Annals of Burns and Fire Disasters - vol. XIV - n.1 - March 2001

USE OF GLUCOCORTICOIDS IN A BURN PATIENT WITH ADULT RESPIRATORY SYNDROME

Ramos G., Patiho O., Sanchez Luceros D., Bolgiani A., Brunoldi D.,Prezzavento C., Benaim F.

CEPAQ, Benaim Foundation, Aleman Hospital, Buenos Aires, Argentina

SUMMARY. We present a male patient with burns in 23% body surface area. He developed adult respiratory distress syndrome (AF;DS) and septic shock during evolution. Although infection was under control, respiratory failure and vasopressor agent requirements did not improve until glucocorticoid treatment was initiated. Glucocorticoid treatment allowed us to take the patient off mechanical ventilation and suspend vasopressor agents during the first week of treatment. The use of glucocorticoids in burn patients, whether suffering from ARDS or septic shock, is controversial. However, it may be effective in the fibroproliferative phase of ARDS and septic shock with infection under control, as in the case presented.

Introduction

Septic shock and adult respiratory distress syndrome (ARDS) are important causes of death in burn patients. These conditions produce an unstable state between the causative agent and the inflammatory response, resulting in a mortality rate of approximately 50%.
Sepsis is defined as the systemic inflammatory response to an infection.'
Septic shock is defined as sepsis with hypotension, despite adequate fluid resuscitation, together with the presence of perfusion abnormalities that may include, but are not limited to, lactic acidosis, oliguria, or an acute alteration in mental status. Patients who are on inotropic or vasopressor agents may not be hypotensive at the moment when perfusion abnormalities are measured.' Septic shock occurs in 20-40% of septic patients, with a mortality rate of 40-90%. ARDS is defined as acute respiratory failure, with diffuse pulmonary infiltrates, a Pa02/FiO2 ratio of less than 200 with a pulmonary wedge pressure (pulmonary artery occlusive pressure) less than 18 cm H20, and absence of evidence of cardiac pump failure.The causes of ARDS are either pulmonary or extra- pulmonary.
In burn patients, the most frequent cause of pulmonary dysfunction is inhalation injury,' while in patients without inhalation injury, ARDS develops as a result of sepsis usually originating from wound infection. Respiratory injury worsens the prognosis in burn patients, independently of the severity of the burn injury. ARDS is characterized by a diffuse alveolar inflammation and increased pulmonary capillary permeability to fluids, whatever the cardiac pre-load volume may be.There are three phases of ARDS:' the exudative phase, the fibroproliferative phase, and the pulmonary fibrosis phase. The first of these phases is characterized by abnormal increases in pulmonary capillary permeability and the damage caused by oxidants and proteolytic enzymes liberated by inflammatory cells. In some patients, ARDS resolves itself at this point, while in others it continues until a fibroproliferative period which prolongs mechanical ventilatory dependence and its related complications, thus increasing mortality. It is known that basal membrane integrity and the capacity of Type H pneumocytes to replicate is fundamental to normal healing and the resolution of ARDS. However, the capacity to recover is dependent not only on the initial pulmonary lesion but also on subsequent and repetitive conditions such as those secondary to persistent endotoxaemia, nosocomial infection, the toxic effects of oxygen therapy, and barotrauma. These complications affect the balance between the reparative and the lytic processes.
The fibroproliferative phase is characterized by the proliferation of myofibroblasts and the deposition of collagen, transforming the initial exudative phase into granulation tissue. Unfortunately, this healing process is inefficient and causes malfunction of the ventilatory system owing to pulmonary restriction, which characterizes the fibrotic phase of ARDS.
After the initial injury, the body responds with the activation of an inflammatory cascade of the coagulation system, the immune system, tissue repair, and the activation of the hypothalamo-hypophyso-suprarenal system, producing glucocorticoids. The activation of this last system moderates the response with the result that it does not get worse than the initial injury. Various methods have been used to treat this inflammatory response, but without any satisfactory results. Glucocorticoids have been considered not to be beneficial, and they are potentially deleterious in septic shock and ARDS,"' especially in burn patients.However, recent studies have demonstrated the utility of glucocorticoids in the late stages of septic shock and ARDS."'
The objective of this report is to define the role of glucocorticoids during the intermediate, or fibroproliferative, phase of ARDS, citing the relevant literature.

