Interferon alfa

Cardiac complications

Severe or life-threatening cardiotoxicity is infrequent and mostly reported in the form of a subacute complication in patients with cancer, and in those with pre-existing heart disease or receiving high-dose interferon alfa. Atrioventricular block, life-threatening ventricular dysrhythmias, pericarditis, dilated cardiomyopathy, cardiogenic shock, asymptomatic or symptomatic myocardial ischemia or even infarction, and sudden death have been observed [ ]. The combination of high-dose interleukin-2 (IL-2) with interferon alfa enhanced cardiovascular complications, namely cardiac ischemia and ventricular dysfunction [ ].

Cardiomyopathy has been attributed to interferon alfa in an infant [ ].

• A 3-month-old boy was given interferon alfa (2.5–5.5 MU/m ² ) for chronic myelogenous leukemia. After 7.5 months he developed progressive respiratory distress, with anorexia, irritability, and nocturnal sweating. A chest X-ray showed cardiomegaly, an echocardiogram showed a markedly dilated left ventricle, and an electrocardiogram showed left ventricular hypertrophy with abnormal repolarization. Viral cultures and serology for cytomegalovirus, parvovirus B19, and enterovirus were negative. Infectious diseases and metabolic disturbances were excluded. Interferon alfa was withdrawn and digoxin, furosemide, and an angiotensin-converting enzyme inhibitor were given. One year later, he was asymptomatic without further cardiac treatment.

Similar, but anecdotal reports were also described in patients without evidence of previous cardiac disease and receiving low-dose interferon alfa [ , ]. In chronic viral hepatitis, only seven of 11 241 patients had severe cardiac adverse reactions [ ]. The exact risk of such cardiovascular adverse reactions is unknown. In patients with chronic viral hepatitis, cardiovascular test results were not modified when patients were re-examined after at least 6 months of treatment, even where there was an earlier cardiac history [ ], but there was a potentially critical reversible reduction in left ventricular ejection of more than 10% in another prospective study [ ].

Myocardial dysfunction can completely reverse after withdrawal of interferon alfa and does not exclude further treatment with lower doses [ ].

• 47-year-old man with renal cancer and no previous history of cardiovascular disease developed gradually worsening exertional dyspnea after he had received interferon alfa in a total dose of 990 MU over 5 years. Echocardiography and a myocardial CT scan confirmed a dilated cardiomyopathy, with left ventricle dilatation and diffuse heterogeneous perfusion at rest. He improved after interferon alfa withdrawal and treatment with furosemide, quinapril, and digoxin. Myocardial scintigraphy confirmed normal perfusion. He restarted low-dose interferon alfa (6 MU/week) 1 year later and had no recurrence of congestive heart failure after a 1-year follow-up period.

• Acute myocardial infarction occurred about 12 hours after the administration of pegylated interferon alfa-2b 1 microgram/kg in a healthy 76-year-old woman during a single-dose pharmacokinetic study in 24 patients, of whom 18 were over 65 years [ ].

Although the case was poorly documented, the close temporal relation strongly suggested a causal role of the drug.

Pulmonary hypertension has previously been reported in one patient and was briefly mentioned as a potential complication of interferon alfa in two other patients who also had multiple ulcers involving the feet or toes [ ]. Other vascular events including Raynaud’s phenomena, digital ulcerations and gangrene, pulmonary vasculitis, and thrombotic thrombocytopenic purpura are occasionally reported [ ].

Pericarditis has been attributed to interferon alfa [ ].

• A 40-year-old woman with chronic hepatitis C developed pericarditis without the typical features of a lupus-like syndrome after 4 months of treatment with low-dose interferon alfa-2b in [ ]. She simultaneously developed a polyneuropathy. Both disorders disappeared after interferon withdrawal.

• A 42-year-old woman with a history of atrioventricular septal defect and atrial fibrillation was given interferon alfa-2b for chronic hepatitis C. Pre-treatment echocardiography showed moderate mitral valve disease and left atrial hypertrophy but no pericardial anomalies. Six hours after her first injection of interferon alfa (3 MU) she had an acute episode of thoracic pain, which disappeared spontaneously within 24 hours. Echocardiography showed a moderate non-constrictive pericardial effusion. Similar symptoms recurred 7 hours after a subsequently reduced dose of interferon alfa (1 MU), and echocardiography again showed a reversible pericardial effusion. Autoimmune diseases, cryoglobulinemia, and myocardial infarction were ruled out. Interferon alfa was withdrawn and no recurrence of pericarditis was noted over 5 years of follow-up.

This seems to have been the first case of direct interferon alfa-induced pericarditis, because lupus-like syndrome and mixed type 2 cryoglobulinemia with vasculitis were deemed to be the direct cause in previously reported cases.

Patients with pre-existing cardiac disease are more likely to develop cardiovascular toxicity while receiving interferon alfa, but these complications are rare. Among 89 patients with chronic hepatitis C, 12-lead electrocardiography monthly during a 12-month treatment period and follow-up for 6-months showed only minimal and non-specific abnormalities in five patients (two had right bundle branch block, one left anterior hemiblock, and two unifocal ventricular extra beats) [ ]. None of these disorders required treatment withdrawal, and complete non-invasive cardiovascular assessment was normal. Overall, the role of interferon alfa was uncertain and the 5.6% incidence of electrocardiographic abnormalities was suggested to be similar to that expected in the general population. Nevertheless, severe cardiac dysrhythmias are still possible in isolated cases, as illustrated by the development of third-degree atrioventricular block, reversible on withdrawal, in a 57-year-old man with lower limb arteritis but no other cardiovascular disorder [ ].

Peripheral vascular complications

Raynaud ’ s phenomenon Raynaud’s phenomenon can occur, particularly in patients with chronic myelogenous leukemia [ ], and severe cases were complicated by digital necrosis [ ].

In 24 cases of Raynaud’s syndrome, interferon alfa was the causative agent in 14, interferon beta in three, and interferon gamma in five [ ]. There was no consistent delay in onset and the duration of treatment before the occurrence of symptoms ranged from 2 weeks to more than 4 years. The most severe cases were complicated by digital artery occlusion and necrosis requiring amputation. Few patients had other ischemic symptoms, such as myocardial, ophthalmic, central nervous system, or muscular manifestations. Severe Raynaud’s phenomenon was also reported in a 5-year-old girl with hepatitis C [ ].

Peripheral ischemia with cryoglobulinemia Although most patients with mild-to-moderate clinical manifestations of hepatitis C virus-associated mixed cryoglobulinemia improved during treatment with interferon alfa, acute worsening of ischemic lesions has been reported in three patients who had prominent cryoglobulinemia-related ischemic manifestations [ ]. All three had acute progression of pre-existing peripheral ischemia or leg ulcers within the first month of treatment, and transmetatarsal or right toe amputations were required in two. The lesions healed after interferon alfa withdrawal. It was therefore suggested that the anti-angiogenic activity of interferon alfa may also impair revascularization and healing of ischemic lesions in patients with initially severe ischemic manifestations.

Venous thrombosis Whereas clinically insignificant coagulation abnormalities have been documented in patients receiving high-dose continuous interferon alfa [ ], isolated cases of venous thrombosis have been observed [ ]. Interferon alfa can also induce the production of antiphospholipid antibodies [ , ]. In one study, antiphospholipid antibodies were found in five of 12 patients with melanoma treated with interferon alfa alone or with interferon alfa plus interleukin-2; deep venous thrombosis occurred in four patients with antiphospholipid antibodies [ ]. Although the underlying neoplasia undoubtedly played a role in the further development of venous thrombosis, the causative role of interferon alfa was suggested by the absence of antiphospholipid antibodies and venous thrombosis in eight patients treated with interleukin-2 alone.

Other vascular complications Other anecdotal reports included acrocyanosis and peripheral arterial occlusion [ , ]. Although the causal relation is unclear, interferon alfa was considered as a possible cause in the triggering of acute cerebrovascular hemorrhage or ischemic neurological symptoms in few patients [ , ]. The pathogenic mechanisms of these vascular effects are still unclear; vasculitis, hypercoagulability, vasospasm, a paradoxical anginal effect of interferon alfa, or an underlying cardiovascular disease have all been suggested as underlying processes.

