Elevated AFP levels can be seen in dying HCC tumors (possibility given the hepatectomy), or in chronic liver disease. Did the patient have severe fibrosis or cirrhosis in addition to the HCC?

If you are worried about a metastasis that the scans did not find, try using other tumor markers in addition to AFP. Things like AFP-L3, des-gamma-carboxyprothrombin (DCP), etc. AFP is not necessarily diagnostic of HCC, only suggestive. The combination of other tumor markers can help to point in the direction of metastasis. 

I have included a few papers and a link below.

https://www.ncbi.nlm.nih.gov/pubmed/18087182

The most probably explanation is the existence of occult multiple functioning metastasis

Another potential cause of high AFP level, apart from those connected with liver diseases that were already mentioned, could be testicular tumor. This however would really be a bad luck to have two types of neoplasm at one time. But still you can check that. 

I can assume that the liver is not cirrhotic.

What was the preoperative AFP level? Did you measure other kind of tumor markers, such as CEA, CA19-9 or C125?

2) There was published article by Ezaki et al, Cancer 1988;61(9):1881-1883. Elevation of alpha-fetoprotein level without evidence of recurrence after hepatectomy for hepatocellular carcinoma. However, the postoperative AFP level was not high as in your case.

3) I had experienced several cases of macro-invisible HCC of the liver, but the AFP level was not high.

4) As Dr. Rodrigo's suggestion, may be there is an occult metastases other than the liver.

5) I had done a research on AFP producing cancer and found that it can occur from endoderm, mesoderm or ectoderm organs and tends to have liver metastasis. The pathological picture of the liver metastasis may be similar to HCC, therefore, it is also called as hepatoid cancer. Sometimes the metastatic lesion is difficult to be differentiated from primary HCC, unless the primary lesion is found.

6) The level of AFP depends on the AFP producing population which varies  in the hepatoid cancer, and not every cancer cell can secret AFP. Small cancer is possible to produce high level of AFP as well.

Any scans beside the pulmonary and abdominal scans? While those are the two most likely metastases, it seems the prevailing opinion of this thread is that there are occult metastases elsewhere in the body. 

Could you please do Bone scan or even PET scan if possible searching for mets other than lung and abdomen..

Thank you for all the answers.

PetCT scan will be performed soon, I will keep this post updates when we'll get the results.

PS:

Testicular tumor was excluded after urological examination. AFP lvl droped to 17.000, still a high value.

alll responces are rihgttt....but first thinkig  that is metastazis and residu tmmm

Should we have further more information?

1. what was the pre-operative AFP level?

2. what was the remnant liver volume?

We recently experienced a post-transplant rebound of AFP with occult metastasis, later revealed to be located in a retro-pancreatic lymph node. 

what is the confidence of PetCT   in the residue tm  ??? 

How was the bilirubin plasma level of the patient at the time of AFP increae? Very high level of AFP can be observed during drug induced hepatitis. Levoxacin can induce hepatitis, mainly cholestatic (low cytolisis with low ALT level) and the timing usually is  2-3 weeks after the administrtation.