Abstract

Pneumonia due to H1N1 infection is now very common. We report a case of ischemic stroke which arose subsequently to H1N1 influenza. The patient was a female who developed acute respiratory distress syndrome (ARDS) after H1N1 influenza, was ventilated as per standard protocol and started treatment with oseltamivir. When sedation was stopped during weaning from the ventilator, she was found to have left hemiparesis resulting from multiple infarctions in the brain. Contrary to thrombocytosis usually seen in acute influenza, the platelet counts in our patient actually dropped. We suspected that increased interleukin release or stickiness of the platelets might have caused this ischemic stroke. In the course of time, she had acceptable neurological recovery following treatment with aspirin and neuro-rehabilitation. This case report provides evidence that a rare, debilitating complication like stroke can occur in H1N1 infection. A high index of suspicion of the probability of a cerebrovascular event should be borne in mind and regular neurological assessment should be done in such cases.

Keywords

ARDS; cerebral infarction; H1N1 infection; influenza A virus, H1N1 subtype; ischemic stroke; respiratory distress syndrome, adult;

---

1. Introduction

H1N1 influenza (influenza A) is now a well known disease causing epidemics all over the world.1, 2, 3 It is seen mostly in the winter season, involving the respiratory system, leading to unexpected morbidity of pneumonia as well as mortality.1, 2, 3, 4, 5 Very rarely, H1N1 influenza may manifest with a variety of non-respiratory complications in elderly and immunocompromized patients as associated co-morbidities.1, 2, 3, 4, 5, 6, 7, 8, 9 A neurological complication, such as ischemic stroke, stands out as such a rare complication leading to increased morbidity and mortality.1, 3, 4, 5, 6, 7 Here, we report an interesting case, where the patient sustained H1N1 influenza and developed an unexpected ischemic stroke during the course of the illness.

2. Case report

A 50-year-old female patient was brought to our emergency department with complaints of breathing difficulties and a flu-like presentation for 3 days. She had a history of exposure to an H1N1-infected child. On examination, her oxygen saturation (SpO₂) in room air was found to be 75%, which improved to 90% with 100% oxygen therapy by a non-rebreathing face mask. Clinically, she was tachypnoeic (respiratory rate = 50/minute), tachycardic (heart rate = 136/minute), febrile (38.6 °C), and hypotensive [blood pressure (BP) = 86/50 mmHg]. Bilateral crepitations were detected on chest auscultation. However, her chest radiograph was found to be within normal limits.

She was transferred to the intensive care unit (ICU) for non-invasive ventilation (NIV) as she was deteriorating clinically. Her SpO₂ improved to 95% on NIV support [inspiratory positive airway pressure = 18, expiratory airway pressure = 10, fraction of inspiratory oxygen concentration (FiO₂) = 0.6] (Table 1). Then she started treatment with intravenous azithromycin (250 mg once/day) and oseltamivir (150 mg twice/day). Her throat and nasal swabs turned out to be positive for H1N1 influenza virus by the real-time reverse transcription and polymerase chain reaction test (RT-PCR test), which is routinely performed for suspected cases in our institute (Table 2, Table 3). On the 3^rd day, her pulmonary condition suddenly deteriorated (SpO₂ dropped to 75% and respiratory rate increased to 50/min); her trachea was thus intubated and invasive ventilation was instituted (Table 1, Table 4). Her C-reactive protein (CRP) was high (Table 2, Table 3). She was found to have developed acute respiratory distress syndrome (ARDS) as a complication of H1N1 pneumonia (Fig. 1A). She was then treated as per net ARDS protocol and sedated with morphine 3–5 mg/hour and midazolam 3–6 mg/hour infusions, in accordance with our institute's protocol for ARDS patients. No muscle relaxant was necessary throughout the period of invasive ventilation. Her respiratory condition and radiographic picture improved (Fig. 1B) after 7 days of mechanical ventilation and thus, the weaning process was started. Sedatives were stopped and over the next 2 days she could tolerate minimal ventilator support. However, we did not venture to extubate her, as she was not waking up (even though she could respond to pain). On neurological evaluation, she was found to have left hemiparesis. An immediate magnetic resonance imaging (MRI) of the brain was done, which showed multiple brain infarcts [right middle cerebral artery (MCA) territory and cerebellum] indicating an acute stroke (Figs. 2A, 2B). MRI angiography revealed complete obstruction of the right MCA (M1 segment), but no significant narrowing or obstruction was seen in the internal carotid or vertebral artery territories.