Peripheral neuropathy

Dose-related distal paresthesia occurred in as many as 7% of patients [ ], but new onset or worsening of neuropathy has rarely been attributed to interferon alfa. A sensorimotor polyneuropathy was the most frequent presentation [ ]. The symptoms usually developed after 2–28 weeks of treatment. Such reports were mostly described in patients who received high cumulative doses of interferon alfa, but induction or exacerbation of peripheral sensorimotor axonal neuropathy, particularly in patients with chronic hepatitis C and mixed cryoglobulinemia, was also observed after long-term or low-dose treatment [ ]. Nerve biopsy showed necrotizing vasculitis or axonal degeneration. Most patients stabilized or improved slowly over several months after interferon alfa withdrawal and/or treatment with corticosteroids. Several patients also required plasmapheresis or cyclophosphamide. Although several authors have suggested an autoimmune process, the underlying pathogenic mechanism is unclear.

Other rare forms of neuropathy include mononeuropathy multiplex [ ], acute axonal polyneuropathy [ ], anterior ischemic optic neuropathy [ ], trigeminal sensory neuropathy [ ], bilateral neuralgic amyotrophy [ ], brachial plexopathy [ ], and symptoms suggestive of leukoencephalopathy [ ].

Interferon can cause multifocal leukoencephalopathy and sensorimotor polyneuropathy, in which case long-term interferon therapy should be reviewed.

• A 77-year-old man developed a progressive exercise-induced gait disturbance with right-sided predominance, non-systemic vertigo, muscle pain in both thighs after exercise, restless legs, stocking-type sensory disturbances bilaterally, and hypesthesia in the left arm [ ]. He had received intramuscular interferon alfa 2b 4.5–5 million units 3 times weekly for 18 years for hairy cell leukaemia. A brain MRI scan showed multiple non-enhancing T2-hyperintense lesions supratentorially, interpreted as vascular lesions. He was also taking pramipexole, levodopa, cabergoline, diltiazem, molsidomine, finasteride, and allopurinol. There was bilateral hypacusis, bilateral dysmetria, exaggerated biceps tendon reflexes on the right side, reduced biceps reflexes on the left, discrete proximal weakness in the left arm, proximal weakness in both legs, reduced ankle reflexes on the right, extensor plantar reflexes bilaterally, left-sided hemihypesthesia, stocking-type sensory disturbances bilaterally, and a tendency to fall during Romberg’s test and Unterberger’s treadmill test. Serum IgG antibodies against Borrelia burgdorferi were positive, but the cerebrospinal fluid was negative.

Cranial nerve palsies

There have been reports of interferon alfa-induced cranial nerve palsies, including Bell’s palsy.

Two patients, one of whom also received ribavirin, had facial nerve palsy after 5 and 8 months of interferon alfa therapy [ ]. The palsy resolved completely in one patient after withdrawal and the administration of prednisolone; however, in the other case, the palsy resolved without drug withdrawal, suggesting coincidence.

Two other cases of Bell’s palsy, which reversed after interferon alfa withdrawal, have been reported [ ]. Although the delay in onset (7.5 and 8 weeks) suggested that interferon alfa might be the cause, one patient had no recurrence after rechallenge. A coincidental adverse event cannot therefore be completely ruled out.

Bell’s palsy occurred in two patients aged 42 and 49 years with chronic hepatitis C [ ]. These patients, who also took ribavirin, developed Bell’s palsy after 2–3 months of treatment. They recovered fully after withdrawal, but prednisone was also required in one. Whether interferon alfa played a direct role is questionable, because the first patient subsequently restarted the same treatment without recurrence, and the second patient had had an attack of Herpes simplex stomatitis shortly before the cranial nerve deficit occurred.

• Three men aged 43–55 years developed Bell’s palsy after 4 days, 5 weeks, and 4 months of treatment with interferon alfa and ribavirin for chronic hepatitis C [ ]. The Bell’s palsy spontaneously resolved within 6 weeks after withdrawal and none of the patients was rechallenged.

Although a coincidental adverse reaction was possible, the authors suggested that immunotherapy may have produced a breakdown in peripheral tolerance to myelin sheath antigens.

Demyelination

Various forms of interferon alfa-induced neuropathy have been reported, but chronic inflammatory demyelinating polyneuropathy has seldom been described [ , ].

In two patients with chronic hepatitis C or malignant melanoma, paresthesia and tiredness occurred after 6 weeks and 9 months of treatment respectively. Despite withdrawal of interferon alfa, the neurological symptoms worsened initially and a diagnosis of chronic inflammatory demyelinating polyneuropathy was finally confirmed several weeks later. One patient improved after an extended course of plasma exchange and the other required immunoglobulins and prednisolone. Mild to moderate neurological abnormalities persisted at follow-up in both patients.

Multiple sclerosis has been attributed to interferon alfa [ ].

• A 29-year-old woman received interferon alfa-2b (6–10 MU/day) for chronic myeloid leukemia, and about 3 years later developed headaches, back pain, progressive visual disturbance, and a sensory deficit in the legs. MRI scans of the brain and spinal cord suggested a first episode of multiple sclerosis. In addition, the myelin basic protein concentration was slightly raised, and perimetry showed bilateral optic neuritis. Most of her neurological symptoms, except central vision impairment, improved after interferon alfa withdrawal and treatment with high-dose methylprednisolone. As a major partial cytogenetic response had been obtained with interferon, she was given natural interferon alfa (3 MU/day) 2 months later. After 2 days of treatment, she complained of transient but severe pain in the back and the legs, and developed acute paraplegia and loss of micturition desire. Again, interferon was withdrawn and she was given high-dose methylprednisolone. There was no further neurological deterioration at follow-up.

This account is reminiscent of the various autoimmune diseases that can be unmasked or exacerbated by interferon alfa.

Features resembling multiple sclerosis have been reported in two women aged 29 and 62 years, treated with recombinant interferon alfa for chronic hepatitis C or chronic myelogenous leukemia [ ]. Both developed bilateral optic neuritis. Other symptoms included increased deep tendon reflexes and reduced sensation of vibration in one patient, and numbness of the legs and bowel and bladder dysfunction in the other. MRI findings and cerebrospinal fluid examination simulated multiple sclerosis. Similar manifestations were again noted in one patient after readministration of natural interferon alfa, and both partially recovered visual acuity after glucocorticoid treatment.

Other demyelinating events attributed to interferon alfa have included a report of Guillain–Barré syndrome 2 months after the end of a course of interferon alfa-2b and suramin sodium for metastatic pulmonary cancer in a 37-year-old woman [ ]. Although the concomitant use of suramin made the association of interferon alfa with the Guillain–Barré syndrome unclear, the authors suggested that the simultaneous occurrence of autoimmune liver disease and hematological changes in this patient argued for a causal role of interferon alfa, and they also suggested that this combination may have had a synergistic effect.

Movement disorders

Extrapyramidal reactions have been occasionally reported as a manifestation of persistent neurotoxicity induced by interferon alfa. This issue has been addressed in a report of severe refractory akathisia [ ].

• A 28-year-old man received interferon alfa (5 MU/day for 28 days) for chronic hepatitis B. At the end of treatment he developed a slight parkinsonian gait, and 8 days later had a fever with vomiting, insomnia, restlessness, and raised serum creatine kinase activity (4946 IU/l). He had severe akathisia with psychomotor excitement and parkinsonism. Despite treatment with clonazepam, thioridazine, propranolol, trihexyphenidyl, and bromocriptine, his condition progressively worsened. He was finally given intravenous levodopa for 8 days and recovered dramatically within the next few days.

This report, together with previous experimental data, suggests that levodopa might be useful in alleviating some of the manifestations of persistent interferon alfa-induced neurotoxicity.

In two other patients, akathisia occurred shortly after they had started to take interferon alfa; one improved after the frequency of injections was reduced [ ]. Unfortunately, this report did not provide sufficient convincing evidence for a causal relation; the development of akathisia may have been coincidental.

Chorea is a very rare manifestation of interferon alfa neurotoxicity [ ].

A 68-year-old woman developed progressive personality changes and 2 months later permanent choreic movements of the four limbs [ ]. She had taken interferon alfa-2b (3 MU/day) and hydroxyurea 50 mg/kg/day for chronic myeloid leukemia for 2 years and had no history of psychiatric disorders. Neuropsychological testing showed frontal subcortical dysfunction. There were no abnormalities in the Huntington disease gene. She progressively worsened over the next 6 months. The electroencephalogram was disorganized, with diffuse slow waves, and she was bedridden. Interferon alfa was withdrawn. The chorea ceased 1 month later and she completely recovered cognitive function. Electroencephalography was normal 6 and 12 months later.

The authors attributed these events to antidopaminergic effects of interferon alfa.

Severe and complex diffuse spinal myoclonus developed after 6 weeks of interferon alfa-2a (27 MU/week) in a 63-year-old man with metastatic renal cancer, but his symptoms were only partially improved after interferon withdrawal and anticonvulsant treatment [ ]. The exact role of interferon was therefore only speculative.

