SEPSIS-LIKE HEMODYNAMIC-CHANGES CAUSED BY INTERLEUKIN-2 INTERFERON-ALPHA IMMUNOTHERAPY IN PATIENTS WITH DISSEMINATED RENAL-CELL CARCINOMA -IMMUNOTHERAPY AS A CLINICAL-MODEL OF SEPSIS SIRS/

Human recombinant interleukin 2 (IL-2), alone or in combination with o ther cytokines, is currently under investigation for the immunotherapy of metastatic tumours. Objective responses of 20%-35% have been repor ted in patients with disseminated melanoma and renal cell carcinoma wh o received high-dose intravenous IL-2 in combination with interferon-a lpha (IFN alpha). However, treatment with IL-2 is complicated by a syn drome of life-threatening adverse reactions such as disseminated vascu lar leakage, fluid retention, severe hypotension, and (reversible) mul tiple organ dysfunction (MODS). A systemic inflammatory reaction (SIRS )/sepsis sepsis-like haemodynamic pattern has been described in patien ts after IL-2 bolus application alone. Our purpose was to study the ha emodynamic changes in patients treated with high-dose IL-2 administere d as a constant infusion and in combination with IFN alpha. Patients a nd Methods. Haemodynamic variables were obtained during therapy course s of 11 patients (aged 48 to 71 years, median 61) with metastatic rena l cell carcinoma receiving immunotherapy with IL-2/IFN alpha. Therapy consisted in (Fig. 1): IFN alpha 10 . 10(10) IU/m(2) body surface area (BSA) once daily on days 1-5 i.m. on a regular ward, followed by IL-2 as a constant infusion of 18 . 10(6) IU/m(2) BSA on days 6-11 in an i ntensive care unit (ICU). Haemodynamics were first measured after 5 da ys of IFN alpha application and transfer to the ICU on day 6, a furthe r 24 h after the beginning of IL-2 infusion (day 7), and at the end of the therapy course (days 10 and 11). Mean arterial pressure (MAP) was measured noninvasively using an oscillometric device (Dinamap(R), Cri tikon). Mixed-venous oxygen saturation (s<(nu)over bar> O-2) was measu red using an CO-oxymeter (OSM 3(R), Radiometer) and peripheral arteria l oxygen saturation (psaO(2)) was recorded continuously with a pulse o ximeter (Oxyshuttle(R), Critikon). In case of haemodynamic instability , stabilisation had priority over invasive haemodynamic measurements, so that nadir values of blood pressure (BP) did not influence mean MAP and are reported separately. Lactate values and criteria for SIRS wer e obtained before and during IL-2 infusion. Lactate measurements were performed using an enzymatic essay (Abbot FL(x)(R)). The mean effect s ize of the haemodynamic values, SIRS criteria, and lactate concentrati ons during IL-2 infusion (days 6-11) were calculated, and 95% confiden ce intervals for the effect sizes are indicated in Table 1. Results. A fter their daily i.m. injections of IFN alpha, patients had short epis odes of fever and tachycardia without significant drops in BP. A few h ours after transfer to the ICU and continuous infusion of IL-2, they d eveloped a syndrome of fever, tachycardia and tachypnoea. The haemodyn amic values after 5 days of IFN alpha therapy remained in the normal r ange, whereas those during IL-2 infusion strongly resembled SIRS and s epsis, with a decrease in MAP (98 to 82 mm Hg) and systemic vascular r esistance (SVR, 1477 to 805 dyn . s . cm(-5)) and an increase in cardi ac output (cardiac index 2.8 to 4.31 . min(-1) . m(-2)) (Fig. 2, Table 1). MAP often had to be stabilised with colloids during the last 48 h of therapy; 5 patients had nadir values below 60 mm Hg, or 30% below basic values in hypertensive patients. Catecholamine therapy became ma ndatory in 1 patient and therapy had to be discontinued. Surprisingly, some patients already had elevated plasma lactate concentrations afte r IFN alpha therapy. During IL-2 infusion mean plasma lactate levels i ncreased from 2.3 to 3.2 mmol . l(-1) and all patients had lactate con centrations above 2.0 mmol . l(-1) at the end of therapy (Fig. 3, Tabl e 1). During the last 48 to 72 h of IL-2 infusion, patients suffered f rom MODS with altered mental state (7 patients), oligoanuria (all pati ents), cardiac dysrhythmias (4 patients), congestive heart failure (1 patient, which led to a second case of therapy interruption), elevated bilirubin (4 patients), and pulmonary dysfunction. in 9 patients supp lementary oxygen was necessary when psaO(2) fell below 92%. Chest X-ra ys showed signs of pulmonary interstitial oedema. All patients develop ed significant generalised oedema due to a vascular leak syndrome, wit h fluid retention and weight gains of 6.3% during IL-2 infusion (Table s 1 and 2). Leukocyte counts dropped to 3670 mu l(-1) after 5 days of IFN alpha injection and rose to 9970 mu l(-1) at the and of 1L-2 infus ion. After discontinuation of IL-2 (day 11) the body temperature, hear t rate, BP, and criteria of impaired organ function rapidly returned t o the normal range, but leukocyte counts rose to 15360-mu l(-1) on day 12 (Table 1). Conclusion. High-dose IL-2 administered as a constant i nfusion and in combination with IFN alpha results in similar haemodyna mic changes to those seen during high-dose IL-2 bolus application alon e. The observed haemodynamic pattern strongly resembles SIRS and sepsi s. MODS, fluid retention, and increases in plasma lactate indicate mic rocircular disorders. Elevated levels of sepsis mediators such as tumo r necrosis factor and interleukin-1, activation of the complement casc ade, activated neutrophil granulocytes and endothelial cells have been reported in patients receiving high-dose IL-2 bolus treatment. Our re sults and data of other investigators lead us to conclude that not onl y the (macro)haemodynamic pattern of IL-2/IFN alpha therapy, but also its pathogenetic pathways parallel SIRS and sepsis. IL-2/IFN alpha imm unotherapy may therefore be used as a clinical ''model'' for sepsis re search.