Her electrocardiogram (ECG) showed sinus rhythm and her vital parameters, excluding SpO₂, were within normal limits throughout the disease period. Her BP showed a trend of rising from the 8th day and normalized from the 14th day onwards. This might have been due to the stroke turning for the better (Table 5, Table 6). Since her admission to the ICU, thromboprophylaxis was started with enoxaparin (1 mg/kg, once/day subcutaneously). On the 6^th day, her platelet counts dropped to 64 × 10⁹/L (Table 3). Enoxaparin was replaced by fondaparinux (7.5 mg subcutaneously, once/day) and a blood sample was sent for heparin induced thrombocytopenia (HIT) screen, which was reported as negative (Table 3). As per standard guidelines of ischemic stroke, aspirin (81 mg once/day orally) was started. Echocardiography was performed, which ruled out any source of thrombus in the heart.

She underwent tracheostomy on the 20^th day because of her depressed neurological status and failure to wean from the ventilator. Neuro-rehabilitation was started. Later, she was successfully weaned off the ventilator and breathed oxygen with the Swedish nose. Slowly, over the next week, she started to respond to verbal commands and became coherent and oriented. However, left sided weakness still persisted. Repeat MRI scan revealed resolution of the cerebellar infarct, but the right MCA infarct still failed to resolve. Finally, she was shifted to a general ward, with the tracheostomy in situ. Her chest physiotherapy and neuro-rehabilitation were continued. Over time, she improved neurologically and achieved motor power up to grade four in the left lower limb. Later, her trachea was decannulated and her tracheostomy was closed after 4 weeks.

3. Discussion

H1N1 influenza is an acute viral illness which usually produces respiratory symptoms and may lead to fulminant viral and superadded bacterial pneumonia.1, 2, 3, 4, 5 Rarely, it can complicate other vital organs such as the heart, brain, or skeleto-muscular system, leading to catastrophic complications like myocarditis, pericarditis, seizures, encephalitis, myositis, rhabdomyolysis, or toxic shock syndrome, in elderly immunocompromized patients with other co-morbid conditions.1, 2, 3, 4, 5, 6, 7, 8, 9 We also noted some commonly reported complications and outcomes in the 184 H1N1 cases admitted to our hospital, which were on a par with those reported in literature (Table 7).1, 3 Ischemic stroke has been reported very rarely.3, 4, 6 However, during an influenza epidemic, such complications can be seen which may be the main reason why morbidity and mortality are increased.3, 4, 7

Any febrile illness, including H1N1 influenza, can lead to a stroke, by stimulating the pro-coagulant properties of blood resulting from hyperviscosity, thrombogenicity, increased platelet aggregation and adhesiveness, thus augmenting the already formed thrombi.4, 7, 8, 9, 10 Virus infected blood and endothelial cells can activate the coagulation system and cause a 4–5 fold increase in the expression of tissue factor.⁷ Recently, it has been found that influenza infection in elderly people can result in thrombin generation, fibrin deposition, and an increased release of the prohemostatic von-Willebrand factor and plasmin-α₂-antiplasmin.8, 9 The use of heparin can reduce this thrombin generation to some extent.⁸ Inflammation is known to cause a down regulation of thrombomodulin and endothelial cell protein C receptor via different cytokines, thereby reducing the ability to generate the natural anticoagulant activated protein C.⁹ Furthermore, inflammation which induces inhibition of fibrinolysis caused by elevated levels of plasminogen activator inhibitor 1 and interleukin 6 (IL-6), could increase the number of platelets and its thrombogenecity.4, 5, 7, 9

Influenza virus induces formation of reactive oxygen radicals, which activate nuclear factor NK-κB, the signaling of which is a prerequisite for influenza infection and subsequent cytokine storm consisting of mostly IL-6.1, 2, 9 The severity of influenza correlates with the amount of the released IL-6.1, 2, 5 Subsequently, IL-6 stimulates the production of CRP, fibrinogen and the coagulation factor (VIII:C), thus increasing the risk of thromboembolism.¹¹ Higher IL-6 production in H1N1 patients is seen to be associated with an increased risk of myocardial infarction and ischemic stroke.2, 5

Contrary to the usual thrombocytosis seen in influenza, platelet count in H1N1 cases is generally seen to be reduced.2, 3 In our case, we encountered a similar picture before the platelet count became normalized (Table 2). The reason for the thrombocytopenia in our patient may have been either severe sepsis, the H1N1 influenza itself, or superimposed disseminated intravascular coagulation (DIC) as reported recently.2, 3 This thrombocytopenia has been seen to be associated with increased morbidity, as well as mortality. Although DIC can result in consumption of platelets, in our case there was no feature suggestive of DIC (fibrinogen and D dimer were within normal limits). Our patient was febrile for some days due to H1N1 infection. This might have triggered the pro-coagulant activity and thrombogenicity of the blood. Hence, our patient most probably had increased stickiness of platelets, activation of pro-coagulants in blood, increased thrombogenicity and IL-6 storm resulting from severe H1N1 influenza combined with sepsis. These probably lead to ischemic stroke.

There are very few data to comment definitely on the reasons of a possible association between acute influenza and the occurrence of venous and pulmonary thromboembolism.² Based on the above evidence, it can be hypothesized that influenza can lead to a pro-coagulant state, resulting in a higher risk of thromboembolism.⁹ However, we did not find any signs of deep venous thrombosis clinically or sonographically in our patient.