Involuntary facial movements have been more convincingly described [ ].

• A 49-year-old woman progressively developed involuntary facial movements after taking interferon alfa-2a (6 MU/day) for 1.5 years for chronic myeloid leukemia. She had bilateral, involuntary, rhythmic, repetitive contractions of her frontalis, nasalis, orbicularis oris, and lower facial muscles. The movements worsened with anxiety and lack of sleep and during talking. Electromyography showed rhythmic synchronous muscle contractions compatible with myorhythmia. Neurological examination was otherwise normal and no other causes were found. The facial movements completely resolved 1 month after interferon withdrawal, and recurred a few weeks after interferon rechallenge.

Interferon alfa-induced acute dystonia has been reported [ ].

A 24-year-old man with chronic hepatitis B who had tolerated interferon alfa for 6 months 9 years before, developed acute involuntary movements with fever and chills 2 hours after his first dose of peginterferon alfa. Electroencephalography and an MRI scan were normal. His symptoms resolved quickly after diazepam injection and he was given peginterferon again uneventfully.

In the context of an acute dystonia, the authors claimed that the development of tolerance to the second injection of interferon alfa further argued for a causal role of the drug.

Restless legs syndrome occurred 2 months after a 60-year-old man started to take interferon alfa and ribavirin for chronic hepatitis C [ ]. Evidence for a causal relation was obtained by complete resolution of the syndrome after withdrawal, careful elimination of other causes, and rapid recurrence of the symptoms on rechallenge with peginterferon alfa and ribavirin. Although both interferon alfa and ribavirin may have been the cause of the syndrome, the authors correctly argued that only interferon alfa is thought to affect dopaminergic neurotransmission, which is strongly suspected to be involved in the pathophysiology of restless legs syndrome.

Seizures

Interferon alfa can cause seizures in patients with no history of epilepsy, and one report of three patients suggested that seizures can occur at any time during treatment [ ]. Seizures were also reported in two of five children after intrathecal interferon alfa for subacute sclerosing panencephalitis, suggesting a much higher incidence in this condition [ ].

Although generalized tonic-clonic seizures have occasionally been described during trials of high doses of interferon alfa, they have also been reported after the use of intermediate or even low doses [ ]. There was a 1.3% incidence of generalized seizures in a retrospective study of 311 patients treated with low doses for chronic viral hepatitis [ ]. In another study, tonic-clonic seizures were identified in 4% of children treated for chronic hepatitis B [ ]. As seizures occurred only in children under 5 years of age with fever or potential perinatal nervous system injury, immaturity of the nervous system was suggested to be an additional factor for interferon alfa-induced neurotoxicity in children.

In three patients, seizures occurred after a cumulative dose of 266–900 MU [ , ]. Two were re-treated with a lower dose and remained free of seizures [ ]. Another patient with a history of bipolar mood disorder had his first four episodes of seizures with a prolonged delirious state 1 week after withdrawal of interferon alfa [ ].

Reversible photosensitive seizures have also been reported [ ].

• A 62-year-old man without a personal or family history of epilepsy received interferon alfa (3 MU three times a week) for 2 years for multiple myeloma. He had frequent episodes of myoclonic jerks in the face, especially when the sun was shining while driving his car. He also had one generalized seizure. Electroencephalography showed a paroxysmal response to intermittent photic stimulation and magnetic resonance imaging was normal. The seizures disappeared and his electroencephalogram normalized after interferon alfa withdrawal.

In this case, the possible role of interferon alfa was suspected only late during treatment, indicating that patients should be regularly questioned about neurological symptoms, because more severe complications might have occurred.

In a prospective study of the effect of interferon alfa (56 MU/day for 4 weeks then 27 MU/week for 20 weeks) in 56 patients with chronic hepatitis C, there was diffuse electroencephalographic slowing at 2 and 4 weeks of treatment, suggesting mild encephalopathy [ ]. These changes completely reversed after withdrawal. However, the dosage used in this Japanese study was relatively high compared with the dosages currently used in Western countries. In addition, the clinical relevance of these electroencephalographic changes was not investigated.

Vision

Ophthalmic disorders can occur during interferon alfa treatment [ ] and some of the literature has been reviewed [ ]. Retinopathy consisting of cotton-wool spots and/or superficial retinal hemorrhages has been reported with a variable incidence (18–86%), and the available data suggest that the increased incidence can be influenced by an initial high dose of interferon alfa. Whereas diabetes mellitus and systemic hypertension have been identified as possible susceptibility factors, the incidence of retinopathy was not significantly increased in 19 patients with chronic renal insufficiency, compared with 17 patients without chronic renal insufficiency [ ]. However, it was felt that renal insufficiency may be associated with the most severe cases, that is those requiring dosage reductions.

In one study, in which prospective ophthalmic examinations were made before and at regular 2-week intervals after the beginning and end of treatment, 28 of 81 patients who received a uniform total dose of natural interferon alfa (478 MU) for chronic hepatitis C developed the typical findings of interferon-induced asymptomatic retinopathy (cotton-wool spots and/or retinal hemorrhages) [ ]. In contrast, there were no lesions in the 25 patients with chronic hepatitis C who did not receive interferon alfa or in the 20 with diabetes mellitus and/or hypertension but without chronic hepatitis C. Most of the cases were observed within 4 months of treatment, and the lesions always abated after withdrawal or even despite continuation, suggesting that treatment can be continued unless patients develop symptoms. Indeed, most patients with retinopathy associated with interferon alfa remained asymptomatic.

Of 42 patients taking interferon alfa-2b and ribavirin for chronic hepatitis C and regularly followed for ocular complications, 27 developed a retinopathy with cotton-wool spots (n = 27), retinal hemorrhages (n = 6), subconjunctival hemorrhage (n = 2), and asymptomatic optic nerve edema (n = 1) [ ]. These lesions were usually transient, but some had recurrent episodes of retinopathy while continuing therapy. Although the retinopathy was considered benign in most cases, two patients complained of ocular symptoms consisting of blurring of vision in one and permanent peripheral monocular scotoma in the other. Three patients had to discontinue treatment because of more severe posterior segment complications. In contrast to several studies, the incidence of retinopathy was not dose related and hypertension was not identified as a significant risk factor.

• A 40-year-old man with multiple sclerosis taking interferon beta-1a developed cotton wool spots in the optic fundi, which seem to have been related to interferon [ ].

• A 37-year-old man developed bilateral ischemic retinopathy and blurred vision and recovered promptly after peginterferon alfa-2b withdrawal [ ].

• A 44-year-old woman developed visual difficulties in dim light, with abnormal electro-oculography; she slowly improved after peginterferon alfa withdrawal [ ]. She had previously tolerated standard interferon alfa for 6 months.

• A 46-year-old man developed visual complaints with predominant blurry vision and the classical cotton-wool spots of interferon-induced retinopathy after treatment with peginterferon alfa-2b; he recovered after 5 months despite continued treatment [ ].

The last report also provided a complete detailed review of 27 reports of interferon alfa-related ocular abnormalities.

The pathogenesis of retinopathy associated with interferon alfa is unclear. In 45 patients with chronic hepatitis C (25 treated with interferon) there was an association between retinal hemorrhages caused by interferon alfa (six patients) and a concomitant significant increase in plasma-activated complement (C5a), compared with baseline C5a serum concentrations [ ]. However, the signification and contribution of raised C5a concentrations in the pathogenesis of ocular complications needs to be clarified, although it has been suggested that retinal hemorrhage could be predicted on the basis of raised C5a concentrations [ ].

Although most patients with interferon alfa retinopathy remain asymptomatic, ocular complications, such as reduced vision or complete visual loss due to occlusive vasculitis, central retinal artery occlusion, or anterior ischemic optic neuropathy, continue to be reported in a very few patients [ , , ]. However, subclinical but eventually long-lasting or even irreversible abnormally long visual evoked responses have been identified in 24% of patients [ ]. Regular ophthalmological monitoring to detect retinal changes, even though the patient is still asymptomatic, is therefore strongly recommended in patients receiving interferon alfa.

Owing to the possible risk of severe retinopathy, regular ophthalmic monitoring before treatment and at regular intervals during treatment is advocated by most authors. That interferon alfa-induced retinopathy is not always self-limiting and benign has been supported by the development of severe proliferative retinopathy in a 69-year-old woman with chronic myeloid leukemia who required extensive laser therapy and unilateral vitrectomy [ ]. Other severe posterior ocular complications recently described in patients with chronic hepatitis C treated with interferon alfa-2b have included:

• the development of an epiretinal membrane in a 47-year-old man [ ];

• permanent visual loss attributed to combined retinal artery and central retinal vein obstruction in a 51-year-old woman [ ];

• bilateral simultaneous anterior ischemic optic neuropathy in two patients [ ].

