|Home | About | Journals | Submit | Contact Us | Français|
Hypoxemia in severe leptospirosis-associated pulmonary hemorrhage syndrome (LPHS) is a challenging clinical scenario not usually responsive to maximal support on mechanical ventilation. We described the efficacy and safety of high frequency oscillatory ventilation (HFOV) as rescue therapy in acute respiratory failure secondary to LPHS. This is a retrospective case study of five patients with diagnosis of severe LPHS, who were admitted to Intensive Care Unit from October 2014 to January 2015. They developed refractory hypoxemia on conventional mechanical ventilation and rescue therapy was indicated. All patients responded rapidly by showing improvements in oxygen index and PaO2/FiO2 ratio within first 72 h of therapy. Despite severity of illness evidenced by high Simplified Acute Physiological II and Sequential Organ Failure Assessment scores, all patients were discharged from hospital alive. In view of the rapid onset and extent of hemorrhage which may culminate quickly into asphyxiation and death, HFOV may indeed be lifesaving in severe LPHS.
Leptospirosis, a zoonotic infection, is an acute febrile illness caused by pathogenic spirochetes of the genus Leptospira. Generally, there are two types of infection that portray different signs and symptoms, which are anicteric and icteric leptospirosis. Ninety percent of cases are anicteric, which means that jaundice does not occur. The clinical course of leptospirosis is variable. It has inconsistent clinical manifestations which range from mild disease to severe form, classically presenting as triad of Weil, i.e. fever, jaundice, and acute kidney injury (AKI).[1,2] Weil syndrome primarily manifests as profound jaundice, renal dysfunction, hepatic necrosis, pulmonary dysfunction, and hemorrhagic diathesis. The icteric phase is much more severe than the anicteric one and may develop into a more fatal condition. Other complications include rhabdomyolysis, myocarditis, and uveitis. Weil syndrome with hepatorenal involvement and jaundice has a mortality rate of 20–40%. Pulmonary manifestations occur in 20–70% of leptospirosis patients but are often overshadowed by other symptoms. Lung involvement may vary from subtle clinical features to acute respiratory distress syndrome (ARDS) and deadly leptospirosis-associated pulmonary hemorrhage syndrome (LPHS).[3,4]
Leptospirosis is distributed worldwide, with majority of cases occurring in impoverished populations inhabiting developing countries with tropical climates. In a recent study by Costa et al., leptospirosis was estimated to cause more than 1 million severe cases and approximately 60,000 deaths each year. Therefore, it has emerged as a leading zoonotic cause of morbidity and accounts for numbers of deaths, which approach or exceed those for other causes of hemorrhagic fever. In the last two decades, there are an increasing number of LPHS cases in patients with severe pulmonary involvement. Segura et al. reported 321 cases with serologic and clinical evidence of leptospirosis in Peru, 189 were from urban location and 132 from rural areas. Seven patients (3.7%) with a history of urban exposure had severe pulmonary manifestations; of these, five patients died; four deaths were caused by severe pulmonary hemorrhage, and one was caused by ARDS and multiorgan failure. In a 12-month study, pulmonary hemorrhage occurred in 14 of 75 patients (19%) admitted to hospital with leptospirosis in Seychelles, causing deaths of six patients (8%). However, the incidence of LPHS in Malaysian population is unknown. According to data from Ministry of Health Malaysia, the annual incidence rate of leptospirosis has increased dramatically from 6.99 to 25.94 cases per 100,000 populations from 2010 to 2014. The mortality rate has also increased from 0.24 to 0.31 per 100,000 populations over the last 10 years.
