]. These doses would correspond to a potentially toxic level on the nomogram.

After an acute ingestion, the serum APAP levels should be drawn 4 hours after ingestion or at any time thereafter and plotted on the nomogram. Patients with APAP levels falling above the possible toxicity line should be treated with NAC. The period of 4 hours after ingestion was arbitrarily chosen to assume full absorption and a peak serum APAP level [16].

The time of ingestion is an essential piece of history when plotting levels on the nomogram. However, the exact time of ingestion is not always clear. Patients' histories are frequently unreliable or intentionally misleading. When possible, it is helpful to corroborate a patient's history with a third party. The earliest possible time of ingestion should be used when no reliable history is available. When in doubt, treatment with NAC is always the safest option when the time of ingestion remains uncertain.

Patients who present late after acute APAP ingestion may present some difficulties in determining the need for NAC therapy, because, as the time after ingestion approaches 24 hours, the treatment line on the nomogram approaches the lower limits of detection of some laboratories. In these cases, the need for NAC treatment is determined using a serum APAP, AST, and ALT levels, along with the patient's reported amount of ingestion (usually >7.5 g in an adult or 140 mg/kg in a child [33]) and the patient's risk factors for APAP toxicity. If APAP is detectable, it should be plotted on the nomogram, and treatment should follow if indicated. When APAP is undetectable and AST and ALT are elevated, it should be assumed that the patient is in the second clinical phase of APAP toxicity, and NAC therapy should be initiated. In patients who have undetectable APAP and normal AST and ALT, NAC may probably be safely withheld. Concern exists that impending APAP-induced hepatotoxicity may be preceded by a brief window of time, 24 hours after ingestion but before AST and ALT begin to rise, in which APAP is undetectable and AST and ALT are normal. Such cases are not reported in the literature. However, because of this theoretic concern, some physicians institute NAC therapy in patients who are at increased risk for APAP-induced hepatotoxicity despite normal laboratory values.

Originally, it was believed that NAC therapy was ineffective when started after 15 hours; therefore, it was frequently withheld from late presenters [34], [35], [36], [37]. It has since been shown that NAC provides significant benefit to patients presenting with APAP-induced FHF regardless of the time after ingestion [38], [39], and it is recommended in any patient with APAP-induced elevations in liver enzymes.

Prognosis 

Patients who are APAP poisoned have a wide spectrum of clinical presentations. Most present in the first 8 hours and before any signs of hepatotoxicity develop. ICU resources are limited and are not necessary for many APAP-poisoned patients. Close observation for progression of serious toxicity is the most common reason for critical care use. Most patients do not require critical care resources for invasive procedures. Patients who exhibit signs of serious hepatotoxicity, such as coagulopathy, encephalopathy, and metabolic derangements, will likely benefit from the ICU and may need invasive procedures. Patients who have renal failure or other end-organ effects are also candidates for the ICU. Patients who have coingestants with a potential for serious toxic effects may also need close monitoring.

NAC therapy is not an indication for critical care admission. Asymptomatic patients presenting early after ingestion may be cared for on a general ward unless concerns about suicide or self-harm warrant otherwise.

It is important for the clinician to identify patients at increased risk for FHF so as to begin the process of transfer to a specialized center should hepatic transplant become necessary. Although the AST and ALT levels are frequently the first laboratory sign of liver injury in APAP toxicity, the rate of rise and peak level give no indication of prognosis. A collection of clinical parameters and laboratory findings known as the King's College Criteria have been established and well validated to predict poor outcome from APAP toxicity. Patients who have isolated APAP overdoses and develop significant metabolic acidosis (pH <7.3 after adequate fluid resuscitation), a serum creatinine >3.3 mg/dL, a prothrombin time >1.8 times control (>100 seconds or an international normalized ratio [INR] >6.5), or grade III or higher encephalopathy have a poorer prognosis [82]. Any patient meeting any one of these criteria should be transferred to a tertiary care center in anticipation of FHF and possible transplantation [82]. These criteria have also been validated by subsequent studies [83], [84], [85].

The King's College Criteria (KCC) are well established, but they are not without their limitations. With the exception of fluid resuscitation, therapeutic interventions affecting these criteria may negate their predictive value. However, once criteria are met, further intervention is no longer problematic. For instance, creatinine levels of 3.3 mg/dL or greater already put the patient at higher risk; hence subsequent measurements with or without dialysis are irrelevant.

Monitoring of coagulopathy is also problematic. To establish a rise in prothrombin time, serial levels must be drawn, a process that again may delay identification of patients at risk for FHF. In one series, a prothrombin time >180 at 4 days was highly suggestive of death [86]. Although this may provide useful information to the liver transplant service, waiting 4 days before transfer to a tertiary center is clearly not beneficial to the patient. The original KCC were developed before the widespread use of the INR, which makes comparison among prothrombin times difficult.