Whether retinopathy is more frequent with peginterferon than with standard interferon is still unknown. In one prospective series of 23 patients co-infected with HIV and hepatitis C, who were treated with peginterferon alfa-2b and ribavirin, eight developed classical interferon retinopathy and two had impaired color vision, of whom one was sufficiently severely affected to require drug withdrawal [ ]. Although the incidence of retinopathy in this small study was within the expected range, the occurrence of two symptomatic cases is noteworthy.

Vogt–Koyanagi–Harada disease, an autoimmune syndrome with potentially severe ocular complications, has been reported [ ].

• A 52-year-old woman developed visual blurring, eye pain, tinnitus, and alopecia after taking interferon alfa-2b 9 MU/week and ribavirin 1200 mg/day for 9 months. Further examination showed poliosis of the left eyelashes and reduced bilateral auditory acuity. Ophthalmic examination showed severe bilateral panuveitis with posterior synechiae, vitritis, and choroidal and scleral thickening. The antiviral treatment was withdrawn and she received local and systemic glucocorticoids.

The authors suggested that Vogt–Koyanagi–Harada disease, the pathophysiology of which involves a T cell-mediated autoimmune response to melanocytes, may have been exacerbated by interferon alfa in a genetically predisposed patient. Vogt–Koyanagi–Harada disease has also been observed after 4 months of interferon alfa-2b treatment in a 36-year-old man with chronic hepatitis C [ ]. Withdrawal and methylprednisolone therapy was followed by slow but complete recovery.

Other isolated reports of ophthalmic abnormalities refer to optic neuritis with blurring of vision, cortical blindness with fatal encephalopathy, mononeural abducent nerve paralysis, and complete but reversible bilateral oculomotor nerve paralysis [ ].

• A 60-year-old smoker was treated with interferon alfa (100 MU/week for 2 months and 9 MU/week for 15 weeks) for cutaneous melanoma. Ocular examination was normal before treatment, but he developed acute loss of peripheral vision in his left eye after 23 weeks. Examination was consistent with anterior ischemic optic neuropathy, and there was optic disc edema, a pupillary defect, and circular visual field constriction in the left eye. There was renal artery constriction in both eyes. Despite treatment with aspirin, high-dose dexamethasone, heparin, and finally withdrawal of interferon alfa, loss of visual function progressed and affected both eyes. Ciclosporin was started, but he was considered to have irreversible loss of visual function.

This report shows that interferon alfa can be a potent precipitator of extremely severe ocular disorders and also argues for careful ocular surveillance in patients receiving adjuvant interferon alfa for high-risk resected melanoma.

Of 57 patients treated with interferon for renal cell carcinoma, two developed multiple retinal exudates associated with visual disturbance; both had taken vinblastine concurrently. The precise role of interferon in this reaction is unknown [ ].

Auditory function

In 49 patients, there was reversible otological impairment with tinnitus, mild-to-moderate hearing loss or both in respectively 8, 16, and 20% of patients after interferon alfa or interferon beta administration [ ]. These disorders tended to occur more frequently in patients on high cumulative doses, but led to withdrawal of treatment in only two patients. Complete but reversible hearing loss, and acute unilateral vestibular dysfunction with spontaneous vertigo and nystagmus have each been reported in one patient receiving interferon alfa [ ].

Of 27 patients treated with interferon alfa 30 MU/week for chronic hepatitis B and prospectively assessed for auditory function, hearing loss was found in nine after 7 days of treatment, with an increase in the degree of hearing loss until day 21 [ ]. However, auditory disability did not exceed 20 dB for any frequency despite continued treatment. Complete recovery was observed 1 month after withdrawal.

Sudden hearing loss has been reported [ ] and in one case of promptly reversible hearing loss on interferon alfa withdrawal, the presence of anti-endothelial cell antibodies was suggested to have played a role in the development of autoimmune microvascular damage [ ].

Sudden hearing loss and tinnitus occurred in six patients treated with peginterferon alfa (mostly alfa-2b); the onset of symptoms after the start of therapy was 1 day in one patient and 4–40 weeks in the other five [ ]. The authors estimated that about 1% of patients would experience this event. The ototoxicity was attributed to reversible biochemical and metabolic changes in the cochlea, but could also have been caused by an autoimmune mechanism with antiendothelial cell antibodies, which were found in nearly half of the patients. Autoimmune hearing loss occurs later in the course of treatment than biochemical and metabolic hearing loss, is mostly bilateral, and has a progressive course over the ensuing days and months. In contrast to direct ototoxicity caused by standard interferon, that caused by peginterferon does not recover fully after withdrawal. However, once symptoms appear they do not progress, and so the decision to continue or abandon treatment has to be individualized.

Taste

In a 40-year-old man taking interferon alfa for chronic hepatitis C ageusia and hyposmia resulted in loss of appetite and weight loss, but reversed on withdrawal [ ].

Olfaction

Patients treated with interferon alfa sometimes complain of transient taste or smell alterations, but anosmia has been reported [ ].

• A 37-year-old man received interferon alfa for chronic hepatitis C. After 2 weeks he complained of smelling difficulties and subsequently developed complete anosmia. There were no other neurological symptoms and complete neurological examination was normal. Anosmia still persisted 13 months after drug withdrawal.

The persistence of anosmia late after interferon alfa was resumed indicates that a causal relation to treatment is purely speculative.

Presentation

Within a large spectrum of symptoms, complications are classified as acute, subacute, or chronic.

Acute neuropsychological disturbances are usually associated with the flu-like syndrome and include headache, fatigue, and weakness, drowsiness, somnolence, subtle impairment of memory or concentration, and lack of initiative [ ]. This pattern of cognitive impairment is similar to changes observed during influenza and has also been described in healthy patients who have received a single dose of interferon alfa [ ]. More severe acute manifestations (for example, marked somnolence or lethargy, frank encephalopathy with visual hallucinations, dementia or delirium, and sometimes coma) have been almost exclusively described in patients receiving more than 20–50 MU [ ]; vertigo, cramps, apraxia, tremor and dizziness were also reported.

The subacute or chronic neuropsychiatric effects of long-term therapy are usually non-specific, with cognitive impairment (for example visuospatial disorientation, attentional deficits, memory disturbances, slurred speech, difficulties in reading and writing), changes in emotion, mood, and behavior (for example psychomotor slowing, hypersomnia, loss of interest, affective disorders, irritability, agitation, delirium, paranoia, aggressiveness, and murderous impulses). Post-traumatic stress symptoms have also been reported [ ]. As a result, severe psychic distress can be observed during long-lasting treatment or in patients who are otherwise not severely affected [ , ]. The most severe psychiatric complications of interferon alfa include rare cases of homicidal ideation, suicidal ideation, and attempted suicide [ ].

The clinical features of mania have been described in four patients with malignant melanoma, with a detailed review of nine other published cases [ ]. Although seven suffered from depression during treatment, the onset of mania or hypomania was often associated with interferon alfa dosage fluctuation (withdrawal or dose reduction) or introduction of an antidepressant for interferon alfa-induced depression. In these patients, the risk of mood fluctuations persisted for several months after interferon alfa withdrawal, and low-dose gabapentin was considered useful in treating manic disorders and in preventing mood fluctuations. Interferon alfa was suggested as a possible cause of persistent manic-depressive illness for more than 4 years in a 40-year-old man [ ]. Although the manic episodes may have been coincidental, the negative history and the age of onset are in keeping with a possible role of interferon alfa treatment.

The clinical features, management, and prognosis of psychiatric symptoms in patients with chronic hepatitis C have been reviewed using data from 943 patients treated with interferon alfa (85%) or interferon beta (15%) for 24 weeks [ ]. Interferon-induced psychiatric symptoms were identified in 40 patients (4.2%) of those referred for psychiatric examination. They were classified in three groups according to the clinical profile: 13 cases of generalized anxiety disorder (group A), 21 cases of mood disorders with depressive features (group B), and six cases of other psychiatric disorders, including psychotic disorders with delusions/hallucinations (n = 4), mood disorders with manic features (n = 1), and delirium (n = 1) (group C). The time to onset of the symptoms differed significantly between the three groups: 2 weeks in group A, 5 weeks in group B, and 11 weeks in group C. Women were more often affected than men. There was no difference in the incidence or nature of the disorder according to the type of interferon used. Whereas most patients who required psychotropic drugs were able to complete treatment, 10 had to discontinue interferon treatment because of severe psychiatric symptoms, 5 from group B and five from group C. Twelve patients still required psychiatric treatment for more than 6 months after interferon withdrawal. In addition, residual symptoms (anxiety, insomnia, and mild hypothymia) were still present at the end of the survey in seven patients. Delayed recovery was mostly observed in patients in group C and in patients treated with interferon beta. Although several patients with a previous history of psychiatric disorders are sometimes successfully treated with interferon alfa, severe decompensation with persistent psychosis should be regarded as a major possible complication [ ].

