2017 Plain English Summaries
Fibrolamellar hepatocellular carcinoma (FLHCC) is a rare cancer that begins in the liver and is known to metastasize, or spread, to nearby organs such as the stomach, diaphragm, and pancreas, as well as metastasis to lymph nodes. Cases of further metastasis to the lung, peritoneum, bone, and adrenal gland has been noted in 30% of patients. Only two previously reported cases of metastasis from FLHCC to the brain existed in the literature before this current paper. This paper details three cases of FLHCC patients with brain metastasis. The authors confirmed the presence of the DNAJB1-PRKACA chimeric RNA transcript found in FLHCC in the brain metastases in all three patients. While these three patients did not share an identical timeline or pattern of metastasis, they all had metastasis to the lung, which is only seen in 30% of FLHCC patients. The authors suggest that given this information and since there currently exists no prescribed recommendations for monitoring brain metastasis in FLHCC patients, patients with advanced stage FLHCC, especially those who have lung metastasis, also be monitored for brain metastasis to help find these tumors before patients exhibit symptoms.
Carney complex is an autosomal dominant disorder, meaning that of the two copies of a particular gene that you have, only one of them needs to have a mutation to cause the disorder. Carney complex results in myxomas, a type of connective tissue tumor, of the heart and skin, as well as irregular pigmentation of the skin and dysregulation of the endocrine system. Endocrine tumors and non-endocrine tumors have also been found in patients with Carney complex.
The most common mutation in DNA that causes Carney complex is a mutation in PRKAR1A that makes it no longer functional. This gene normally leads to a protein that regulates protein kinase A (PKA, sometimes noted as PRKACA when referring to the active unit). PKA is what is called a holoenzyme or heterotetramer, meaning it is a complex of four different proteins. Two of the proteins are regulatory proteins, like PRKAR1A, and two of the proteins are the catalytic, or active, proteins, like PRKACA. It is thought that when the protein complex is intact, it is not active, but when PRKAR1A releases from the complex, it frees the catalytic proteins to act in the cell.
The authors of this paper are in favor of the hypothesis that fibrolamellar hepatocellular carcinoma (FLC) is caused by over-activation of the protein kinase A activity, since tumor samples from these patients commonly have the same break in one copy of chromosome 19 that leads to the fusion of part of a heat shock protein, DNAJB1, and the majority of PRKACA, including the catalytic site, in addition to normal PRKACA that is still intact on the other copy of chromosome 19. Based on this hypothesis, the authors of this paper looked for Carney complex patients who also had liver tumors that appeared to be FLC.
The authors found three patients with a history of Carney complex who had tumors that were diagnosed as FLC. The FLC diagnosis was based on the tumors displaying the typical morphology, or shape of structures in the cell, of FLC and positive markers for keratin 7 and CD68, which were previously published markers for FLC diagnosis. Also, based on a method known as FISH (to see an explanation of this method, please see the patient summary for “Molecular testing for the clinical diagnosis of fibrolamellar carcinoma.”), the authors could not detect the DNAJB1-PRKACA fusion, which is found in tumors of FLC patients who do not have Carney complex, in these tumors. But the tumors from these patients did not have the PRKAR1A protein present, suggesting over-activation of PRKACA in these tumors. Other liver tumors, known as hepatic adenomas, were also found in Carney complex patients, and these tumors also did not have PRKAR1A.
The authors suggest that, based on these patients, FLC be included as part of the Carney complex, though it has a different underlying DNA mutation than the FLC patients who do not have Carney complex. Carney complex patients would benefit from screening for FLC and other liver tumors as well, to help with early detection and treatment.
- Melissa Jarmel, Rockefeller University
All tumors of fibrolamellar hepatocellular carcinoma have a single common alteration in their DNA: a deletion of 400,000 bases in one of the two copies of chromosome 19 that results in the loss of seven genes that encode for proteins and the fusion of parts of two genes to create a new protein-coding gene known as DNAJB1-PRKACA. These researchers used new genome-editing techniques to ask what was the more important driving factor for this cancer: the presence of this new protein coded by the DNAJB1-PRKACA chimeric gene or the loss of the seven protein-coding genes.
The seven genes between the mouse DNAJB1 and PRKACA are orthologs (genes in different species that have the same function) of the seven genes found between the human DNAJB1 and PRKACA. The researchers used a new biological tool, known as CRISPR-Cas9, to cut out these 7 genes in mice and stitch together the parts of the DNAJB1 and PRKACA that are found fused together in this human cancer; CRISPR-Cas9 genome editing is a cutting-edge technique that lets us cut DNA with great precision using what are known as guide RNAs and the DNA-cutting enzyme known as Cas9. Using two sets of guide RNAs in mice to create the fusion of DNAJB1-PRKACA, these researchers were able to create tumors that resemble the human tumors for this cancer in many ways that can be observed by pathologists, though no fibrosis or metastases were observed.
