The lower proportion of HIV patients receiving surgical treatment in China
There are six related studies on the treatment of HIV+ HCC patients from the United States, France, Italy and other European countries. Overall, the rate of curative and TACE treatments was lower in HIV+ than HIV− patients (Fig. 2). Currently, the above reports on HIV+ HCC patients are mainly concentrated in Europe or America, where HIV treatment is relatively mature, while there are no reports on HIV+ HCC patients in Asia and Africa. We are the largest center for HIV patients in Southwest China. Through nearly 20 years of clinical treatment, nearly 2,000 patients with HIV have been treated annually. Approximately 3,000 of these have been treated surgically, of whom over 800 were treated in 2018. However, only 30 patients with HIV+ have been diagnosed with HCC, including 21 patients with stage BCLC-B, and 3 patients with stage BCLC-A. Pinato et al. published the largest HIV+ clinical study in 2018, involving 387 patients, of whom 33% were BCLC-A and 18% BCLB-B11. The reasons for the high rate of BCLC-B patients in our hospital were that most of them received no surgical treatment before they were transferred to our hospital; and that during this period, their staging also progressed from A to B. Although the current law of our country stipulates that HIV patients should enjoy an equal right to medical treatment, in fact HIV patients are mainly concentrated in specialized infectious disease hospitals. These are deficient in surgical technology compared with large general hospitals. Surgical resection and liver transplantation are the main treatment methods for hepatocellular carcinoma and have a good curative effect, but due to the particularity of HIV, the implementation of treatment methods is often inadequate.
Occurrence and development of HCC in HIV+ patients
The shorter onset time of HCC in HIV patients is still the focus of research and discussion12. Because of the shared route of transmission and the immunosuppressive characteristics of HIV, HIV+ patients are more likely to develop HBV or HCV infection13,14,15. Our data show that the median age of HIV infected patients at first diagnosis of HCC in our hospital was 50.24 + 12.10 years, while that of the overall HIV− population was 54.07 ± 13.04 years (P = 0.181). The overall patients we included, including non-HBV HCC patients, were no younger than HIV HCC patients. This difference may be due to the first time of HIV exposure. Puoti et al.12 reported that the possible age of exposure is during patients’ twenties while our center’s HIV+ patients, the first diagnosis of HIV antibody positivity is 48.07 ± 12.03 years, and possible exposure age was 44.57 ± 10.99 years.
In addition, the weakening of immune response caused by HIV will increase the viral load and accelerate the progression of hepatitis cirrhosis and hepatocellular carcinoma through the carcinogenic characteristics of HBV or HCV themselves. Since anti-hepatitis B drugs have a significant impact on viral load, we chose HIV+ HBV co-infected HCC patients who did not receive HBV antiviral therapy at the time of consultation, to compare with HIV− HBV HCC patients. Our results showed that there was no significant difference in viral load and onset time between HIV HBV co-infected HCC patients and HIV− HBV infected HCC patients. (P = 0.612).
However, due to HIV-induced immunosuppression, the reduced immune response to HBV or HCV may actually reduce liver inflammation and injury, thereby slowing the progression of cirrhosis and cancer. Meta-analysis16 showed that Fibrosis-4 index (FIB-4) was a more accurate non-invasive method for evaluating the degree of fibrosis. Because biopsy results cannot be obtained at the TACE procedure itself, we chose FIB-4 to reflect the degree of liver fibrosis. Our data showed that the degree of hepatic fibrosis in 21 HIV+ patients was 6.06 ± 5.00 and in 1443 non-HIV patients was 5.14 ± 6.03 (P = 0.486, 95% CI -1.67 to 3.51).
Viral load, hepatic fibrosis index and onset time did not show a significant impact on HIV positive patients. However, the degree of liver fibrosis has not been determined at biopsy, too few cases of viral load were measured and the short interval between HIV exposure time and HCC detection time may have some impact on these results. We suggest that a prospective multi-center study be undertaken specifically to validate these indicators.
Prognostic risk factors in patients with HCC
It is still controversial whether HIV combined with hepatitis virus may increase the risk of HCC occurrence13,17,18; however, what can be confirmed is that HAART based on survival analysis is currently recognized as an independent protective factor for HIV co-infected patients19,20,21,22. Since the introduction of HAART therapy, there have been several reports from HIV+ and HIV− clinical control studies10,12,23,24,25,26. The current consensus is: 1. BCLC grade is the main risk factor affecting prognosis of HIV+ HCC patients; 2. The prognosis of untreated patients is significantly worse than that of patients receiving treatment. Our univariate and multivariate analyses showed that HIV+ was not a prognostic risk factor for BCLC-B patients. For all patients of BCLC-B stage, our data show that TACE treatment has the most significant protective significance. In addition, combined RR values, AFP and Child–Pugh Score are the most important risk factors. This may be due to that BCLC-B staging limited liver function, size and number of tumors.