The neuropsychological adverse effects of long-term treatment have been assessed in 14 patients with myeloproliferative disorders using a battery of psychometric and electroneurological tests before and after 3, 6, 9, and 12 months of treatment (median dose 25 MU/week) [ ]. In contrast to several previous studies, there was no significant impairment of neurological function, and attention and short-term memory improved during treatment. Despite the small number of patients, these results suggest that prolonged interferon alfa treatment did not cause severe cognitive dysfunction, at least in patients with cancer.

The well-known psychiatric adverse reactions to pegylated interferons have again been reported. Of 56 patients with hemophilia infected with hepatitis C who were treated with peginterferon alfa-2b 1.5 micrograms/kg/week and ribavirin 800–1200 mg/day, 22% developed depression that warranted the use of antidepressants; one developed a psychosis; 11% discontinued therapy owing to adverse reactions [ ].

Of 375 patients with renal cell carcinoma who were treated with subcutaneous interferon alfa 9 MU three times a week, 12% developed grade 3 or 4 fatigue [ ].

Diagnosis

Electroencephalographic (EEG) findings show reversible cerebral changes with slowing of dominant alpha wave activity, and occasional appearance of one and two activity in the frontal lobes, suggesting a direct effect on fronto-subcortical functions. Marked electroencephalographic abnormalities are sometimes observed in asymptomatic patients. The pattern of changes is identical whatever the dose, but the severity of symptoms is dose- and schedule-related. Most patients improve or recover after dosage reduction or withdrawal, and protracted toxicity, with impaired memory, deficits in motor coordination, persistent frontal lobe executive functions, Parkinson-like tremor, and mild dementia, have been occasionally reported [ ].

In a study of 67 patients with chronic viral hepatitis, the self-administered Minnesota Multiphasic Personality Inventory (MMPI), which determines the patient’s psychological profile, significantly correlated with the clinical evaluation and was a sensitive and reliable tool for identifying patients at risk of depressive symptoms before the start of interferon alfa therapy [ ]. It was also successfully used to monitor patients during treatment.

Frequency

The most typical psychiatric symptoms reported by patients taking interferon alfa are depressive symptoms, at rates of 10–40% in most studies [ ]. In four clinical studies in a total of 210 patients with chronic hepatitis C, the rate of major depressive disorders during interferon alfa treatment was 23–41% [ ].

Suicidal ideation or suicidal attempts have been reported in 1.3–1.4% of patients during interferon alfa treatment for chronic viral hepatitis or even within the 6 months after withdrawal [ , ], but the excess risk related to interferon alfa is not known.

Time-course

Subacute or chronic neuropsychiatric manifestations are more typically identified after several weeks of treatment and are among the most frequent treatment-limiting adverse reactions [ , , ]. The onset can be insidious in patients treated with low doses, or subacute in those who receive high doses. Most patients develop severe depressive symptoms within the first 3 months of treatment [ , , , ].

Although psychiatric manifestations usually appear during interferon alfa therapy, delayed reactions can occur.

• A 37-year-old man without a previous psychiatric history developed major depression with severe psychotic features within days after the discontinuation of a 1-year course of interferon alfa-2b [ ].

In 10 patients with melanoma and no previous psychiatric disorders, depression scores measured on the Montgomery-Asberg Depression Rating Scale were significantly increased after 4 weeks of high-dose interferon alfa [ ]. Patients whose scores were higher before treatment developed the worst symptoms of depression during treatment. This positive correlation provides striking evidence that baseline and regular assessment of mood and cognitive functions are necessary to detect disorders as early as possible.

Mechanisms

Although very few studies have specifically investigated the role of the underlying disease, the findings of significant neuropsychiatric deterioration during interferon alfa treatment compared with placebo or no treatment in chronic hepatitis C, chronic myelogenous leukemia, or amyotrophic lateral disease strongly suggested a causal role of interferon alfa [ , , ].

The mechanism by which the systemic administration of interferon alfa produces neurotoxicity is unclear, and might result from a complex of direct and indirect effects involving the brain vasculature, neuroendocrine system, neurotransmitters and the secondary cytokine cascade with cytokines which exert effects on the nervous system, for example interleukin-1, interleukin-2, or tumor necrosis factor alfa [ ]. Whether a clinical effect is directly mediated through the action of a given cytokine or results from a secondary pathway through the induction of other cytokines or second messengers is difficult to determine.

A study in 18 patients treated with interferon alfa for chronic hepatitis C has given insights into the possible pathophysiological mechanism of depression [ ]. Depression rating scales, plasma tryptophan concentrations, and serum kynurenine and serotonin concentrations were measured at baseline and after 2, 4, 16, and 24 weeks of treatment with interferon alfa 3–6 MU 3–6 times weekly. During treatment, tryptophan and serotonin concentrations fell significantly, while kynurenine concentrations rose significantly. Depression rating scales also rose from baseline after the first month of treatment, with continued increases thereafter. In addition, there was a relation between increased scores of depression and changes in serum kynurenine and serotonin concentrations. These changes suggested a predominant role of the serotonergic system in the pathophysiological mechanisms of interferon alfa-associated depression. Accordingly, 35 of 42 patients included in three open trials of antidepressant treatment responded to a selective serotonin reuptake inhibitor drug, such as citalopram or paroxetine, and were able to complete interferon treatment [ , , ].

Susceptibility factors

Various possible susceptibility factors have been analysed in several studies [ , ]. Sex, the dose or type of interferon alfa (natural or recombinant), a prior personal history of psychiatric disease, substance abuse, the extent of education, the duration and severity of the underlying chronic hepatitis, and scores of depression before interferon alfa treatment were not significantly different between patients with and without interferon alfa-induced depression. Advanced age was suggested to be a susceptibility factor in only one study [ ]. Although a worsening of psychiatric symptoms was noted during treatment in 11 patients receiving psychiatric treatment before starting interferon, only one was unable to complete the expected 6-month course of interferon alfa and ribavirin therapy [ ].

Of 91 patients treated with interferon alfa-2b and low-dose cytarabine for chronic myelogenous leukemia, 22 developed severe neuropsychiatric toxicity [ ]. Their symptoms consisted mostly of severe depression or psychotic behavior, which resolved on withdrawal in all patients. The time to toxicity ranged from as early as 2 weeks to as long as 184 weeks after the start of treatment. Five of six patients had recurrent or worse symptoms after re-administration of both drugs. Several baseline factors were analysed, but only a pretreatment history of neurological or psychiatric disorders was considered to be a reliable susceptibility factor. Severe neuropsychiatric toxicity developed during treatment in 63% of patients with previous neuropsychiatric disorders compared with 10% in patients without. It is unlikely that the combination of interferon alfa-2b with low-dose cytarabine potentiated the neuropsychiatric adverse reactions to interferon alfa in this study. Indeed, previous experience with this combination, but after exclusion of patients with a psychiatric history, was not associated with such a high incidence of neuropsychiatric toxicity or any significant difference in toxicity between interferon alfa alone and interferon alfa plus low-dose cytarabine.

Patients receiving high doses of interferon alfa or long-term treatment are more likely to develop pronounced symptoms [ ]. A previous history of psychiatric disorders, organic brain injury, or addictive behavior are among potential susceptibility factors, but worsening of an underlying psychiatric disease is not the rule, provided that strict psychiatric surveillance and continuation of psychotropic drugs are maintained [ ]. Other putative susceptibility factors include the intraventricular administration of interferon alfa, previous or concomitant cranial irradiation, asymptomatic brain metastases, and pre-existing intracerebral ateriosclerosis [ , ]. Despite early findings, co-infection with HIV has not been confirmed to be a susceptibility factor [ ].

The occurrence of psychiatric disorders has been prospectively investigated in 63 patients who received a 6-month course of interferon alfa (9 MU/week) for hepatitis C [ ]. All were assessed at baseline with the Structured Clinical Interview for DSM-III-R (SCID) and monitored monthly with the Hopkins Symptoms Checklist (SCL-90). Most had a history of alcohol or polysubstance dependence, and 12 had a lifetime diagnosis of major depression. There were no significant changes in the SCL-90 scores during the 6-month period of survey in the 49 patients who completed the study, even in those who had a lifetime history of major depression. At 6 months, there was probable minor depression in eight patients and major depression in one; none had attempted suicide.

