Alfa feto protein or AFP

Alfa feto protein or AFP

1. Alpha-fetoprotein  

2. Alpha-fetoprotein   —  Alpha-fetoprotein is a glycoprotein that is normally produced during gestation by the fetal liver and yolk sac, the serum concentration of which is often elevated in patients with HCC. Serum levels of AFP do not correlate well with other clinical features of HCC such as size, stage, or prognosis. Elevated serum AFP occurs in pregnancy ( figure 1 ), with tumors of gonadal origin (both germ cell and non-germ cell [ 56 ]) and in a variety of other malignancies, of which gastric cancer is the most common [ 57 ].

3. Elevated serum AFP may also be seen in patients with chronic liver disease without HCC such as acute or chronic viral hepatitis [ 58,59 ]. AFP may be slightly higher in patients with cirrhosis due to hepatitis C [ 60 ]. In one report, serum AFP decreased significantly in patients with cirrhosis from hepatitis C, treated with peginterferon plus  ribavirin  [ 60 ].

4. A rise in serum AFP in a patient with cirrhosis or hepatitis B should raise concern that HCC has developed. It is generally accepted that serum levels greater than 500 mcg/L (normal in most laboratories is between 10 and 20 mcg/L) in a high-risk patient is diagnostic of HCC [ 61 ]. However, HCC is often diagnosed at a lower AFP level in patients undergoing screening [ 58,62 ].

5. Not all tumors secrete AFP, and serum concentrations are normal in up to 40 percent of small HCCs [ 63 ]. Furthermore, an elevated AFP may be more likely in patients with HCC due to viral hepatitis compared with alcoholic liver disease [ 64 ].

6. In a study of 357 patients with hepatitis C and without HCC, 23 percent had an AFP >10.0 mcg/L [ 65 ]. Elevated levels were associated with the presence of stage III or IV fibrosis, an elevated international normalized ratio, and an elevated serum aspartate aminotransferase level.

7. AFP levels are normal in the majority of patients with fibrolamellar carcinoma, a variant of HCC ( picture 1 ) [ 66 ]. (See  "Pathology of malignant liver tumors", section on ' Fibrolamellar carcinoma ' .)

8. Patients with cirrhosis and persistently elevated AFP values have an increased risk of developing HCC compared with those who have fluctuating or normal levels (29 versus 13 and 2.4 percent, respectively, in one report) [ 67 ].

9. The sensitivity, specificity, and predictive value for the serum AFP in the diagnosis of HCC depends upon the characteristics of the population under study, the cutoff value chosen for establishing the diagnosis, and the gold-standard used to confirm the diagnosis. A number of studies have described test characteristics in different settings [ 68,69 ]. The following estimates were based upon a cutoff value of >20 mcg/L in a systematic review that included five studies [ 70 ]:

10. ●Sensitivity 41 to 65 percent ●Specificity 80 to 94 percent ●Positive likelihood ratio 3.1 to 6.8 ●Negative likelihood ratio 0.4 to 0.6 Three examples (one of which was included in the above study) illustrate the range of findings in the individual reports:

11. One study included 1069 patients who were chronic carriers of hepatitis B who underwent regular screening with a serum AFP [ 68 ]. Serum AFP was greater than 20 mcg/L in 9 of 14 patients with HCC compared with 91 of 964 patients without the tumor, giving an overall sensitivity and specificity of 64 and 91 percent, respectively. However, only 9 of the 100 patients with AFP values greater than 20 mcg/L had HCC (positive predictive value 9 percent) even in this population of patients with a high prevalence of HCC ( table 1 ).

12. In a similar study of 1531 patients with hepatitis B who were receiving  entecavir  and undergoing screening, 57 patients developed HCC [ 71 ]. An AFP value ≥20 mcg/L had a sensitivity and specificity for HCC of 33 and 98 percent, respectively (positive predictive value of 45 percent, negative predictive value of 98 percent). If the cutoff was dropped to ≥6 mcg/L, the values for sensitivity and specificity were 79 and 80 percent, respectively (positive predictive value of 12 percent, negative predictive value of 99 percent).

13. A case-control study evaluated the diagnostic characteristics of the serum AFP in screening for HCC in patients with different types of chronic liver disease. The following sensitivities and specificities were observed [ 69 ]: ●AFP cutoff 16 mcg/L (sensitivity 62, specificity 89 percent) ●AFP cutoff 20 mcg/L (sensitivity 60, specificity 91 percent) ●AFP cutoff 100 mcg/L (sensitivity 31, specificity 99 percent) ●AFP cutoff 200 mcg/L (sensitivity 22, specificity 99 percent )

14. At a prevalence of HCC of 5 percent, a serum AFP of ≥20 mcg/L (the cutoff value the authors considered to be best) had a positive and negative predictive value of 25 and 98 percent, respectively. At a prevalence of 20 percent, these numbers were 61 and 90 percent, respectively. The low positive predictive values observed in these studies [ 68,69 ] (and the test characteristics described in the systematic review above [ 70 ]) underscore the limitation of using the serum AFP as a screening test for HCC.

15. Despite the issues inherent in using AFP for the diagnosis of HCC, it has emerged as an important prognostic marker, especially in patients undergoing resection and those being considered for liver transplantation. Patients with AFP levels >1000 mcg/L have an extremely high risk of recurrent disease following transplantation, irrespective of the tumor size