Gross pathology of the lung in severe pulmonary form of leptospirosis generally shows extensive pulmonary hemorrhage, with numerous bleeding foci of different sizes. Histological sections usually reveal pulmonary congestion, pulmonary edema, and several foci of interstitial and intra-alveolar bleeding. In a recent study by Croda et al., it was found that the linear deposition of immunoglobulins (IgA, IgG, and IgM), and complement on the alveolar surface may play a role in the pathogenesis of pulmonary hemorrhage in LPHS. This is due to a toxin-mediated process which induces vascular injury, particularly a small vessel vasculitis. There is remarkable leakage of fluid and proteins into intravascular space with extrusion of erythrocytes culminating into widespread hemorrhagic infiltrates. Electron microscopy finds damage into capillaries as the primary lesion. Endothelial cells swell and detach from the basement membrane leaving areas of exposed interstitium, even in areas free of hemorrhage.[3,11]
A high index of suspicion is required to make the diagnosis based on epidemiologic exposure and clinical manifestations. As the clinical features are nonspecific, leptospirosis can be confirmed only by laboratory tests. Microscopic agglutination test is a reference standard but not readily available. A number of simple rapid tests such as IgM ELISA and macroscopic slide agglutination test have been developed; however, these tests become positive only after a week. The organism can be cultured, but this may take several weeks. Therefore, definitive guidelines for the diagnosis of human leptospirosis are necessary such as the modified Faine's criteria with amendment (2012) from the original WHO guidelines 1982. In these guidelines, the diagnosis is based on three categories, i.e., clinical data (Part A), epidemiological factors (Part B), and bacteriological and lab findings (Part C). In the first week, clinical and epidemiological factors (A + B) are of value, whereas in the 2nd week, laboratory tests should be used to confirm diagnosis. When available, polymerase chain reaction can be used as a diagnostic tool in the 1st week. In the recent 2012 guidelines, hemoptysis and dyspnea were included in the clinical criteria as pulmonary hemorrhage is recognized as an important complication. Presumptive diagnosis of leptospirosis is made based on scoring when Part (A + B + C) is 25 or more. In the 1st week when laboratory tests are not available, a score of 26 or more for Part (A + B) is diagnostic. The (A + B) criteria should be used to start empiric therapy even for possible leptospirosis (A + B = 20 − 25).
Severe LPHS is a serious complication of leptospirosis and has reported mortality rates of up to 75% even when optimal treatment is provided.[3,13] Accumulation of hemorrhagic fluid in the alveoli may result in airway obstruction and ventilation-perfusion mismatch with acute gas exchange failure. In this situation, hypoxemia in LPHS may be difficult to treat in spite of timely initiation of conventional mechanical ventilation (CMV). Data on the use of high frequency oscillatory ventilation (HFOV) for this disease are scarce in the literature. We report our experience with HFOV as rescue therapy in patients with massive pulmonary hemorrhage who deteriorated on CMV.
This is a retrospective case series, reviewed from patient records in a tertiary adult Intensive Care Unit (ICU) in Pahang, Malaysia. All five patients were admitted to ICU in between October 2014 and January 2015, which is within Northeast monsoon season that brings more rainfall compared to other months. Publication of patients' data, informed consent, and anonymity requirements is in agreement with the rules of the Ethical Committee of our institution. All leptospirosis patients with pulmonary hemorrhage were initially ventilated with either bilevel positive airway pressure control ventilation or volume control ventilation. HFOV (SensorMedics 3100B, Yorba Linda, CA, USA) was started as rescue therapy for ARDS when there was severe hypoxemia (PaO2/FiO2 <100 mmHg, arterial saturation SaO2 <90%) with or without ventilation failure (pH <7.25) and plateau airway pressure ≥30 cmH2O. Oxygen index (OI) was calculated using formula; 100 × mPaw × FiO2/PaO2. Mean airway pressure (mPaw) was initially set as 5 cmH2O above plateau pressure (plateau P) on CMV but not exceeding 40 cmH2O. All patients were adequately volume resuscitated and had mean arterial pressure (MAP) ≥65 mmHg with or without continuous infusion of vasoactive agents before transition to HFOV. Recruitment maneuver was performed with mPaw of 40 cmH2O for 40 s with precautions and terminated immediately if associated with hypotension (MAP <60 mmHg or decrease by>20 mmHg) or oxygen desaturation (decrease in oxygen saturation, SpO2 to ≤85%, or decrease of more than 5%). Results are presented as mean ± standard deviation for normally distributed variables or median (interquartile range) for nonnormally distributed variables. Comparisons of oxygenation index, PaO2/FiO2, and PaCO2 over time were analyzed using the repeated measures analysis of variance (ANOVA). All nonnormally distributed variables were log-transformed before repeated measures ANOVA analysis.
Five patients who progressed to severe LPHS and treated with HFOV as rescue therapy were included in the study. Their main clinico-epidemiological and laboratory blood results are demonstrated in Tables Tables11 and and2,2, respectively. Four patients had clinico-epidemiological information in agreement with those observed in icteric leptospirosis and one suffered anicteric leptospirosis. Their hemoglobin levels dropped to below 10 g/dL within 24 h or at the same day of ICU admission. All patients had high Sequential Organ Failure Assessment (SOFA) scores due to rapid progression to multiple organ dysfunction syndrome and septic shock. They required vasoactive drugs to maintain appropriate MAP ≥60 mmHg. All five patients had Stage 3 AKI and were prescribed renal replacement therapy (RRT) within 24 h of ICU admission. Duration of HFOV and CMV, length of ICU and hospital stays, and in-hospital mortality are shown in Table 3. Despite severity of illness evidenced by high Simplified Acute Physiological II and SOFA scores, all patients survived and were discharged from hospital alive.