In a prospective study, 50 patients with chronic hepatitis B or C who received 18–30 MU/week of natural or recombinant interferon alfa were followed for 12 months [ ]. The SCID before starting interferon alfa identified 16 patients with a current psychiatric diagnosis and eight with a previous psychiatric disorder; 26 patients free of any psychiatric history constituted the control group. Psychiatric manifestations during treatment occurred in 11 patients (five from the control group), major depression in five, depressive disorders in three, severe dysphoria in two, and generalized anxiety disorder in one. Most of them were successfully treated with psychological support and drug therapy. Overall, 20 patients interrupted interferon alfa (10 in each group), including three for psychiatric adverse reactions, but patients with a pre-existing or recent psychiatric diagnosis were no more likely to withdraw from treatment than the controls.

Of 33 patients with chronic hepatitis C treated with interferon alfa, 9 MU/week for 3–12 months, prospectively evaluated using the Montgomery-Asberg Depression Rating Scale (MADRS) before and after 12 weeks of treatment, eight developed depressive symptoms, of whom four had major depression without a previous psychiatric history [ ]. All four recovered after treatment with antidepressants. This study confirmed that a high baseline MADRS is significantly associated with the occurrence of depressive symptoms.

These studies have confirmed that previous psychiatric disorders are not necessarily a contraindication to a potentially effective treatment. However, patients with depressive symptoms immediately before treatment are still regarded at risk of severe psychiatric deterioration with treatment [ ].

Management

The management of the psychiatric complications of interferon alfa has not been carefully investigated, but multiple approaches are theoretically possible. Various pharmacological and non-pharmacological interventions have been discussed [ ], and prompt intervention should be carefully considered in every patient who develops significant neuropsychiatric adverse reactions while receiving interferon alfa. Depending on the clinical manifestations, proposed treatment options include antidepressants, psychostimulants, or antipsychotic drugs.

Based on a possible reduction in central dopaminergic activity mediated by the binding of interferon alfa to opioid receptors, naltrexone has been proposed as a means of improving cognitive dysfunction [ ].

Selective serotonin re-uptake inhibitors have been advocated as the drugs of choice to allow completion of interferon alfa treatment [ ], but that was based on very limited experience and the unproven assumption that SSRIs are safe in patients with underlying liver disease. The preliminary results of a double-blind, placebo-controlled study showed that 2 weeks of pretreatment with paroxetine significantly reduced the occurrence of major depression in 16 patients on high-dose interferon alfa for malignant melanoma [ ]. In a placebo-controlled trial, the preventive effects of paroxetine (mean maximal dose of 31 mg) were studied in 40 patients with high-risk malignant melanoma and interferon alfa-induced depression [ ]. Treatment started 2 weeks before adjuvant high-dose interferon alfa. Paroxetine significantly reduced the incidence of major depression (45% in the placebo group and 11% in the paroxetine group) and the rate of interferon alfa withdrawal (35 versus 5%). Although the number of patients was small and the duration of the survey short (12 weeks), this suggests that paroxetine effectively prevents the risk of depressive disorders in patients eligible for high-dose interferon alfa. However, these results are limited, because patients with melanoma who receive adjuvant high-dose interferon alfa are particularly likely to develop depression. The safety of prophylaxis with paroxetine also requires additional data, because three patients taking paroxetine developed retinal hemorrhages, including one with irreversible loss of vision.

In contrast to this study, a 31-year-old woman with major depressive disorder, which responded to paroxetine and trazodone, had progressive recurrence of mood disorders after the introduction of interferon alfa for essential thrombocythemia [ ].

This suggests that interferon alfa can also reverse the response to antidepressants.

Pituitary

Interferon alfa can stimulate the hypothalamic–pituitary–adrenal axis, with a marked increase in cortisol and adrenocorticotrophic hormone secretion after acute administration [ ]. No further stimulation was observed after several weeks of treatment, pointing to possible down-regulation of the ACTH secretory system. As a result, long-term treatment with interferon alfa is not thought to influence pituitary hormones significantly, and the concentration of several hormones, for example calcitonin, LH, FSH, prolactin, growth hormone, ACTH, cortisol, testosterone, and estradiol, were not modified by prolonged interferon alfa treatment [ , ]. No clinical endocrinopathies attributable to such disorders in the regulation of these hormones have yet been reported.

Although the rate of growth was significantly lower than predicted in 35% of children receiving long-term treatment for recurrent respiratory papillomatosis [ ], only one case of growth retardation has been reported in other settings [ ]. A significant reduction in weight and nutritional status was observed during treatment with interferon alfa for 6 months for chronic viral hepatitis in children aged 4–16 years, but this was transient and not associated with growth impairment [ ].

Hypopituitarism has rarely been reported in patients with chronic hepatitis C.

A 44-year-old woman developed severe weakness after 2 months of interferon alfa 9 MU/week [ ]. Interferon was withdrawn and subsequent treatment with 4.5 MU/week produced similar symptoms within 3 months. Serum cortisol concentrations were dramatically reduced and plasma ACTH was undetectable. Other hormonal concentrations were within the reference ranges and adrenal cortex antibodies were negative. Biological anomalies normalized after interferon alfa withdrawal and 6-month substitutive hydrocortisone treatment.

• A 39-year-old man received interferon alfa-2b 9 MU/week and ribavirin 400 mg/day for 1 year, and amantadine 200 mg/day for 9 months [ ]. He had major weight gain (23 kg), reduced libido, and neuropsychiatric disturbances during treatment, and there was only partially improvement 1 year after the completion of treatment. There was testosterone, gonadotrophin, and growth hormone deficiency, but antipituitary antibodies were not detected. Although his symptoms markedly improved with recombinant human growth hormone, permanent pituitary impairment was suspected.

The authors suggested that in the second case pituitary dysfunction might have caused persistent symptoms after interferon alfa treatment. Although involvement of the immune system was suggested in both cases, autoantibodies were not detected.

Reversible hypopituitarism with antibodies to pituitary GH3 cells and exacerbation of Sheehan’s syndrome have been reported [ , ].

A syndrome resembling inappropriate antidiuretic hormone secretion has been described in a few patients receiving high-dose interferon alfa [ ].

Thyroid

Since the original 1988 report of hypothyroidism in patients with breast cancer receiving leukocyte-derived interferon alfa [ ], numerous investigators have provided clear clinical and biological data on thyroid disorders induced by different forms of interferon in patients with various diseases [ , ]. Two of these reports also mentioned associated adverse reactions that developed concomitantly, namely myelosuppression and severe proximal myopathy (Hoffmann’s syndrome).

Among 100 patients with chronic hepatitis C who were treated with subcutaneous interferon alfa-2b (3 million units three times a week) and oral ribavirin (1000–1200 mg/day, overt thyroid disease (hyper- and hypothyroidism) developed in 13% and subclinical thyroid disease in 5% [ ].

Presentation and outcomes The spectrum of interferon alfa-induced thyroid disorders ranges from asymptomatic appearance or increase in antithyroid autoantibody titers to moderate or severe clinical features of hypothyroidism, hyperthyroidism, and acute biphasic thyroiditis. Antithyroid hormone antibodies have also been found in one patient, and this could have been the cause of erroneously raised thyroid hormone concentrations [ ].

The clinical, biochemical, and thyroid imaging characteristics of thyrotoxicosis resulting from interferon alfa treatment have been retrospectively analysed from data on 10 of 321 patients with chronic hepatitis (75 with chronic hepatitis B and 246 with chronic hepatitis C) who developed biochemical thyrotoxicosis [ ]. Seven patients had symptomatic disorders, but none had ocular symptoms or a palpable goiter. Six had features of Graves’ disease that required interferon alfa withdrawal in four and prolonged treatment with antithyroid drugs in all six. Three presented with transient thyrotoxicosis that subsequently progressed to hypothyroidism and required interferon withdrawal in one and thyroxine treatment in all three.

