Earliest case reports and systematic published studies supporting the notions that “cryptogenic cirrhosis” equals “NAFLD-cirrhosis” in most cases; and that PLCs, hepatocellular carcinoma (HCC), and cholangiocarcinoma (CC) may indeed occur as a complication of NAFLD and related metabolic disorders

Author, year [Ref]MethodFindingsConclusion
Caldwell et al. [8], 1999Findings from 70 consecutive CryptoCir probands reassessed for alcohol consumption, evaluated by the IAH score and for risks of viral hepatitis and NASH were compared to 50 consecutive NASH patients, 39 nonalcoholic patients with HCV cirrhosis, and 33 with cirrhosis owing to PBCThe CryptoCir group, that comprised 70% of women in their 60s, had a prevalence of T2D and obesity significantly higher than those with cirrhosis owing to either PBC or HCV. Conversely, the prevalence of obesity and T2D was similar to the NASH patients who were, on average, 10 years youngerData suggest that NASH is an under-recognized cause in many CryptoCir patients, most of whom are older ladies with T2D and obesity
Zen et al. [9], 2001A 58-year-old lady who did not drink alcohol and was negative for all serological markers of HBV and HCV infection received a diagnosis of T2D, treated with insulin therapy. Four years later, liver biopsy (performed to investigate altered liver tests) was compatible with NASHIn the follow-up, the patient was re-biopsied for multifocal hepatopathy and 3 out of the 4 liver nodules were moderately differentiated HCC (10 years after the diagnosis of NASH), well-differentiated HCC (11 years later) and dysplastic nodule (11 years later)This case study is the first proof-of-concept published anecdotal evidence that HCC may develop as a late NASH complication
Bugianesi et al. [10], 2002Twenty-three out of forty-four CryptoCir patients retrospectively identified among 641 cirrhosis-associated HCCs were actively followed up and compared to viral- and alcohol-associated HCCThe prevalence of obesity and T2D was significantly higher in patients with CC, who also had higher glucose, cholesterol, triglyceridemia, and IR; aminotransferase levels were lower. Iron status and prevalence of mutations in the HFE gene did not differ. At LRA hypertriglyceridemia, T2D, and normal aminotransferases were independently associated with HCC arising in CryptoCirCharacteristics compatible with NASH are more common in HCC arising in patients with CryptoCir than in age- and sex-matched HCC cases owing to viral or alcoholic etiology suggesting that HCC may occur as a late complication of NASH-cirrhosis
Marrero et al. [11], 2002Among 105 consecutive HCC patients, the most common etiologies of CLD were HCV and CC (51% and 29%, respectively). Half of the CryptoCir patients exhibited either histologic or clinical features associated with NAFLD. In 50% of cases, HCC was diagnosed during surveillance (group I); in the remaining patients, HCC was symptomatic (group II)Group I patients had smaller cancers (P = 0.01), were more likely to be eligible for surgical treatment (P = 0.005) and had higher survival rates than group II patients (P = 0.001). CC patients were less likely to have been submitted to surveillance for HCC and, accordingly, were diagnosed larger tumor burdensHCV and CryptoCir were the most common etiologies of HCC. NAFLD accounted for at least 13% of the cases
Michelini et al. [12], 2007The authors speculated that IR might be a risk factor also for CC, like other cancer typesTo illustrate their speculation, these authors reported on 3 ICC cases that exhibited clinical manifestations of IR such as obesity, dyslipidemia, and NAFLD as the common grounds predisposing to CCThis case study is the first proof-of-concept published anecdotal evidence that conditions belonging to the domain of MetS may be the only biologically plausible risk factor in a fraction of CC cases
Welzel et al. [13], 2007The SEERM database was utilized to evaluate comorbid conditions of 535 ICC patients, 549 ECC patients, and 102,782 cancer-free controls. Data were analyzed with LRAFurther to established risk factors, several endocrine and metabolic comorbidities were strongly associated with both ECC and ICC: cholelithiasis (P < 0.001), diabetes (P < 0.001), thyrotoxicosis (ECC, P = 0.006; ICC, P = 0.04). Conditions associated with ICC alone included obesity (P = 0.01) and NAFLD (P = 0.02)This pioneering study identified several novel metabolic risk factors for ECC and ICC

CLD: chronic liver disease; CryptoCir: cryptogenic cirrhosis; ECC: extrahepatic cholangiocarcinoma; HBV: hepatitis B virus; HCV: hepatitis C virus; IAH: International Autoimmune Hepatitis; ICC: intrahepatic cholangiocarcinoma; IR: insulin resistance; LRA: logistic regression analysis; MetS: metabolic syndrome; PBC: primary biliary cholangitis; SEERM: Surveillance, Epidemiology and End Results-Medicare; T2D: type 2 diabetes