Case report

We admitted a 26-year-old male patient with a direct flame burn and a prior history of schizophrenia. The burn affected 23% of the body surface area, of which 16% was partial thickness and 7% full thickness. There was no inhalation injury. The burns were located in the face, the anterior trunk and abdomen, the right arm, and the bilateral upper thighs. Silver sulphadiazine was applied to the wounds and reapplied every 12 h. On day post-burn 6, the patient was febrile and presented the clinical criteria of burn wound infection." Proteus mirabilis and P. aeruginosa, sensitive to ciprofioxacin and amikacin, were isolated. No other infectious loci were identified. The wounds were excised tangentially and covered with silver sulphadiazine dressings. On day 8 post-burn, the patient continued to be febrile and developed episodes of hypotension, accompanied by oliguria, tachycardia, tachypnoea, disorientation, and hallucinations. Two litres of crystalloid were rapidly infused without any response, and dopamine infusion was therefore initiated at a rate of 10 mcg/kg/min. Additionally, supplemental oxygen was initiated with aggressive respiratory therapy, because of a deterioration in gas exchange (OZ saturation fell to less than 90% of room air). Diffuse pulmonary infiltrates were noted on chest X-ray, and arterial blood gases were: pH, 7.43; PaC02, 29; Pa02, 83; bicarbonate, 22; 02 saturation, 96.9% on 35% Fi02 by mask. On day 9 post-burn, a second tangential excision of suspicious wound infections was performed. Twelve hours post-operatively, the patient was orotracheally intubated and mechanical ventilation (MV) was initiated following a deterioration of mental status and further oxygen desaturation. Initially, the patient required 100% Fi02 with a PEEP of 5 cm H20 to produce the following ABG: pH, 7.2; PaC02,46; Pa02, 49; BE, -10; 02 saturation, 73%. Additionally, the patient became haemodynamically unstable, with hypotension refractory to fluid expansion and little response to dopamine titration. Haemodynamic stability was achieved with epinephrine (0.9 mcg/kg/min). A Swan Ganz catheter was inserted, revealing a PAOP of 8 mm Hg and a cardiac output of 12 1/min. The following days the patient remained in a hyperdynamic state with a Pa02/Fi02 ratio between 150 and 200. On day 10 (MV day 2), fiberoptic bronchoscopy and bronchoalveolar lavage demonstrated a significant number of gram-positive cocci and gram-negative bacilli, for which reason the antibiotic regimen was altered to include vancomycin and imipenen, while the ongoing aminoglycoside therapy was continued. On day 13 (MV day 5), S. aureus and Acinetobacter were isolated and ampicillin-sulbactame was initiated on the basis of MIC sensitivities. Gas exchange improved (Pa02/Fi02 ratio, 294; PEEP, 10) and the epinephrine and dopamine drips were decreased to 0.57 mcg/kg/min and 2..7 mcg/kg/min, respectively. A fascial excision of approximately 10% BSA was performed on the trunk and abdomen due to doubts about tissue viability, and the wounds were occlusively covered with polymyxin B and polyurethane dressings. Post-operatively, the patient developed signs of disseminated intravascular coagulation (prothrombin time, 18; partial thromboplastin time, 300; fibrinogen, 120 mg/dl, fibrinogen degradation products, +++; platelet count 100,000/mm3).The patient was transfused with 4 units of fresh frozen plasma and 2 units of red blood cells. The bleeding of the burn wounds was not very important and was easily stanched. Low molecular weight heparin therapy (parnaparine 0.3 ml/12 h) was initiated and 2 units of fresh frozen plasma were given every day for 5 days. On day 16 (MV day 8), the patient experienced a further gas exchange deterioration (P:F ratio, 120; PEEP, 10) and new pulmonary infiltrates were observed by chest X-ray, and bronchoalveolar lavage was therefore performed. On day 18 (MV day 10), the BAL results revealed S. aureus and P. aeruginosa, and vancomycin was therefore continued with the addition of arbekacin and imipenem. Intermittent colistin nebulization was incorporated into the ventilatory circuit. Autografts were applied following determination of adequate excision and topical polymyxin B was continued on the grafted sites. On day 22 (MV day 14), PEEP was increased to 15-20 cm H20 following a deterioration in the gas exchange (Pa02/Fi02 ratio down to 116), and fiberoptic bronchoalveolar lavage was performed. Two days later (day 24), BAL cultures revealed 50-100 colonies of P. aeruginosa. Ultrasound of the abdomen and para-sinus X-rays revealed no occult infectious locus. Blood and urine cultures were negative. There was a 100% take of all autografts and closure of the interstices was observed. Ventilatory support continued with a PEEP of 15 and a P:F ratio of 124. The patient continued to be haemodynamically unstable, requiring epinephrine infusion at a rate of 0.7 mcg/kg. In response to vasopressor dependence, a cortisol test was performed to rule out suprarenal insufficiency. Basal cortisol levels were 25 mcg/dl, while after the ACTH test cortisol was 4!a mcg/dl. On day 25 (MV day 17), treatment with methylprednisone 200 mg q6h was initiated. On day 27 (methylprednisone day 3), the epinephrine infusion was discontinued. Sedative agents were discontinued on day 28 and pressure support ventilation initiated at 5 cm PEEP with a P:F ratio of 226. The patient was extubated on day 29 (MV day 21 and methylprednisone day 5). ABG on 50% Fi02 by mask demonstrated the following: pH, 7.49; PaC02, 44; Pa02, 62; bicarbonate, 3 1; saturation, 93%. The patient continued to require intermittent non-invasive ventilatory support for another four days, supported by aggressive pulmonary therapy. The methylprednisone dose was reduced by one half every day until it was discontinued on day 47. On day 55, the patient was transferred to a psychiatric institution without need of supplemental oxygen (ABG pH, 7.41; PaC02, 36; Pa02, 82; bicarbonate, 26; saturation, 96%) and with complete closure of the bum wounds.