Although much work on thyroid autoimmunity associated with interferon alfa has accumulated, little is known about the very long-term outcome of this disorder. In 114 patients with chronic hepatitis C and no previous thyroid disease who were treated with interferon alfa-2a for 12 months, data on thyroid status were retrospectively obtained at the end of treatment, 6 months after withdrawal, and after a median of 6.2 years [ ]. Among 36 patients who had thyroid autoantibodies at the end of treatment, the authors identified three groups according to the long-term outcome: 16 had persistent thyroiditis, 10 had remitting/relapsing thyroiditis (that is antibodies became negative after 6 months of therapy and were again positive thereafter), and 10 had transient thyroiditis. Therefore, 72% of these patients had chronic thyroid autoimmunity at the end of follow-up and 12 developed subclinical hypothyroidism. In contrast, only one of 78 patients negative for thyroid autoantibodies developed thyroid autoantibodies. Although none of the patients had clinical thyroid dysfunction, this study suggests that long-term surveillance of thyroid disorders is useful in patients who have high autoantibody titers at the end of treatment with interferon alfa.

Although thyroid disorders in patients treated with interferon alfa generally follows a benign course after interferon alfa withdrawal or specific treatment, severe long-lasting ophthalmopathy resulting from Graves’ disease has been described in a 49-year-old woman [ ].

Time - course Clinical symptoms usually occur after 2–6 months of treatment and occasionally after interferon alfa withdrawal.

• A middle-aged woman developed subacute thyroiditis by the sixth month of treatment with interferon alfa [ ]. She also had the classic symptoms of hyperthyroidism, although it is clear that these could easily have been mistaken for adverse reactions to interferon alfa itself, for example weakness, weight loss, and palpitation.

After 6 months of treatment, 12% of patients with chronic hepatitis C had thyroid disorders, compared with 3% of patients with chronic hepatitis B. This study also suggested a possible relation between low free triiodothyronine serum concentrations before treatment and the subsequent occurrence of thyroid dysfunction. After a follow-up of 6 months after the end of interferon alfa treatment, 60% of affected patients with chronic hepatitis C still had persistent thyroid dysfunction; all had been positive for thyroid peroxidase antibodies before treatment. Long-term surveillance is therefore needed in these patients.

Frequency In a prospective study, the overall incidence of biochemical thyroid disorders was 12% in 254 patients with chronic hepatitis C randomized to receive ribavirin plus high-dose interferon alfa (6 MU/day for 4 weeks then 9 MU/week for 22 weeks) or conventional treatment (9 MU/week for 26 weeks) [ ]. There was no difference in the incidence or the time to occurrence of thyroid disorders between the groups. Of the 30 affected patients, 11 (37%) had positive thyroid peroxidase autoantibodies (compared with 1% of patients without thyroid dysfunction), nine developed symptomatic thyroid dysfunction, and only three had to discontinue treatment. There was no correlation between the viral response and the occurrence of thyroid disorders, and only female sex and Asian origin were independent predictors of thyroid disorders.

Data on interferon alfa-associated thyroid disease have been comprehensively reviewed [ , ]. There was a mean prevalence of 6% for incident thyroid dysfunction, and treatment for malignancies was associated with the highest prevalence (11%). Hypothyroidism occurs more often than hyperthyroidism, and spontaneous resolution is expected in almost 60% of patients with or without interferon alfa withdrawal. Finally, female sex and the presence of baseline thyroid autoimmunity were confirmed to be the most significant risk factors. The mechanisms of interferon alfa-induced thyroid dysfunction are not yet fully clarified. Although an autoimmune reaction or immune dysregulation are the most likely mechanisms, a direct inhibitory effect of interferon alfa on thyrocytes should be considered in patients without thyroid antibodies.

Data on the incidence of thyroid disorders in interferon alfa-treated patients vary, largely because the follow-up duration, the nature of the study (prospective or retrospective), biological monitoring, diagnostic criteria, and the underlying disease differ from study to study [ ]. The incidence of clinical or subclinical thyroid abnormalities is generally 5–12% in large prospective studies in patients with chronic hepatitis C treated for 6–12 months, but it reached 34% in one study [ , ]. The incidence was far lower in patients with chronic hepatitis B, at 1–3%. A wider range in incidence was found in patients with cancer, with no clinical thyroid disorders in 54 patients treated during a mean of 16 months for hematological malignancies [ ], whereas in many other studies there was a 10–45% incidence [ ]. Even more impressive was the escalating incidence of thyroid disorders in patients with cancer receiving both interferon alfa and interleukin-2 (qv).

Mechanisms Possible mechanisms need to be clarified. Since thyroid autoantibodies are detected in most patients who develop thyroid disorders, the induction or exacerbation of pre-existing latent thyroid autoimmunity is the most attractive hypothesis. This is in accordance with the relatively frequent occurrence of other autoantibodies or clinical autoimmune disorders in patients who develop thyroid disorders [ ]. However, 20–30% of patients who develop thyroid diseases have no thyroid antibodies, and it is thus not yet proven that autoimmunity is the universal or primary mechanism. In fact, there were subtle and reversible defects in the intrathyroidal organification of iodine in 22% of antithyroid antibody-negative patients treated with interferon alfa [ ]. In addition, the acute systemic administration of interferon alfa in volunteers or chronic hepatitis patients reduces TSH concentrations [ ], and in vitro studies have suggested that interferon alfa directly inhibits thyrocyte function [ ]. Finally, the thyroid autoantibody pattern in patients who developed thyroid dysfunction during cytokine treatment was not different from that of patients without thyroid dysfunction, but differed significantly from that of patients suffering from various forms of spontaneous autoimmune thyroid disease [ ].

Susceptibility factors In addition to the underlying disease, there are many potential susceptibility factors [ , ]. There is as yet no definitive evidence that age, sex, dose, and duration of treatment play an important role in the development of thyroid disorders. However, patients with previous thyroid abnormalities are predisposed to develop more severe thyroid disease [ , ]. The incidence of thyroid disease was not different between natural and recombinant interferon alfa. Although this should be taken into account, a previous familial or personal history of thyroid disease was generally not considered a major risk factor. Finally, only pre-treatment positivity or the development of thyroid antibodies during treatment seem to be strongly associated with the occurrence of thyroid dysfunction.

In 175 patients with hepatitis B or C virus infections, women with chronic hepatitis C and patients with previously high titers of antithyroid autoantibodies were more likely to develop thyroid disorders [ ].

The immunological predisposition to thyroid disorders has been studied in 17 of 439 Japanese patients who had symptomatic autoimmune thyroid disorders during interferon alfa treatment [ ]. There was a significantly higher incidence of the human leukocyte antigen (HLA)-A2 haplotype compared with the general Japanese population (88 versus 41%), suggesting that HLA-A2 is a possible additional risk factor for the development of interferon alfa-induced autoimmune thyroid disease.

Among other potential predisposing factors, treatment with iodine for 2 months in 21 patients with chronic hepatitis C receiving interferon alfa did not increase the likelihood of thyroid abnormalities compared with eight patients who received iodine alone, but abnormal thyroid tests were more frequent compared with 27 patients who received interferon alfa alone [ ]. This suggests that excess iodine had no synergistic effects on the occurrence of thyroid dysfunction induced by interferon alfa.

The occurrence of thyroid dysfunction in 72 patients treated with interferon alfa plus ribavirin (1.0–1.2 g/day) has been compared with that of 75 age- and sex-matched patients treated with interferon alfa alone for chronic hepatitis C [ ]. Of the former, 42 patients, and of the latter, 40 patients had received previous treatment with interferon alfa alone. There was no difference in the rate of thyroid autoimmunity (antithyroglobulin, antithyroid peroxidase, and thyroid-stimulating hormone receptor antibodies) between the two groups, but the patients who received interferon alfa plus ribavirin developed subclinical or overt hypothyroidism more often (15 versus 4%). Similarly, the incidence of hypothyroidism increased to 19% in patients who underwent a second treatment with interferon alfa plus ribavirin compared with 4.8% after the first treatment with interferon alfa alone, while the incidence remained essentially the same in patients who had two consecutive treatments with interferon alfa alone (4.7 and 7.1% respectively). Furthermore, there was no higher incidence of thyroid autoimmunity or clinical disorders after a second course of interferon alfa whether alone or combined with ribavirin in patients who had no thyroid autoantibodies at the end of a first course of interferon alfa alone, suggesting that these patients are relatively protected against the development of thyroid autoimmunity.

Management The management of clinical thyroid dysfunction depends on the expected benefit of interferon alfa. Assay of thyroid antibodies before treatment, and regular assessment of TSH concentrations in treated patients, even after interferon alfa withdrawal, are useful as a means of predicting and detecting the risk of thyroid disorders. Complete recovery of normal thyroid function is usually observed after thyroxine replacement but sometimes requires interferon alfa withdrawal. Sustained hypothyroidism requiring long-term substitution treatment has occasionally been observed [ ], and is more likely in patients with initially severe hypothyroidism and raised thyroid antibody titers [ ]. By contrast, hyperthyroidism generally requires the prompt withdrawal of interferon alfa, and severe forms may require radical radioiodine therapy. Although not enough data are available on the long-term consequences of interferon alfa-induced thyroid dysfunction, the recurrence of thyroid abnormalities after the administration of pharmacological doses of iodine should be borne in mind [ ].

