Two Episodes of Systemic Capillary Leak Syndrome in an 8-year-old Boy, Following Influenza A Virus Infection

Abstract

Systemic capillary leak syndrome is a rare condition, characterized by hypotension, edema, hemoconcentration and hypoalbuminemia. We describe 2 episodes of systemic capillary leak syndrome, following influenza A virus infection, occurring during 2 subsequent influenza seasons, in an 8-year-old boy.

Systemic capillary leak syndrome (SCLS) is a rare condition that presents with acute episodes of hypotension, generalized or localized edema, hemoconcentration, represented by markedly elevated hemoglobin and hematocrit levels, and hypoalbuminemia in the absence of albuminuria.¹ The manifestations are thought to be the result of transient systemic capillary hyperpermeability caused by endothelial dysfunction.^1,2 Acute episodes of SCLS usually develop in a matter of hours and can be either idiopathic or secondary to a known causative factor, such as hepatitis C virus infection or malignancies.¹ These acute episodes normally follow a triphasic course. At first, the signs are nonspecific, with coughing, fever, nausea and vomiting. The second phase is characterized by generalized or localized edema, which can lead to hypovolemic shock. During the third phase, spontaneous shift of fluid back to the intravascular space occurs, which can result in fatal pulmonary edema, especially in the presence of extensive fluid resuscitation during the state of shock.^1,3 The resolution of such acute episodes usually occurs in less than a week.¹ Diagnosis is based on the presence of characteristic triad of symptoms, namely hemoconcentration, hypotension and hypoalbuminemia, in the absence of other causes of shock.² The diagnosis is further supported by the presence of monoclonal gammopathy, present in more than 90% of patients with SCLS.¹ However, the mechanism of its occurrence is not yet clear.² Because the diagnosis of SCLS is 1 of exclusion, several differential diagnoses are first considered; most common are infections, anaphylaxis and malignancies. The most important part of the treatment strategy is extensive fluid resuscitation and inotropic and vasopressor support.^1–3 The syndrome usually presents in adults, most commonly after the age of 40 years. There have only been a handful of reports of this syndrome in children.³ In this report, we will detail 2 episodes of SCLS in an 8-year-old boy, both of which occurred after influenza A virus infection.

CASE PRESENTATION

In January 2011, an 8-year-old boy presented with vomiting and nausea. His past medical history was unremarkable. The symptoms began 2 days before presentation with nasal discharge, general weakness, cough and an elevated body temperature of 39°C. Later he developed facial edema, vomiting and tachycardia. Physical examination revealed a body temperature of 34.1°C, eyelid edema, elevated heart rate of 155/min and prolonged capillary refill. Blood pressure could be measured only on his leg and was 61/44 mmHg; femoral pulses were not palpable and his calves and thighs were painful on palpation. Laboratory tests showed marked hemoconcentration (Hb 201 g/L, Ht 0.61), leukocytosis (29 × 10⁹/L) and hypoalbuminemia (21 g/L) with normal inflammation markers (C-reactive protein < 3 mg/L). While thrombocytes were within the normal range (437 × 10⁹/L), coagulation tests showed signs of disseminated intravascular coagulopathy (prothrombin time 0.48 seconds, international normalized ratio 1.54 seconds, partial prothrombin time 58.2 seconds, fibrinogen 0.99 g/L, thrombin time 35.2 seconds, antithrombin 0.52 seconds, D dimer 874 µg/L,). Coagulation markers were repeatedly measured but stayed within normal range throughout. Blood gas analysis showed metabolic acidosis (pH 7.12, pCO2 5.2, pO2 5.5, base excess −15.4). Muscle enzymes remained within normal range, apart from creatine kinase, which rose to 750.9 µKat/L on day 3, but then steadily returned to normal. Chest radiograph showed normal lungs and heart. Urinalysis was performed and was within normal range. Rapid influenza diagnostic test was positive for influenza A (H3N2) virus, with the sample obtained by nasopharyngeal swabbing. The patient was hemodynamically unstable, for which extensive fluid resuscitation with crystalloids and colloids was required. Because of progressive and resistant shock, the patient was intubated on admission and was mechanically ventilated for 5 days. Vasopressor and inotropic support were initiated, with noradrenalin and dopamine infused for 4 and 7 days, respectively. He received continuous infusion of hydrocortisone in stress levels for 24 hours. Four days after admission, rhabdomyolysis and myoglobinuria (Mb 31948 mg/L) developed, which led to acute renal insufficiency and arterial hypertension. Renal tests were markedly elevated, with urea and creatinine 32.9 mmol/L and 424 µmol/L, respectively. Hemofiltration and hemodyalisis were required for 8 days, after which renal function returned to normal; however, persistent arterial hypertension was treated with amlodipine. Influenza infection was treated with oseltamivir for 7 days. He was discharged after 2 weeks, with normal laboratory values. After discharge, severe muscle atrophy persisted, for which he required intensive physiotherapy. At the follow-up examination 10 months after discharge, he was without sequelae.

