Abstract
Background: Malignant infiltration (MI) is a very rare cause of acute liver failure (ALF), associated with very poor prognosis. Discrimination from ALF of indeterminate (IND) etiology may be difficult, but early recognition and diagnosis is crucial to avoid inadvertent liver transplantation (LT) and enable chemotherapy. We compared clinical, radiologic and laboratory characteristics of patients with ALF from IND and MI to identify key discriminatory features.
Patients and Methods: 90 patients with ALF admitted to a single specialist ICU were studied, 24 with ALF from first presentation biopsy-confirmed MI and 72 with non-malignant IND disease. Statistical comparison utilised non-parametric testing and receiver operating characteristic (ROC) techniques.
Results: In MI cases, median admission INR was (1.9 (1.6-2.8), Aspartate transaminanse (AST) 451 IU/l (231-2298) bilirubin 220 µMol/l (152-324) and peak encephalopathy grade 3 (2-3). 13/24 (54%) MI cases had hematologic malignancies; (8 lymphoma, 5 leukaemia), 10 carcinomas and 1 angiosarcoma. 3/24 (13%) MI patients survived with chemotherapy, 2 with lymphoma and 1 leukaemia. Time from admission to death in remaining patients was 4 (2-18) days. One 30 yr old patient underwent LT; adenocarcinoma was unexpectedly confirmed in the explant and she died 27 days post-operatively. Overall, survival in INT patients was 69%; 42 underwent LT; 41 (98%) survived.Compared to INT cases, MI Patients were older (52 yrs (36-59) vs. 38 (30-54), p<0.03) Bilirubin, AST or creatinine levels did not differ but MI cases had higher Alkaline Phospatase (ALP) (260 IU/l (163-643) vs.147 (102-220). P<0.001), Lactate dehydrogenase (LDH) (1652 IU/l (754-5095) vs.384 (240-522), p<0.001), urea (11.3 mMol/l (8.4-23) vs. 5.3 (2.7-9.9), p<0.001) and Ferritin (6068 µg/l (1610-31515) vs. 1600 (476-5882), p<0.01) and lower platelet count (46 10E9/l (26-89) vs. 159 (105-244), p<0.001). Urea/Creatinine ratio was higher in MI cases (52.3 (38-63) vs. 21.1 (13-33), p<0.001). Imaging-confirmed hepatomegaly was present in 18/24 (75%) MI but only 6/70 (9%) INF cases, (p<0.001). 15 MI patients underwent bone marrow biopsy; 10/15 (67%) biopsies confirmed the haematologic malignancy. On ROC analysis, AUROC for Urea/Creatinine ratio was 0.890, Platelet count 0.831, LDH 0.817 and ALP 0.751.
Conclusion: MI is a rare cause of ALF and associated with high mortality but is not uniformly fatal. More than half of cases resulted from haematologic disease and many could be diagnosed with BM biopsy. Simple clinical and biochemical features are strong indicators of the presence of malignant disease and can aid in earlier diagnosis and treatment.
Patients and Methods: 90 patients with ALF admitted to a single specialist ICU were studied, 24 with ALF from first presentation biopsy-confirmed MI and 72 with non-malignant IND disease. Statistical comparison utilised non-parametric testing and receiver operating characteristic (ROC) techniques.
Results: In MI cases, median admission INR was (1.9 (1.6-2.8), Aspartate transaminanse (AST) 451 IU/l (231-2298) bilirubin 220 µMol/l (152-324) and peak encephalopathy grade 3 (2-3). 13/24 (54%) MI cases had hematologic malignancies; (8 lymphoma, 5 leukaemia), 10 carcinomas and 1 angiosarcoma. 3/24 (13%) MI patients survived with chemotherapy, 2 with lymphoma and 1 leukaemia. Time from admission to death in remaining patients was 4 (2-18) days. One 30 yr old patient underwent LT; adenocarcinoma was unexpectedly confirmed in the explant and she died 27 days post-operatively. Overall, survival in INT patients was 69%; 42 underwent LT; 41 (98%) survived.Compared to INT cases, MI Patients were older (52 yrs (36-59) vs. 38 (30-54), p<0.03) Bilirubin, AST or creatinine levels did not differ but MI cases had higher Alkaline Phospatase (ALP) (260 IU/l (163-643) vs.147 (102-220). P<0.001), Lactate dehydrogenase (LDH) (1652 IU/l (754-5095) vs.384 (240-522), p<0.001), urea (11.3 mMol/l (8.4-23) vs. 5.3 (2.7-9.9), p<0.001) and Ferritin (6068 µg/l (1610-31515) vs. 1600 (476-5882), p<0.01) and lower platelet count (46 10E9/l (26-89) vs. 159 (105-244), p<0.001). Urea/Creatinine ratio was higher in MI cases (52.3 (38-63) vs. 21.1 (13-33), p<0.001). Imaging-confirmed hepatomegaly was present in 18/24 (75%) MI but only 6/70 (9%) INF cases, (p<0.001). 15 MI patients underwent bone marrow biopsy; 10/15 (67%) biopsies confirmed the haematologic malignancy. On ROC analysis, AUROC for Urea/Creatinine ratio was 0.890, Platelet count 0.831, LDH 0.817 and ALP 0.751.
Conclusion: MI is a rare cause of ALF and associated with high mortality but is not uniformly fatal. More than half of cases resulted from haematologic disease and many could be diagnosed with BM biopsy. Simple clinical and biochemical features are strong indicators of the presence of malignant disease and can aid in earlier diagnosis and treatment.
| Original language | English |
|---|---|
| Article number | 309 |
| Pages (from-to) | 360A-360A |
| Number of pages | 1 |
| Journal | Hepatology |
| Volume | 58 |
| Publication status | Published - Oct 2013 |
| Event | 64th Annual Meeting and Postgraduate Course of the American-Association-for-the-Study-of-Liver-Diseases - Washington, United Kingdom Duration: 1 Nov 2013 → 5 Nov 2013 |
Keywords
- Acute liver failure
- Hepatic Malignant Infiltration