To the Editor:
We read the study by Dam et al. with great interest. They compared the cerebral oxygen metabolism (CMRO2), cerebral blood flow (CBF), and metabolic rate of blood ammonia (CMRA) in patients with cirrhosis during and after recovery from hepatic encephalopathy (HE) and concluded that the changes in CMRO2 and CBF could not link to ammonia concentration or CMRA. Their findings are of great potential for clinical applications; however, we have some concerns.
First, the authors reported that CBF increased significantly and arterial ammonia concentration decreased markedly after recovery from HE, which tended to show a negative correlation between them. A 14-month arterial-spin labeling (ASL) magnetic resonance imaging (MRI) follow-up study in one of our patients with cirrhosis after transjugular intrahepatic portosystemic shunt (TIPS) showed a recovery process similar to the one reported by Dam et al. and indicated the venous blood ammonia correlated slightly with CBF (r = −0.86, P = 0.061). The patient had no sign of HE, and his global CBF was 66.29 mL·min−1·100 g−1 before TIPS. Five days after TIPS insertion, he showed no sign of HE and his CBF decreased to 55.51 mL·min−1·100 g−1. Ninety-seven days later, the patient had three episodes of acute HE, and his CBF decreased to 33.58 mL·min−1·100 g−1, the lowest in the 14-month follow-up (Fig. 1A). About 4 months after HE, he was free of HE after treatments, and his CBF recovered to 61.20 mL·min−1·100 g−1. The patient's venous blood ammonia level reached a peak value of 65 mL/L during HE (Fig. 1B), indicating that ammonia correlated negatively with the development of HE. Thus, contrary to the authors' conclusion, we suggest that CBF changes might be associated with ammonia level.
Second, CMRA could be saturated during and after HE. Although the concentration of venous ammonia is always lower than that of arterial blood, it has the same positive correlation with HE grade as arterial ammonia. If we use venous ammonia to approximate arterial ammonia in patients with cirrhosis, the estimated ammonia delivery can be calculated by the product of venous ammonia and CBF. In our case, the estimated ammonia delivery increased from approximately 0.9 μmol·min−1·100 g−1 before HE to 2.18 μmol·min−1·100 g−1 during HE, decreased slightly to 1.84 μmol·min−1·100 g−1 4 months after recovery from HE, and dropped to 1.43 μmol·min−1·100 g−1 1 year after recovery from HE (Fig. 1C). Because the estimated ammonia delivery remained at a high level after recovery, it is possible that CMRA was still at a high level to detoxify ammonia as much as possible. We suggest that CMRA before HE should be included to show the relationships among CMRA and HE.
Third, 1,000 MBq 15O-oxygen, 500 MBq 15O-water, and 700 MBq 13N-ammonia and low-dose computed tomography were performed in Dam et al.'s study, delivering a high radiation dose to the patients. Other imaging modalities without radiation dose, such as MRI including ASL, T2-Relaxation-Under-Spin-Tagging, and phase-based oxygen metabolism MRI, should be performed to replace (at least partly) invasive nuclear medicine imaging techniques in longitudinal studies for patients with cirrhosis.
In conclusion, venous blood ammonia level could be related to changes in CBF. A longitudinal MRI study is the preferred modality to show the relationship between ammonia level, CBF, CMRO2, and CMRA.
Long Jiang Zhang1
Guang Ming Lu1
1Department of Medical Imaging, Jinling Hospital, Clinical School of Medical College, Nanjing University, Nanjing, China
2College of Civil Aviation, Nanjing University of Aeronautics and Astronautics, Nanjing, China
3Kreiger School of Arts and Science, Johns Hopkins University, Baltimore, MD