Discussion

Respiratory insufficiency in burn patients is often related to inhalation injury. However, burn patients without inhalation injury can develop acute respiratory insufficiency (ARI).The causes of ARI in burn patients may be direct pulmonary injury as in the case of bronchoaspiration, pneumonia, inhalation injury, or pulmonary trauma. But A.RI can also be a manifestation of multi-organ dysfunction syndrome, in which the lung suffers indirectly? ARDS is the worst stage of respiratory dysfunction. The incidence of ARI and ARDS in burn patients with inhalation injury was found to be 73% and 20% respectively. In bum patients without inhalation injury the incidence of ARI and ARDS was 5% and 2% respectively.' Clinical management usually consists of support modalities that provide adequate tissue oxygenation, early infection control, and nutritional and ventilatory support, above all avoiding complications while pulmonary recovery is in progress. There appears to be agreement that the use of glucocorticoids should be avoided in the acute management of sepsis and ARDS." The administration of glucocorticoids did not appear to improve physiological parameters or mortality rates in ARDS of various aetiologies.' In all those studies, glucocorticoid administration was initiated during the initial stages of the illness. However, Jones" and Hakkinen2' demonstrated in experimental studies that the early administration of glucocorticoids produced an increased deposition of collagen in the pulmonary parenchyma. In contrast, the administration of glucocorticoids in the later stages of disease prevents excessive deposition of collagen in the lung. These findings were clinically confirmed by Meduri,'° who found that glucocorticoids were useful in intermediate states, i.e. during the fibroproliferative phase. Glucocorticoid treatment appears to limit the progression of fibrosis, reducing the deposition of collagen matrices, promoting their reabsorption, and facilitating endothelial and epithelial repair. This makes it possible to preserve and recuperate pulmonary structure and function.
Ashbaugh and Maier' described ten patients suffering from severe ARDS with fever, leukocytosis, and deterioration of respiratory status but with no evidence of infection in the first week after the onset of ARDS. Biopsies revealed marked cellular proliferation with obliteration of alveoli and fibrosis, and the condition was defined "idiopathic pulmonary fibrosis". Treatment was with high doses of glucocorticoids (500 mg methylprednisone/day) over a prolonged period (4-5 weeks), and there was a survival rate of 80%. Hooper and Kearl' described the benefits of glucocorticoid treatment in patients suffering from ARDS but with no evidence of infection. They use gallium scintography to confirm active pulmonary inflammation in nine out of the ten patients. They did not use biopsy confirmation in any patient. Both groups demonstrated that rapid discontinuation of glucocorticoids was associated with a deterioration in the respiratory parameters. This effect has also been described in experimental studies." Meduri" studied eight patients with late ARDS suffering from fever, progressive infiltrates, diffuse radioisotope captation (gallium scans), leukocytosis, purulent tracheal secretions, and ruled out pneumonia and extrapulmonary infection. The patients received methylprednisone in doses of 2-3 mg/kg/day and the survival rate in the group was 75%. In seven of the eight patients, lung biopsies confirmed the fibroproliferative phase of ARDS and the absence of pulmonary infection. We believe that the clinical characteristics of this state, together with the high sensibility of brochoalveolar lavage for ruling out infection,"" allows us to initiate glucocorticoid therapy without the need for histological confirmation, thus avoiding the hazards and risks of the pulmonary biopsy procedure. It is important to note that in this last study 50% of the patients with haemodynamic monitoring demonstrated low systemic vascular resistance. The patient we present did not have haemodynamic monitoring during this phase, but his dependence on vasopressor agents allows us to suppose with some degree of certainty that the patient persisted with low vascular resistance probably as a consequence of an inflammatory response. Years later, Meduri'° described the different patterns of response to glucocorticoid. Rapid responders were defined as those who improved at least 1 point on the lung injury scale' in the first week. Slow responders improved in the first 14 days, while nonresponders did not improve at all. Survival rates were 86% among rapid responders, 83% among slow responders, and 25% among non-responders. The patient we present was a a rapid responder to glucocorticoid therapy (Fig. 1).