The largest single-arm study of 387 HIV+ HCC patients in a multi-center trial showed that CD4 is a potential prognostic factor for the survival of HIV+ HCC patients11. In order to study some specific indicators of HIV infection, we conducted single and multi-factor analyses of HIV patients to assess the significance of varying levels of CD3, CD4 and CD8. Our central data show that the levels of CD3, CD4, and CD8 were not considered to be a prognostic factor. However, three of the eight patients whose CD4 was below 200/μL had opportunistic bacterial infections, and the CD4 level also remained low, after surgery. None of the 13 patients whose CD4 was greater than 200/μL had opportunistic bacterial infections (P = 0.042). The HIV RNA is also considered as a potential independent predictor. The three patients with opportunistic bacterial infection had HIV replication > 5 × 102, 2.11 × 103, 1.91 × 106, while 5 patients in the other 18 patients without opportunistic bacterial infection had HIV RNA levels > 5 × 102 before operation (P = 0.274). From the statistical results of P-value, the prediction value of CD4 for postoperative opportunistic infection was better than that of HIV RNA. Unfortunately, due to the limited number of cases, we were unable to conduct a control study to guide whether CD4 count or HIV RNA is more predictive for the postoperative opportunities. Interestingly, we found that Plt count and AST are protective factors for HIV patients, but RR is of less significance; this may reflect that there were only 21 samples, and we look forward to a larger control study to verify this.
Overall, the AFP, Child–Pugh score, and tumor size may increase with the progression of the disease in both the general population and the independent HIV population. Therefore, TACE should be performed as early as possible for BCLC-B HCC patients, including HIV-positive patients, to improve prognosis.
Survival time of HIV patients with HCC
The prognosis of HIV+ and HIV− patients is still inconsistent. The 5-year survival rate is generally believed to be worse for HIV+ than for HIV−25. Gelu-Simeon et al.10 is currently the largest multicenter study, showing that the 1-and 2-year survival rates of HIV+ patients are significantly worse than that of HIV− patients. Pinato et al. mainly studied the survival of HIV positive patients with different albumin levels. There was no HIV negative patients as control group, so we did not include their research in meta-analysis11. From Fig. 2, we can see that only 2 of the 6 studies have significant poor prognosis in patients with HIV+. In addition, we found that the proportion of HIV− patients receiving TACE treatment in one of the above two studies is significantly higher than that in HIV+ patients (71% vs. 51%, P = 0.007), while another one is lack of relevant data. Considering the low proportion of HIV+ patients receiving curative treatment, Lim et al.24 proposed a 1:2 matching method to control the deviation caused by the proportion of treatment modalities. This report had the smallest difference in the composition ratio between HIV+ group and HIV− treatment mode (73.9% vs.69.6%, P = 0.704). As a result, there was no statistical difference in the overall survival rate. In the study of Berretta et al.25 the curative or TACE treatment of HIV+ patients accounted for 75% and of the HIV− group accounted for 63%; the overall survival rate of the HIV+ group was poor, but there was no significant difference in survival rates at 1 and 2 years. In other studies where the ratio of curative and TACE treatment was slightly different between HIV+ and HIV− groups, there was no statistical difference in 1 or 2-year prognosis.
Overall, through meta-analysis of previous studies, the survival rate of HIV+ patients is still lower than that of HIV− patients. Previous studies show that BCLC-A patients account for a higher proportion of HIV+ than HIV− patients, while HIV positive patients receive a lower proportion of the treatment recommended by BCLC. As a result, this may also be the cause of the poor prognosis of HIV+ patients. Based on the shortcomings of previous studies, we hope to be able to strictly match HIV+ and HIV—patients with the data of our center over 16 years. Matching will strictly control the selection of patients in a single stage, i.e. BCLC-B stage, and treatment guidelines, i.e. recommended treatment TACE, as well as preoperative conditions. Since the median survival rate of BCLC-B patients is about 2.5 years, it would be more important to ensure the 1 and 2-year survival rate of these patients. Our data show that there is no significant difference in 1-year and 2-year survival rates between HIV+ and HIV− patients, and they are much higher than those without surgical treatment. In conclusion, we believe that TACE treatment should be actively carried out for HIV+ patients in the BCLC-B stage.
Prospects and limitations
HIV therapy is still developing in China. In order to create a mature medical system for HIV-HCC patients and to realize equal surgical opportunities for HIV and ordinary patients, policies such as HIV surgical treatment still need to be improved at the national level, followed by appropriate education for doctors and patients. The retrospective design should be acknowledged as a limitation to our work. Because of the retrospective study, it is difficult to avoid the selective bias, even when using a case–control study design based on propensity score matching. Furthermore, the case–control matching may lead to other selective biases due to unmeasured patient characteristics which may have influenced outcomes. There were only 21 HIV+ patients undergoing TACE in the BCLC-B stage; however, this study is—to our knowledge—the largest case control study of survival outcomes of BCLC-B stage HIV patients who underwent TACE.