Additionally, the researchers used a well-established technique known as the Sleeping Beauty transposon system to add a copy of the DNAJB1-PRKACA protein-coding gene or the PRKACA protein-coding gene to mice without deleting the seven protein-coding genes previously mentioned. This allowed them to ask if the presence of the new DNAJB1-PRKACA protein or more PRKACA protein was enough to make tumors in the mice. The mice with the DNAJB1-PRKACA protein were able to make lethal tumors similar to the CRISPR-Cas9 mice and human FLHCC. Additionally, mice just making more PRKACA protein only had abnormal liver cells and no lethal tumors. Thus, it was the formation of the new fusion gene, not the loss of the genes in between, that was important to make tumors. To further support this point, the researchers repeated this experiment using a version of the DNAJB1-PRKACA protein that could not function and found that they were not able to get abnormal liver cells. This is a promising sign that the development of a drug that can make the DNAJB1-PRKACA protein non-functional could be an effective therapy.
Lastly, the researchers asked what, in addition to the DNAJB1-PRKACA fusion protein, could enhance the growth of these tumors in mice to potentially give clues about how this cancer develops. They used the transposon system to introduce other known drivers of liver cancer with the DNAJB1-PRKACA protein-coding gene to see if tumor growth was enhanced. The only protein-coding gene they tested that showed enhanced tumor growth with DNAJB1-PRKACA was CTNNB1T41A, which makes an activated form of β-catenin, a protein involved in how genes are made into proteins and in how cells stick together. Significantly, this combination was also found in a primary FLHCC sample and corresponding brain metastasis. Next, they used a drug, known as hepatotoxin 3,5-diethoxycarbonyl- 1,4-dihydrocollidine (DDC), that causes liver damage and leads to fibrosis while expressing these two protein-coding genes and observed a dramatic decrease in survival of these mice. The development of these mouse models with tumors that closely resemble tumors found in FLHCC patients creates a viable system to test any new therapies that target this DNAJB1-PRKACA protein.
-Melissa Jarmel, Rockefeller University
The DNAJB1-PRKACA gene fusion is found in almost all fibrolamellar hepatocellular carcinomas. This fusion is formed by a ~400 kilobase deletion in chromosome 19. While this gene fusion has been identified in almost all tumors from patients with fibrolamellar hepatocellular carcinoma, the gene fusion has not been proven to be the causative agent in the formation of these tumors.
The authors of this paper used the CRISPR/Cas9 technique to introduce the chimeric DNAJB1-PRKACA gene fusion into livers of healthy mice to determine if this gene fusion can cause tumors.
The CRISPR/Cas9 system uses guide RNAs that target specific sequences in DNA, which then tell the Cas9 enzyme to cut the DNA sequences at those particular points. The authors of this paper designed guide RNAs to target the DNAJB1 and PRKACA genes, causing them to fuse and create the DNAJB1-PRKACA chimeric gene. They tested this system in mouse cells grown in a dish. After introducing the guide RNAs into the mouse cells in a dish, the cells were lysed and the DNA was harvested and sent for sequencing. The sequencing showed that the mouse cells now had the DNAJB1-PRKACA DNA fusion, when they previously did not.
Knowing that they could successfully create the DNAJB1-PRKACA DNA fusion using the CRISPR/Cas9 system, the authors moved on to an animal model. They injected the guide RNAs into the tail veins of healthy young mice. From the tail vein, the blood drains into the liver, bringing the guide RNAs into contact with the liver cells. The mice were sacrificed 14 months after the injection and their livers were examined for the presence of alterations. Twelve of the 15 (80%) of the mice had developed tumors in their livers.
The authors examined these tumors for similarities to the human fibrolamellar hepatocellular carcinoma by looking at cell shape and other tissue characteristics under a microscope. They determined that the mouse liver tumors had a similar appearance and immunohistochemical characteristics to the human FLHCC. Immunohistochemistry allows scientists to see if certain proteins are present in a tissue sample by using antibodies that target those proteins. Most importantly, the authors harvested DNA from the mouse liver tumors and found that the DNAJB1-PRKACA DNA fusion was present in the tumor cells.
The authors of this paper have shown that the DNAJB1-PRKACA DNA fusion appears to cause FLHCC tumor formation in mouse livers in the absence of other genetic mutations or liver damage, meaning that this DNA fusion may be the only thing necessary to cause FLHCC.
If this DNA fusion is the genetic cause of FLHCC, therapies targeting the fusion should be successful in treating the disease. Such therapies have not yet been developed.
