Toxic Effects of Nucleoside Reverse Transcriptase Inhibitors on the Liver- 4 of 4

[Guest guest]

Toxic Effects of Nucleoside Reverse Transcriptase Inhibitors

from American Journal of Clinical Pathology

Discussion

We have described NRTI-induced hepatotoxic effects in 13 HIV-infected

patients with proof of the mitochondrial toxic effects by liver biopsy.

Excluding the classic but rare dramatic hepatic failure with lactic

acidosis,[8-11, 18] our major finding is that NRTI-induced hepatotoxic

effects may be present as a mildly symptomatic and nonspecific disease.

In the present study, although liver biologic features always were abnormal,

NRTI-induced toxic effects were clinically suspected in fewer than half of

the cases, based on elevated serum lactate levels or extrahepatic symptoms

(ie, neuromuscular abnormalities). The remaining patients were mildly

symptomatic, complaining of asthenia, weight loss, or nausea. In patients

with HIV infection in whom the liver test results were abnormal, the

differential diagnosis of drug-induced toxic effects includes opportunistic

infections, hepatic neoplasia, alcoholic liver disease, and the toxic

effects of intravenous drug abuse. Concerning the hepatotoxic effects of

antiretroviral drugs, most patients are treated simultaneously with several

potentially toxic drugs. Liver toxic effects have been well described with

NNRTIs and PIs.[19-21] Nevirapine and efavirenz have been associated with

hepatotoxic effects.[20] The clinical and histologic features of the toxic

effects of PIs seem to be quite variable and have been described mainly with

ritonavir and indinavir.[19, 21] Moreover, an assessment of synergistic

toxic effects of the combination of multiple NRTIs, NNRTIs, and PIs is

particularly difficult.

Concerning NRTI class, mildly abnormal serum liver test results are

well-known side effects of HAART and of regimens containing AZT, ddI, or

stavudine.[1, 15] A study based on abnormalities of liver test results and

not on liver biopsies found an incidence of hepatic toxic effects during

NRTI-containing regimens in 21 cases per 1,000 treated person-years.[22]

Several studies have evaluated the hepatotoxic properties of different drugs

of the NRTI class.[6, 23] Among these drugs, stavudine alone or associated

with didanosine (9 patients from our series were treated with this

combination) has been involved frequently in hepatotoxic effects.[24]

Diffuse macrovesicular steatosis, megamitochondria, and cholestasis that

strongly suggest NRTI-induced toxic effects were not a constant feature in

our series. Severe microvacuolar steatosis was not observed in our series,

probably owing to the case selection that excluded patients with liver

failure. Additional lobular inflammation and acidophilic or ballooning

hepatocyte necrosis with Mallory bodies also were observed frequently in our

series. These findings may support an associated drug-induced

steatohepatitis.[10, 25, 26] Perisinusoidal fibrosis is a common finding in

liver biopsy specimens of HIV-infected patients and patients coinfected with

HIV and HCV.[27] Several physiopathologic hypotheses have been proposed to

explain such a fibrosis. A direct toxic effect of the NRTI via mitochondrial

defects cannot be excluded.[3]

To date, ultrastructural studies and biochemical mitochondrial tests are

available to confirm the diagnosis of NRTI-induced mitochondrial

toxicity.[23, 28, 29] We confirmed the ultrastructural mitochondrial

abnormalities, mainly reported in clinical and experimental studies in the

liver[30] and in muscular tissue.[7] These mitochondrial alterations include

megamitochondria, loss of the cristae, autophagic vacuoles, and

electron-dense and crystalline inclusions. The absence of mitochondrial

abnormalities in the 2 control patients supports that these ultrastructural

defects are specific to NRTI-induced mitochondrial cytopathy. Only a few

studies have dealt with liver ultrastructural findings and mitochondrial

pathology in HIV-infected patients.[31, 32] Megamitochondria and

mitochondrial inclusions (without any other features reported in the present

study) have been reported, but these studies were limited strongly by the

absence of clinical and biologic data.