Duration of patients on CMV before transition to HFOV is shown in Table 3. After omitting the time taken for transportation, the mean duration on CMV before changing to HFOV was 9.8 h. The CMV values before transition are presented in Table 4. They had mixed metabolic and respiratory acidosis and categorized as severe ARDS. Three patients had driving pressure (PlateauP– positive end-expiratory pressure [PEEP]) that was more than 15 mmHg, which in ARDS lungs are associated with increased mortality. They received high levels of PEEP, at the range of 12–20 cmH2O. After transition to HFOV, the median maximum mPaw was 34 cmH2O (interquartile range [IQR] 31–37). The improvement of OI during the first 72 h is shown in Figure 1. Compared to the baseline values, there was a marked improvement in oxygenation as depicted by reduction in median OI at 72 h. The median (IQR) of OI at baseline versus 72 h was 46.0 (37.9–53.3) versus 11.3 (11.2–15.6), with repeated measures ANOVA, P <0.001 and post hoc Tukey analysis, P < 0.005. Figure 2 shows the boxplot of PaO2/FiO2 ratio (vs. time) of all patients before and during the first 72 h on HFOV. Significant improvement in median PaO2/FiO2 was observed as early as 24 h after commencement of HFOV. After 48 and 72 h, the median value of PaO2/FiO2 ratio had improved significantly. The median (IQR) of PaO2/FiO2 ratio at baseline, 24, 48, and 72 h was 69.5 (56.3–84.6), 155.2 (114–280), 213 (184–388), and 232 (217–303), respectively, with repeated measures ANOVA, P = 0.002 and post hoc Tukey analysis P <0.005. Figure 3 shows the PaCO2 values of all patients at Time 0 and during first 72 h of HFOV. The median (IQR) of PaCO2 at baseline, 24, 48, and 72 h was 51.9 (48.6–68.6), 42.5 (36.8–67.9), 48.5 (36.6–58.8), and 36.7 (31.9–55.0), respectively. No significant improvement in PaCO2 was observed within first 72 h (P = 0.26). Table 5 demonstrated important hemodynamic data for all patients during the first 24 h of starting HFOV. The infusion rates for the vasoactive agents were the same as baseline or decreased after 4 h of initiation of therapy. Only two patients (number 3 and 5) received blood transfusion during treatment period.
Pulmonary hemorrhage which may be sudden, unexpected, and massive may be the major cause of death in patients with leptospirosis. Our current case series suggested that the use of HFOV is not only safe but also able to rapidly improve oxygenation and ventilation. All patients had severe ARDS and responded rapidly by showing improvements in OI and PaO2/FiO2 ratio within first 72 h of HFOV. In view of the rapid onset and extent of hemorrhage which may culminate quickly into asphyxiation and death, HFOV may indeed be lifesaving.
All patients had compatible clinical data, epidemiological factors, and serological diagnosis of leptospirosis with modified Faine's criteria scores, part (A + B + C) of more than 25. They presented with an acute clinical course of fever, acute respiratory insufficiency, bilateral diffuse pulmonary infiltrates in chest radiographs [Figure 4], thrombocytopenia, septic shock, and mixed respiratory and metabolic acidosis. All patients developed AKI and required RRT. Total bilirubin levels were increased in four patients, associated with increased liver transaminases (alanine transaminase and aspartate transaminase) and creatine kinase. Previous reports have shown that the severity of pulmonary leptospirosis does not correlate with the presence of icterus. Besides, severe pulmonary hemorrhage may occur in the icteric form or in the absence of jaundice. The evolution of LPHS is rapidly progressive, with acute gas exchange failure and asphyxia ensuing in just a few hours.
Leptospirosis is considered to be severe if a patient has acute febrile illness with complications. Empiric antibiotic therapy should be started as soon as clinical diagnosis is made without waiting for the confirmatory results of laboratory tests. In our practice, coinfection of leptospirosis can occur with malaria, dengue, viral hepatitis, and melioidosis. It is essential to investigate for these diseases even in patients who are confirmed to have leptospirosis. Severe leptospirosis can be treated with intravenous (IV) penicillin or ceftriaxone. Our patients received empirical IV meropenem which were later de-escalated to IV ceftriaxone 2g daily for 7 days once diagnosis was confirmed by rapid tests. Severe leptospirosis may require correction of hypovolemia, shock, and electrolyte abnormalities. Transfusion of blood and blood products may be required to manage severe bleeding which is due to LPHS or associated thrombocytopenia and consumptive coagulopathy. RRT for AKI and early mechanical ventilation support for respiratory failure would definitely reduce mortality.