-Nikki Croteau, Rockefeller University
“Molecular testing for the clinical diagnosis of fibrolamellar carcinoma” Published, September, 2017 in Modern Pathology Currently, fibrolamellar hepatocellular carcinoma (FLHCC) is diagnosed based on the histology and morphology of the liver cells in tumor tissue, meaning pathologists examine the shape of the cells and what else is present in the tissue after surgery using a microscope to determine the type of tumor that was removed. Pathologists can also confirm diagnosis using immunohistochemistry based on morphology. In other words, they can see if certain proteins found in fibrolamellar, like cytokeratin 7 and CD68, are more present by using antibodies that target these proteins. Previous studies have shown that a fusion of part of the DNAJB1 gene and part of the PRKACA gene is present in the DNA of FLHCC tumors. The authors of this paper propose another method to confirm diagnosis of FLHCC in addition to morphology of the cells and immunohistochemistry of the tissue based on the presence of this DNAJB1-PRKACA DNA fusion event in FLHCC tumors known as break-apart fluorescent in situ hybridization (break-apart FISH). This is a technique that allows one probe made of a specific DNA sequence to emit green light when it binds to its complementary sequence and another probe made of a different specific DNA sequence to emit a red light when it binds to its separate complementary sequence in a cell. When green and red light emit from the same place, it is visualized as a yellow light. When green and red light emit separately, the sequence has “broken-apart.” The authors used probes made from the DNA sequence of the PRKACA gene; the probes use sequences of DNA that are next to each other in the PRKACA gene so when the PRKACA gene remains whole, the probes bind to it and yellow light is emitted, and when the PRKACA gene breaks apart, like it does with the DNAJB1-PRKACA fusion event in FLHCC, the probes bind and red or green light is emitted separately. The authors’ findings show that this method is a clinically useful tool to confirm the diagnosis of FLHCC, in addition to morphology and immunohistochemistry. More accurate diagnosis will aid in providing better treatment for patients. –Melissa Jarmel, Rockefeller University
CD47 is not Over-Expressed in Fibrolamellar Hepatocellular Carcinoma” Published in Annals of Clinical and Laboratory Science, August 2017. One of the ways the body removes unwanted cells is through phagocytosis, a process where other cells engulf and breakdown unwanted cells and other small particles. This process is balanced by the number of pro-phagocytic markers and anti-phagocytic markers displayed on the outside of a cell. Cancer cells can survive by producing more of these anti-phagocytic markers than pro-phagocytic markers. CD47 is one of these anti-phagocytic markers that is over-expressed in some cancers and is currently being targeted in clinical trials. Since CD47 over-expression is found in conventional hepatocellular carcinoma, the authors of this paper wanted to know if the same was true for fibrolamellar hepatocellular carcinoma. Looking at the protein and gene expression of CD47 in fibrolamellar hepatocellular carcinoma, the authors did not see strong expression of this anti-phagocytic marker. For this reason, they believe, there is no evidence for patients with this cancer to undergo clinical trials for anti-CD47 treatments. This paper provides additional support to the finding that fibrolamellar hepatocellular carcinoma is a distinct cancer from hepatocellular carcinoma. -Melissa Jarmel, Rockefeller University
A Proposed Physiopathological Pathway to Hyperammonemic Encephalopathy in a Non-Cirrhotic Patient with Fibrolamellar Hepatocellular Carcinoma without Ornithine Transcarbamylase (OTC) Mutation Published March 2017 in The American Journal of Case Reports High levels of ammonia have been a serious complication of advanced fibrolamellar. Clinicians have found that treatments such as lactulose, which work well for other liver cancers, do not work well for fibrolamellar. In this paper, the authors propose and test a novel treatment on one patient which proves successful in treating his severe hyperammonia. The authors noted two earlier findings: that a protein called Aurora Kinase A is increased in fibrolamellar; and that Aurora Kinase A is expected to increase the expression of a protein called ornithine decarboxylase, which can produce an imbalance in the urea cycle and thus, high levels of ammonia. The authors then propose a novel treatment of sodium benzoate (3 g) and arginine (3 g) administered every four hours via nasogastric tube. The authors demonstrate the successful use of this treatment on one 31-year-old fibrolamellar patient. Sodium benzoate is a food preservative found in salad dressings, carbonated drinks, jams and fruit juices, pickles, and condiments. It has been successfully used before to treat urea disorders but never, apparently, fibrolamellar. This research offers the hope of alleviating the effects of hyperammonemia in fibrolamellar. –Dr. Sandy Simon, Rockefeller University
Environmental Exposures as a Risk Factor for Fibrolamellar Carcinoma Published in Modern Pathology, March, 2017 Searching for causes of fibrolamellar, researchers noticed a potentially important chronological clue: all identifications of fibrolamellar occurred after 1939, when World War II began. The researchers then went into the pathology archives at the Mayo Clinic, in Rochester, Minnesota, and analyzed the original slides of tissues created from surgeries that took place between 1905 and 1939. Further, they reviewed the patients’ clinical charts and performed a variety of state-of-the-art tests (at the histologic, ultrastructural and proteomic levels) to confirm the diagnoses associated with the slides. Using these methods, they were able to positively identify two cases of fibrolamellar, from 1915 and 1924. The existence of these two cases from before WWII implies a “reduced likelihood” that exclusively post-WWII factors, such as the rise of the plastics industry, cause fibrolamellar. The authors report that their findings, along with the fact that fibrolamellar tumor cells have aryl hydrocarbon receptors on them, suggest the possibility that environmental exposures that are new since World War II may not be exclusively responsible for causing fibrolamellar. However, they note that their findings do not exclude a role for other potential environmental risk factors. Nor does this work address the change in the frequency of fibrolamellar diagnoses before and after World War II. –Rachael Migler, The Fibrolamellar Registry