Based on these mitochondrial defects, electron microscopic study permitted

the diagnosis of NRTI-induced hepatotoxic effects in all cases without

typical diffuse steatosis. However, and in view of the limited number of

cases in our study, the severity of mitochondrial damage did not correlate

with the severity of steatosis and with the lactate level. Our EM findings

in mildly symptomatic patients are similar to those reported during dramatic

lactic acidosis and severe liver damage.[9, 11, 17] This absence of

correlation among histologic findings, electron microscopic findings, and

liver biologic test results suggests that EM assessment is valuable in the

positive diagnosis of NRTI-induced mitochondrial cytopathy. However, EM

findings cannot be considered a prognostic factor of liver involvement.

The clinical management of patients with NRTI-induced hepatotoxic effects is

still a matter of debate. Although discontinuation of the NRTI is required

in cases of life-threatening lactic acidosis, there still is insufficient

information on the follow-up of patients with only mild hepatic

abnormalities. In the present study, discontinuation of the NRTI was

associated with improved liver test results. Some authors have proposed

regular monitoring of liver enzyme levels without discontinuation of the

HAART regimen if the aminotransferase levels remain less than 5 times the

upper limit of the reference range.[3, 24, 33]

Liver biopsy is necessary for the initial evaluation of liver dysfunction in

patients with HIV infection who are treated by HAART and with NRTI-based

regimens. The diagnosis of adverse events associated with NRTIs is important

because of the virologic implications of discontinuing antiretroviral

treatment. The ultrastructural assessment of mitochondrial abnormalities is

a valuable tool for the diagnosis of NRTI-induced toxic effects because

standard histologic examination may be nonspecific in about 50% of cases.

Funding Information

Supported by a grant from the French Society of Pathology, Paris, and by

grant 2000/151 from the National Agency for AIDS Research, Paris.

Reprint Address

Address reprint requests to Dr Duong Van Huyen: Laboratoire d'Anatomie

Pathologique, Hôpital Européen s Pompidou, 20 rue Leblanc, 75908 Paris

Cedex 15, France.

[Guest guest]

Toxic Effects of Nucleoside Reverse Transcriptase Inhibitors

from American Journal of Clinical Pathology

Discussion

We have described NRTI-induced hepatotoxic effects in 13 HIV-infected

patients with proof of the mitochondrial toxic effects by liver biopsy.

Excluding the classic but rare dramatic hepatic failure with lactic

acidosis,[8-11, 18] our major finding is that NRTI-induced hepatotoxic

effects may be present as a mildly symptomatic and nonspecific disease.

In the present study, although liver biologic features always were abnormal,

NRTI-induced toxic effects were clinically suspected in fewer than half of

the cases, based on elevated serum lactate levels or extrahepatic symptoms

(ie, neuromuscular abnormalities). The remaining patients were mildly

symptomatic, complaining of asthenia, weight loss, or nausea. In patients

with HIV infection in whom the liver test results were abnormal, the

differential diagnosis of drug-induced toxic effects includes opportunistic

infections, hepatic neoplasia, alcoholic liver disease, and the toxic

effects of intravenous drug abuse. Concerning the hepatotoxic effects of

antiretroviral drugs, most patients are treated simultaneously with several

potentially toxic drugs. Liver toxic effects have been well described with

NNRTIs and PIs.[19-21] Nevirapine and efavirenz have been associated with

hepatotoxic effects.[20] The clinical and histologic features of the toxic

effects of PIs seem to be quite variable and have been described mainly with

ritonavir and indinavir.[19, 21] Moreover, an assessment of synergistic

toxic effects of the combination of multiple NRTIs, NNRTIs, and PIs is

particularly difficult.