Finally, in the first prospective randomized study of ;lucocorticoid treatment in patients in the fibroproliferative phase of ARDS, Meduri2' demonstrated improved gas exchange, decreases in mufti-organ dysfunction scoring, increases in rapid extubation, and a reduction in mortality in patients treated with methylprednisone compared with patients receiving conventional treatment. Vasopressor dependence to compensate for septic shock is another interesting point.21 Various studies have demonstrated increased mortality and vasopressor dependence in septic patients with suprarenal insufficiency." Others have reported patients without absolute suprarenal failure but with an insufficient ACTH response whose prognosis and response to glucocorticoids were various." In the case presented here, the patient did not present suprarenal failure, as demonstrated by the basal cortisol level. He did not present relative suprarenal failure either, and the basal cortisol level was therefore doubled in response to the ACTH challenge. However, the administration of glucocorticoids led to rapid removal of epinephrine (Fig. 1). The clinical use of glucocorticoids in septic shock has been practised for more than 30 years but there is no conclusive evidence of their benefit as no well-designed studies have been conducted."'Two recent meta-analyses demonstrated no benefit of the same, and noted that such treatment could be deleterious. However, some studies have used glucocorticoid treatment even in later stages of septic shock, with satisfactory results, noting an improvement in haemodynamics and a decrease in vasopressor treatment." In Briegel's study" patients who required catecholamiries for more than 48 h and received glucocorticoids reversed shock within 7 days in 68% of cases, while patients in the control group achieved this in only 21 % of cases. Mortality was 32% in the glucocorticoid group and 63% in the control group. This study explicitly excluded patients with suprarenal insufficiency and demonstrated that the response to glucocorticoids was similar in both responders and non-responders to ACTH. It is possible to demonstrate that glucocorticoids can be beneficial if used during the opportune period and that the response to glucocorticoid administration is independent of the suprarenal reserve. To our knowledge, this is the first report of successful treatment with glucocorticoids of a burn patient presenting septic shock and ARDS.

Conclusions

Both endogenous and exogenous glucocorticoids protect the body from an exaggerated inflammatory response, an event that can be more deleterious than the initial injury or trauma. It is clear that glucocorticoid therapy cannot be administered haphazardly in ARDS or septic shock conditions. However, glucocorticoids appear to be beneficial in the fibroproliferative stage of ARDS and septic shock with controlled infection, when the inflammation response is out of control. This is independent of the suprarenal reserve. In the patient we present the use of glucocorticoids facilitated rapid weaning off mechanical ventilation and vasopressor agents.

RESUME. Les Auteurs presentent le cas d'un patient male atteint de brulures daps 23% de la surface corporelle qui a developpe le syndrome de detresse respiratoire des adultes (SDRA) et le choc septique pendant I'evolution. Wine si 1'infection etait maintenue sous controle, 1'insuffisance respiratoire et les besoins d'agents vasocompresseurs ne se sont pas ameliores sans le traitement glucocorticoide. Ce type de traitement a permis d'interrompre la ventilation mecanique et de suspendre les agents vasocompresseurs pendant la premiere semaine du traitement. L'emploi des glucocorticoides dans les patients brines atteints d'SDRA ou de choc septique est controverse, mais cc traitement pent se reveler efficace dans la phase fibroproliferative de 1'SDRA et du choc septique si I'infection est sous controle, comme dans le cas presente.