More Evidence for Centralization of Care for Hepatocellular Carcinoma

I n Western countries, hepatocellular carcinoma (HCC) most commonly presents in the setting of chronic liver disease or cirrhosis.Improvements in screening along with better treatment regimens have translated into improved outcomes, evidenced by recent population-based studies. 1,2owever, a substantial portion of patients presenting with HCC are from racial/ethnic minorities and/or of lower socioeconomic status, who may not benefit from the aforementioned improvements. 3n general, these patients are frequently diagnosed with more advanced cancers, not limited to HCC. 4 Possible explanations for this phenomenon include variations in access to care and screening; historic, justified mistrust of the health care system 5 ; and quality concerns in lower-volume hospitals due to limited exposure to rare diseases. 6Non-White individuals, Medicaid-insured patients, and those who are uninsured or underinsured are also less likely to receive care at high-volume (HV) referral hospitals, which may influence long-term oncologic outcomes. 7n the hepatopancreatobiliary surgical oncology literature, there is a well-established association between higher surgical volume and improved perioperative outcomes across pancreas 8 and liver 9 cancer resections.It also has been previously established that hospital volume in general correlates with improved clinical outcomes in these cancers (pancreatic cancer, 10 gallbladder cancer, 11 and HCC 12,13 ).In this issue of JNCCN , Whitham et al 14 examined the comparative effects of safety-net hospital (SNH) status and hospital volume on clinical outcomes and first oncologic treatment choice in patients with newly diagnosed HCC.SNHs are defined by the Institute of Medicine as facilities that, by mission, provide care to a substantial portion of vulnerable patients, regardless of their ability to pay.These hospitals will on average have higher proportions of patients who are uninsured or on Medicaid, receive public assistance, and are typically located in metropolitan areas of poverty.The authors hypothesized that disparities in cancer care experienced by patients treated at SNHs could be mitigated by overall hospital volume.To test their hypothesis, the authors queried the California and Texas statewide cancer registries from 2004 to 2019 to identify patients with newly diagnosed HCC. 14 CMS files were used to identify SNH status by calculating the proportion of Medicaid and low-income Medicare patients out of the total number of patients treated at each individual facility.Patients were categorized according to SNH status, and a "volume designation" was assigned to the reporting hospital where the diagnosis was rendered or a decision for first course of treatment was made.The investigators accomplished this by using an outcome-oriented statistical approach involving a series of Cox regression models that incorporated mortality risk to define a cutoff of 21 cases, distinguishing low-volume (LV) and HV hospitals.Of the 535 hospitals, 18% were LV SNHs (LV/SNHs), 76% were LV non-SNHs (LV/non-SNHs), 2% were HV SNHs (HV/SNHs), and 4% were HV non-SNHs (HV/non-SNHs)-highlighting the small number of HV, primarily academic referral centers.Teaching status and medical oncology, radiation oncology, and palliative care services were more frequently available in HV hospitals regardless of SNH status, whereas HV/non-SNHs were the most likely to have liver transplant services.When the 535 hospitals were categorized into quartiles by HCC volume, the annual cases declined to 4.75 per year even in the top quartile.Of the 50,167 patients, 12% were treated at LV/SNHs, 36% at LV/non-SNHs, 13% at HV/SNHs, and 38% at HV/non-SNHs.SNHs were more likely to care for Hispanic and uninsured or underinsured patients.Regardless of SNH status, LV hospitals encountered a higher proportion of patients with advanced disease.The proportion of patients with HCC who were not recommended for treatment was lower at HV and non-SNHs: 44% at LV/SNHs, 36% at LV/non-SNHs, 23% at HV/SNHs, and 17% at HV/non-SNHs.Similar trends were observed when tumor stage was categorized as localized, regional, or metastatic, with patients treated at LV/SNHs (39%-50%) having the highest odds of receiving no treatment, and those treated at HV/non-SNHs (13%-31%) having the highest odds of receiving treatment across all stages of disease.Of the patients who received any treatment, chemotherapy was the most common