Although the use of oseltamivir is reported to reduce the incidence of ischemic stroke, whether this protection is effective in the acute phase or not is still not clear.4, 5, 6 Our patient was started on a high dose of oseltamivir, considering the severity of her illness,1, 2, 3, 4, 5, 6, 10 but this did not confer any protection against the occurrence of stroke in our patient or in a recently reported case.³ Influenza vaccine has been shown to reduce the risk of ischemic stroke, but the response to the vaccine varies among different age groups.⁶ Unfortunately, our patient was not vaccinated, which might have increased the risk in our case. Even in the only reported case, the patient was fully vaccinated, but that vaccination could not prevent the stroke either.³

Ischemia of brain tissue, either from poor oxygenation or reduced perfusion, can lead to neurological damage, culminating in ischemic stroke. The longer the period of oxygen deprivation, the more extensive and irreversible is the damage. Therefore, to prevent stroke, it is critical to maintain an adequate oxygen supply and perfusion. Although our patient was on NIV initially in the ICU, we instituted invasive ventilation at the earliest moment when her oxygenation dropped, to provide her with the best possible oxygenation. Her sedation was managed according to the protocol of our institute, based on sedation criteria of net ARDS recommendation. Our ARDS patients are usually sedated with morphine and midazolam infusions, titrated to maintain Ramsay sedation level 5 or 6, for ventilation support. Sedation breaks (Ramsay sedation level of 4 or less) are allowed when the patient achieves an acceptable oxygenation at FiO₂ < 0.6 for more than 1day, in order to assess the neurological status. When the FiO₂ can be reduced further, sedation breaks are carried out daily. In our case, however, we could not do it earlier as the patient was not meeting the criteria. We kept the patient deeply sedated to reduce the oxygen requirement, which allowed the patient to survive with lower oxygen tension. This measure would certainly delay the neurological assessment. This maneuver could mask the clinical picture and delay diagnosis. In the recently reported case, the caregivers allowed regular sedation breaks for neurological assessment, but still could not prevent the occurrence of multiple strokes in their patient.³ It is better to taper the sedation and allow regular neurological assessment in such cases, even though the risk of stroke is high in severe H1N1 patients complicated with sepsis or DIC. Stroke should also be suspected when the blood pressure rises in an otherwise normotensive and well sedated patient, as happened in our case.

Until our case, ischemic stroke in H1N1 patients had not been reported and the only case reported recently was in the USA.³ Although they cited DIC and severe H1N1 infection superadded with sepsis as the precipitating cause for ischemia, they did not stratify any predetermined risk factor.³ Even in the available literature, the risk stratifications have only been reported in relation to susceptibility to H1N1 infection, non-stroke complications of H1N1 and mortality.1, 2, 3 Hitherto, no risk stratifications have been devised for stroke complicating an H1N1 infection. Our patient was only a case diagnosed as having mild schizophrenia, without any risk factors evidently pointing to a complication of H1N1 influenza. The liver enzymes were mildly elevated initially, but settled after a few days. Albumin were close to normal levels initially, but was subsequently reduced, possibly due to the ongoing sepsis. The patient's renal function was normal throughout and her creatinine kinase, (determined only once) was normal. These laboratory parameters are reported to be abnormal in H1N1 patients who are at risk of developing complications.1, 2, 3 Coincidentally, in our case, the BP showed a trend of increasing, which was found retrospectively after the stroke. Thus, this was a rise in BP in a well sedated normotensive patient without any rationale pointing to ischemic stroke. As we did not suspect such an event beforehand, and our patient did not have DIC, we were unable to pinpoint any specific risk factor in our patient. However, after observing this incident, we searched the literature for possible causes. In our unvaccinated patient, severe H1N1 infection could possibly have triggered pro-coagulant activity of blood, increased stickiness of platelets and IL-6 storm, all of which could increase the thrombogenicity, leading to ischemic stroke. As there is nothing clearly responsible, we could only extrapolate the possible risk factors reported in the literature related to influenza. There is the need of more such case reports or case series to comment firmly on risk factors predisposing H1N1 patients to ischemic stroke. The rarity of this sort of complication in H1N1 infection, in the absence of DIC, prompted us to report this case.

4. Conclusion

Although rare, ischemic stroke can sometimes occur in patients with severe H1N1 infection complicated with ARDS. Diagnosis in such cases may be delayed or masked if heavy sedation is used without giving sedation breaks for neurological assessment. This rare complication should be suspected in severe H1N1 influenza patients complicated by sepsis, if such a patient develops signs and symptoms suggestive of neurological damage, even if they do not develop thrombocytosis. Hence, sedation breaks should be given in such cases, whenever possible weighing the risk benefits of oxygenation versus stroke.

Financial support

The authors received no financial support for performing this case as well as writing this article.