Concerning NRTI class, mildly abnormal serum liver test results are

well-known side effects of HAART and of regimens containing AZT, ddI, or

stavudine.[1, 15] A study based on abnormalities of liver test results and

not on liver biopsies found an incidence of hepatic toxic effects during

NRTI-containing regimens in 21 cases per 1,000 treated person-years.[22]

Several studies have evaluated the hepatotoxic properties of different drugs

of the NRTI class.[6, 23] Among these drugs, stavudine alone or associated

with didanosine (9 patients from our series were treated with this

combination) has been involved frequently in hepatotoxic effects.[24]

Diffuse macrovesicular steatosis, megamitochondria, and cholestasis that

strongly suggest NRTI-induced toxic effects were not a constant feature in

our series. Severe microvacuolar steatosis was not observed in our series,

probably owing to the case selection that excluded patients with liver

failure. Additional lobular inflammation and acidophilic or ballooning

hepatocyte necrosis with Mallory bodies also were observed frequently in our

series. These findings may support an associated drug-induced

steatohepatitis.[10, 25, 26] Perisinusoidal fibrosis is a common finding in

liver biopsy specimens of HIV-infected patients and patients coinfected with

HIV and HCV.[27] Several physiopathologic hypotheses have been proposed to

explain such a fibrosis. A direct toxic effect of the NRTI via mitochondrial

defects cannot be excluded.[3]

To date, ultrastructural studies and biochemical mitochondrial tests are

available to confirm the diagnosis of NRTI-induced mitochondrial

toxicity.[23, 28, 29] We confirmed the ultrastructural mitochondrial

abnormalities, mainly reported in clinical and experimental studies in the

liver[30] and in muscular tissue.[7] These mitochondrial alterations include

megamitochondria, loss of the cristae, autophagic vacuoles, and

electron-dense and crystalline inclusions. The absence of mitochondrial

abnormalities in the 2 control patients supports that these ultrastructural

defects are specific to NRTI-induced mitochondrial cytopathy. Only a few

studies have dealt with liver ultrastructural findings and mitochondrial

pathology in HIV-infected patients.[31, 32] Megamitochondria and

mitochondrial inclusions (without any other features reported in the present

study) have been reported, but these studies were limited strongly by the

absence of clinical and biologic data.

Based on these mitochondrial defects, electron microscopic study permitted

the diagnosis of NRTI-induced hepatotoxic effects in all cases without

typical diffuse steatosis. However, and in view of the limited number of

cases in our study, the severity of mitochondrial damage did not correlate

with the severity of steatosis and with the lactate level. Our EM findings

in mildly symptomatic patients are similar to those reported during dramatic

lactic acidosis and severe liver damage.[9, 11, 17] This absence of

correlation among histologic findings, electron microscopic findings, and

liver biologic test results suggests that EM assessment is valuable in the

positive diagnosis of NRTI-induced mitochondrial cytopathy. However, EM

findings cannot be considered a prognostic factor of liver involvement.

The clinical management of patients with NRTI-induced hepatotoxic effects is

still a matter of debate. Although discontinuation of the NRTI is required

in cases of life-threatening lactic acidosis, there still is insufficient

information on the follow-up of patients with only mild hepatic

abnormalities. In the present study, discontinuation of the NRTI was

associated with improved liver test results. Some authors have proposed

regular monitoring of liver enzyme levels without discontinuation of the

HAART regimen if the aminotransferase levels remain less than 5 times the

upper limit of the reference range.[3, 24, 33]

Liver biopsy is necessary for the initial evaluation of liver dysfunction in

patients with HIV infection who are treated by HAART and with NRTI-based

regimens. The diagnosis of adverse events associated with NRTIs is important

because of the virologic implications of discontinuing antiretroviral

treatment. The ultrastructural assessment of mitochondrial abnormalities is

a valuable tool for the diagnosis of NRTI-induced toxic effects because

standard histologic examination may be nonspecific in about 50% of cases.

Funding Information

Supported by a grant from the French Society of Pathology, Paris, and by

grant 2000/151 from the National Agency for AIDS Research, Paris.

Reprint Address

Address reprint requests to Dr Duong Van Huyen: Laboratoire d'Anatomie

Pathologique, Hôpital Européen s Pompidou, 20 rue Leblanc, 75908 Paris

Cedex 15, France.