Reported cases also exist of prolongation of prothrombin time without clinical or laboratory signs of hepatotoxicity [87]. In fact, in both retrospective and small prospective studies, prothrombin time prolongation appeared to correlate with IV NAC use rather than hepatic failure [88], [89], [90]. This rise in INR without overt FHF may be due, in part, to decreased factor VII levels [88].

Coagulopathy in the face of FHF, and the associated increase in INR, is due to the reduction of clotting factors. The admission ratio of factor VIII to factor V, and the overall concentration of factor V have both been found to be helpful in predictors of outcome. Factor V levels less than 10 percent were a poor prognostic findings. A factor VIII/V ratio of greater than 30 is also a poor prognostic sign [91]. Because of these observations, some recommend factor monitoring [87]. While this may be useful for the transplant team, it should not be relied upon as a screening tool.

Treatment of coagulopathy can be accomplished with administration of FFP. While many clinicians treat this coagulopathy without signs of bleeding, there is evidence that it is safe to withhold treatment in APAP induced coagulopathy. Gazzard et al compared two groups with APAP induced coagulopathy. The control group was treated with FFP while the experimental group was not unless the INR was over 7. There was no difference in outcome or complications. In the face of bleeding FFP and factor administration is necessary, but with bleeding, it may be possible to with hold blood products [92].

The limitations of the KCC have led to a search for other screening tools. Although many parameters have been suggested and evaluated in small numbers, none are currently as well established or widely used as the KCC.

The acute physiology and chronic health evaluation II (APACHE II) score has shown promise as an early and accurate indicator of impending hepatic failure and the need for transplantation [93]. In a small study, this evaluation was found to be more sensitive and specific than KCC. It was based on patient assessment on the first day of hospital admission, making it possible to determine the need for transplant referral at an earlier time. The disadvantage of the APACHE is that it is cumbersome and difficult to remember. Although this evaluation is intriguing, further study is required.

Changes in serum phosphate levels have been suggested as a prognostic indicator in APAP-induced hepatotoxicity. One prospective study looking at 125 patients who had suspicion of severe APAP-induced hepatotoxicity evaluated the prognostic value of serial serum phosphate measurements [94]. The study revealed that a serum phosphate level >1.2 mmol/L 48 to 96 hours after an APAP overdose specifically and sensitively identifies patients with little chance of survival. The authors hypothesize that the liver consumes phosphate during hepatic regeneration. Thus, phosphate levels represent a balance of renal failure and liver regeneration. Similar findings in patients who have FHF have been reported [95]; however, other studies show mixed results [96]. Further study is needed to clarify the role of phosphate as a prognostic indicator.

Elevated arterial lactate has been proposed as a prognostic marker for poor outcomes in APAP overdose. A retrospective cohort study revealed that lactate levels >3.5 mmol/L before fluid resuscitation or >3.0 mmol/L after fluid resuscitation were sensitive and specific indicators of survival [97]. A subsequent prospective cohort study was performed for validation [97]. Comparison with the KCC revealed that early lactate levels >3.5 mmol/L identified at-risk patients earlier but had a lower sensitivity and accuracy than the KCC criteria. Postresuscitation lactate concentrations >3.0 mmol/L had equivalent sensitivity and higher specificity and accuracy than the KCC criteria. However, there was no significant difference in time to identification of patients between KCC criteria and postresuscitation lactate [97].

Recently, investigators have shown interest in hyperamylasemia as a possible prognostic indicator in APAP toxicity. A large retrospective study at a Danish referral center found amylase levels >1.5 times normal to correlate with poor outcome and the need for transplantation. Hyperamylasemia did not appear to be due to clinical pancreatitis; however, speciation of the amylase was not performed [98]. Prospective validation of amylase with speciation is needed.

Although it does not directly predict the need for transplantation, a novel approach using easily measurable criteria has been prospectively used to predict encephalopathy in APAP toxicity. Time to NAC therapy, platelet count, and INR are used to determine a prognostic index [99]. Because hepatic encephalopathy requires specialized care, referral based on these criteria may prove useful.

Level of alfa-fetoprotein (AFP) be used to predict favorable outcome in APAP induced FHF. A small increase in AFP one day after peak of ALT was found in survivors compared to non-survivors. Because this increase was small, a highly sensitive assay is required to make reliable predictions [100].

Summary 

APAP is likely to remain a common toxic exposure and continue to cause significant morbidity and mortality. To minimize the harm to patients, it is necessary for the clinician to be aware of the current diagnostic and therapeutic management of APAP poisoning. Despite the bulk of literature on APAP, management strategies are likely to continue to change as more studies are conducted to improve our understanding of nonacute ingestions and the role of prognostic markers in defining those most at risk for life-threatening hepatotoxicity.