Parathyroid

Exacerbation of secondary hyperparathyroidism occurred in a 20-year-old renal transplant patient who also developed psoriasis during interferon alfa treatment [ ]. Both disorders resolved after withdrawal.

Adrenal

Of 62 initially autoantibody-negative patients treated with interferon alfa for chronic hepatitis C for a mean of 8 months, three developed antibodies to 21b-hydroxylase, a sensitive assay of adrenocortical autoimmunity [ ]. However, there were no cases of Addison’s disease or subclinical adrenal insufficiency. This study suggested that the adrenal cortex is another potential target organ of autoimmune effects of interferon alfa, along with thyroid and pancreatic islet cells.

Diabetes mellitus

The development or worsening of insulin-dependent diabetes mellitus is limited to isolated case reports in patients treated with interferon alfa or interferon alfa plus interleukin-2 [ ]. In chronic hepatitis, diabetes mellitus was noted in only 10 of 11 241 treated patients [ ]. Although a relation between chronic hepatitis C and the occurrence of glucose metabolism disorders is possible [ ], reports of diabetes mellitus in patients treated with interferon alfa were probably more than coincidental. Indeed, there have been reports of prompt amelioration or complete recovery after interferon alfa withdrawal [ , ] and of successive episodes of diabetes after each course of interferon alfa [ ].

• In three middle-aged patients, diabetes was diagnosed after 3–7 months of treatment with interferon alfa-2b and ribavirin, and two presented with severe ketoacidosis [ , ]. There was a family history of diabetes in one patient and two had high titers of glutamic acid decarboxylase antibodies before treatment. One patient never had diabetes-related serum autoantibodies before or after interferon alfa therapy. All three required permanent insulin treatment despite withdrawal of interferon alfa.

• Insulin-dependent diabetes mellitus has been reported after 2 weeks to 6 months of treatment with interferon alfa in four patients with chronic hepatitis C [ ]. All discontinued interferon alfa, and one woman who restarted treatment had a subsequent increase in insulin requirements.

A 75 g oral glucose tolerance test was performed before and after 3 months of interferon alfa treatment in 32 patients with chronic hepatitis C, of whom 15 also had an intravenous glucose test [ ]. Baseline evaluation showed that five patients had mild diabetes mellitus, three had impaired glucose tolerance, and 24 were normal. After 3 months of treatment, two patients with diabetes mellitus shifted to impaired glucose tolerance, and all patients with impaired glucose tolerance had normal glucose tolerance. Only three initially normal patients developed impaired glucose tolerance and none had newly diagnosed diabetes mellitus. From these results, and in contrast to previous reports [ ], it appears that interferon alfa did not have any adverse effects on insulin sensitivity and glucose tolerance after 3 months of treatment.

Interferon alfa may produce more severe changes than interferon beta [ ].

• A 39-year-old man with diabetes, stabilized with insulin 22 U/day for 13 years, received interferon beta (6 MU/day) for chronic hepatitis C. His diabetes progressively worsened, necessitating insulin 50 U/day. After 4 weeks, interferon beta was replaced by interferon alfa (10 MU/day). Shortly afterwards he developed severe diabetic ketoacidosis and shock, which reversed after hemodynamic support and continuous hemodiafiltration.

• A 39-year-old man who had received peginterferon alfa-2b and ribavirin for 9 months developed diabetes mellitus 3 months after interferon withdrawal [ ]. He required oral hypoglycemic drugs and insulin, but fasting plasma glucose and glycosylated hemoglobin concentrations later normalized despite withdrawal of antidiabetic treatment.

Incidence In patients with chronic hepatic B or C the respective prevalences of pancreatic autoantibodies increased from 2% and 3% at baseline to 5% and 7% after interferon [ ]. In all, 31 published cases of type 1 diabetes mellitus attributed to interferon alfa treatment were detailed, mostly in patients with hepatitis C. Irreversible diabetes required permanent insulin treatment in all but eight cases. At least one marker of pancreatic autoimmunity was positive in nine of 18 patients before treatment, and in 23 of 30 patients at the onset of diabetes. In accordance with these results and the likelihood of a genetic predisposition, the authors recommended screening for islet cell and glutamic acid decarboxylase autoantibodies before and during interferon alfa treatment. However, owing to the low number of reported cases and the paucity of studies that have examined the relation between pancreatic autoimmunity and the occurrence of diabetes, further research on the predictive potential of such a systematic investigation is warranted.

In a randomized trial in 74 patients with chronic hepatitis C treated with interferon alfa-2b and ribavirin, plus placebo or amantadine, two developed glutamic acid decarboxylase (GAD) autoantibodies, but none developed IA-2 or insulin autoantibodies [ ]. One had an increased titer of GAD autoantibodies during a first sequence of interferon alfa monotherapy, then a further rise during subsequent combination therapy, and finally developed diabetes mellitus after 5 months of treatment. The authors suggested that repetitive treatment with interferon alfa could increase the risk of type 1 diabetes in patients previously positive for islet antibodies.

Mechanisms Autoimmunity was suggested as a likely mechanism, with HLA-DR4 haplotype and/or islet cell antibody (ICA) positivity at the time of diagnosis in several patients. Because the induction of ICA antibodies in patients treated with interferon alfa has never been otherwise demonstrated [ ], the triggering, rather than the induction, of a latent autoimmune phenomenon in patients with a genetic susceptibility is probable [ ].

More direct interference with glucose metabolism cannot be excluded. Interferon alfa can reduce the sensitivity of peripheral tissues or liver to insulin and accelerate the destruction of stimulated pancreatic beta-cells [ , ]; this could be a possible mechanism in patients not exhibiting islet cell antibodies. This is also in keeping with rare instances of induction or exacerbation of type II non-insulin dependent diabetes mellitus [ , ].

Insulin antibodies were also found in six of 58 patients treated for chronic viral hepatitis [ ] and that was associated with signs of insulin allergy in one patient [ ].

Susceptibility factors Patients with obesity and a family or previous history of glucose intolerance should be considered more predisposed to interferon alfa-induced diabetes, but the association is not consistently found [ ].

Dyslipidemia

Interferon alfa often affects lipid metabolism and produces a reversible reduction in cholesterol and, more consistently, increases in triglyceride concentrations [ ]. Meticulous blood lipid investigation showed a significant rise in serum triglyceride and lipoprotein(a) concentrations and reductions in total cholesterol, HDL cholesterol, LDL cholesterol, and apoprotein A1.

Marked hypertriglyceridemia (10–20 μg/ml), which abates when treatment is withdrawn, has sometimes been observed [ , ]. Inhibitory effects of interferon alfa on lipoprotein lipase and triglyceride lipase or increased hepatic lipogenesis have been suggested [ , ]. Diet and lipid-lowering drugs have been proposed as means of maintaining acceptable triglyceride concentrations during long-term interferon alfa therapy. Although the possibility of pancreatic or cardiovascular complications should be borne in mind, no secondary clinical consequences of interferon alfa-induced blood lipid disorders have been so far reported.

In a prospective study of lipid changes in 36 patients with chronic hepatitis C treated with interferon alfa for 6 months, the most prominent findings included increases in triglycerides, VLDL cholesterol, and apolipoprotein B, and falls in HLD cholesterol and apolipoprotein A1 [ ]. Three patients also developed chylomicronemia and two of those had severe hypertriglyceridemia. All three patients had triglycerides over 2 μg/ml before treatment, suggesting that patients with abnormal serum triglyceride concentrations at baseline are more likely to develop marked hypertriglyceridemia.

Severe hypertriglyceridemia (7.5–19 mmol/l; 653–1644 mg/dl) has been reported in three patients receiving adjuvant high-dose interferon alfa for malignant melanoma [ ]. The authors reviewed the available literature and proposed a detailed surveillance and management plan for this metabolic disorder in patients with melanoma treated with interferon.

Porphyria

• A severe acute flare of porphyria cutanea tarda has been reported in a 61-year-old man after 4 months of treatment with interferon alfa-2b plus ribavirin for chronic hepatitis C [ ]. No further relapse was observed after chloroquine treatment, despite continuation of the antiviral drugs.

This patient had previously had episodes of small blisters that spontaneously resolved, and hereditary porphyria cutanea tarda was demonstrated by chromatographic and mutation analysis.