Fourteen months later, in March 2012, the patient presented again with eyelid edema, signs of hypovolemic shock with blood pressure on his arm 95/59 mmHg, tachycardia of 135/min, capillary refill >5 seconds and with impalpable peripheral pulses. His body temperature was 35.3°C. Laboratory tests revealed marked hemoconcentration (Hb 208 g/L, Ht 0.61) and leukocytosis (18 × 10⁹/L) with normal inflammation markers (C-reactive protein <3 mg/L) and hypoalbuminemia (16 g/L). Blood gas analysis showed metabolic acidosis (pH 7.21, base excess −12.7) and rapid influenza diagnostic test was positive for influenza A (H1N1) virus. Renal function was normal. Extensive fluid resuscitation with crystalloids and colloids was required and he received inotropic support with dopamine on the first day and human albumins due to hypoalbuminemia. Treatment with oseltamivir was started on day 1 and was discontinued after 5 days. Diagnosis of SCLS was made, after infections, adrenal insufficiency and other metabolic disturbances were excluded with appropriate blood tests. Biopsy of the skin and thigh muscle was unremarkable. No specific therapy for SCLS was introduced, because the patient had recovered by the time the diagnosis was made. The patient recovered without sequelae and was discharged after 3 days with normal laboratory values.

During his second episode and at the follow-up visits 6 weeks, 3 months and 1 year after the second episode his immune status was evaluated. All lymphocyte subsets remained within normal range together with normal results of in vitro lymphocyte activation. Normal serum immunoglobulin levels were found with good specific immunoglobulin response to varicella zoster virus infection and specific immunoglobulin response to tetanus and diphtheria vaccination. Complement studies including total hemolytic activity, C1-inhibitor activity and serum levels of complement proteins C3, C4 and C1inh were within normal range. There was no monoclonal gammopathy present during the follow-up period, but low levels of circulating immune complexes were detected throughout the whole period. After the second episode, the patient was vaccinated against influenza and developed good immunogenity to influenza A vaccine. In 1-year follow-up period, we did not observe similar episodes.

DISCUSSION

SCLS is a rare disorder with the acute attacks occurring spontaneously, without a clear precipitating factor. However, as was the case in our patient, a causative factor of SCLS is sometimes suspected. In our patient, both episodes of SCLS were concurrent with influenza A virus infection, which was later suspected as the triggering factor. There have been previous reports of pulmonary capillary leak syndrome following influenza A virus infection, confirming the possibility of the virus being responsible for the development of SCLS in our patient.⁴ Because the patient recovered from every other infection without developing symptoms and signs of SCLS, an inappropriate immune response to influenza virus or specific failure of immunity to influenza was a possibility. Many studies thus far suspect an immunologic basis for the syndrome, reporting a high prevalence of monoclonal gammopathy of unknown significance in SCLS cases.^5,6 If this were the case, the SCLS in our case could have been caused by endogenous immunoglobulins released as a response to virus infection. However, monoclonal gammopathy was not detected in our patient. Alternatively, nonimmunoglobulin humoral factors could contribute to vascular hyperpermeability as an underlying pathology in SCLS. In this case, virus infection could function as a triggering factor for the release of leak-inducing mediators, among which vascular endothelial growth factor is thought to play an important role.⁷ Our case is hence interesting because both episodes of the syndrome coincided with influenza A virus infection, while the patient recovered from every other infection without developing symptoms and signs of SCLS. It would thus be prudent to further delve into the mechanisms of endothelial dysfunction in SCLS and what the specific link between influenza virus infection and damage to the endothelium could be.

Furthermore, the first episode of SCLS in our patient was complicated by the development of rhabdomyolysis and acute renal insufficiency. At first they were thought to be complications of influenza infection; however, in retrospect, rhabdomyolysis, which led to acute renal failure due to myoglobinuria, could also have been a direct manifestation of SCLS. There have been few reports of renal damage and rhabdomyolyis as rare manifestations of this syndrome.⁸ Systemic capillary hyperpermeability leads to generalized edema, which can cause increased compartment pressure. This in turn leads to ischemic muscle necrosis, resulting in myoglobinuria and acute renal insufficiency.^3,6,8 During the second episode, rhabdomyolysis or renal failure did not occur.

Moreover, our case emphasizes the role of sufficient yet judicious use of fluids as the most important part of treatment strategy.^1,2,5,9 Both episodes in our patient followed the described triphasic course.³ During the phase of vascular hyperpermeability, extensive fluid resuscitation was required, with crystalloids as well as with colloids. In addition, inotropic and vasopressor support were instituted and substantial amounts of human albumins were infused because of marked hypoalbuminemia. Because septic shock was first suspected as the primary diagnosis in our patient, central venous catheter was inserted and central venous oxygen saturation was measured continuously. Such continuous monitoring of several variables (ie, blood pressure, body temperature, heart rate, blood oxygen saturation, urine output, central venous oxygen saturation) is crucial during fluid resuscitation. The transition between the phase of hyperpermeability and the phase of recovery can be difficult to detect and if the amount of fluids infused is too large, fatal pulmonary edema can occur.¹ In our case, every infusion of fluids was carefully documented, as was the patient’s urine output, which enabled us to taper down the fluid infusions accordingly.

Finally, as the diagnosis of SCLS is 1 of exclusion, it is very important to keep SCLS as a diagnostic option in cases where no other cause for hypovolemic shock is apparent.

CONSENT

Written informed consent was obtained from the patient for publication of this Case report and any accompanying images. A copy of the written consent is available for review by the Editor of this journal.