Four of our patients received high dose of IV methylprednisolone. The increasing evidence that LPHS is immune-mediated has led to suggestion of outcome improvement through immunomodulation, for example, plasmapheresis, Ig, glucocorticoids, or cyclophosphamide. Several reports from India advocate the use of glucocorticoids in LPHS to reduce the need for ventilator support. The mortality benefit was not significant in cases where hemorrhage had already occurred and mechanical ventilation initiated. Corticosteroids would therefore be beneficial only if given within first 24 h of onset of pulmonary symptoms. The uses of plasma exchange and cyclophosphamide have been evaluated in patients with LPHS. However, concurrent removal of coagulation factors and consequent complications from dilutional coagulopathy and thrombocytopenia renders it unsuitable for these patients. Hypoxemia in LPHS may be difficult to treat despite maximal CMV given its rapid onset and extensive nature culminating into asphyxia and death. Arokianathan et al. have reported successful treatment of refractory hypoxemia with extracorporeal membrane oxygenation (ECMO). However, the use of HFOV in severe hypoxemia due to LPHS has not been extensively described in the literature.
ARDS may result in refractory and life-threatening hypoxemic respiratory failure. In severe cases, clinicians commonly utilize interventions termed “hypoxemic rescue therapies” in an attempt to improve oxygenation. Without these therapies, CMV is associated with high mortality. The ICU physicians may make a transition to rescue therapies such as HFOV or ECMO when both these modalities are at their disposal. The results of the recently published OSCILLATE and OSCAR trials that compared HFOV and lung protective strategy will likely discourage the routine use of HFOV in ARDS.[21,22] In these studies, many factors influence the outcome. Majority of the ICUs participating in the studies were new and inexperienced in HFOV. In the OSCILLATE trial, the HFOV group required higher vasopressor, more positive fluid balance and were more heavily sedated as compared to control group. Aggressive HFOV strategy may create significant hemodynamic instability due to the effect of very high intrathoracic pressure on the right ventricular afterload. A more positive fluid balance in the HFOV group would also be detrimental and affect mortality. Patient selection may be an important factor. In ARDS patients with heterogenous and nonrecruitable lung, the associated increased mPaw in HFOV may lead to overdistension of healthy lung regions without increased in aeration of collapsed alveoli.
However, HFOV may still have a role in selected groups of patients such as in pulmonary hemorrhage. In the current study, all patients required rescue therapy within 24 h of CMV initiation. HFOV uses high continuous distending mPaw to achieve lung recruitment and improve oxygenation. Such pressures tamponade the transudation of hemorrhagic edema and decrease blood flow from ruptured capillaries by increasing intrathoracic pressure and reducing pulmonary blood flow. Furthermore, continuous high pressure will result in fewer pressure swings during the respiratory cycle thus minimize alveolar bleeding. All the above mechanisms will reduce the acute extrusion of hemorrhagic fluid from injured capillaries as the result of toxin-mediated vasculitis. Duval et al. described the safety and efficacy of HFOV as rescue therapy for lung hemorrhage in a retrospective case study of nine children. HFOV was started when conventional ventilation was ineffective in controlling hemorrhage resulting in hypoxemia and hypercarbia. They reported treatment benefit in seven infants and mortality in two. A few factors may have contributed to the success of HFOV in our cohorts. Other than the possible mechanisms explained above, we noted that we were able to wean the vasoactive agents within 24 h of initiation of HFOV. We included cautious recruitment maneuver in our protocol. Aggressive volume resuscitation before transition to HFOV was paramount in achieving optimal hemodynamic goals during the first few hours of therapy. To the best of our knowledge, this is the first case study demonstrating the effectiveness and safety of HFOV intervention in adult patients with acute refractory hypoxemic respiratory failure due to LPHS.
Hypoxemia in LPHS is a challenging clinical scenario not usually responsive to maximal support on CMV. Our case study demonstrated that in such patients, HFOV was a safe and effective rescue therapy with a good outcome. This intervention may indeed be lifesaving in an otherwise rapid, overwhelming, and potentially fatal form of pulmonary hemorrhage.
There are no conflicts of interest.