January 2019—Ten years ago, says Manhal Izzy, MD, the approach might have seemed quixotic: performing liver transplants in patients with early intrahepatic cholangiocarcinoma. Even today, it’s hardly standard of care. Nevertheless, it has moved well beyond the impossible dream category.
The five-year survival rate without any treatment is dismal—less than 10 percent. “That has not changed over time,” says Sumera Rizvi, MBBS, assistant professor of medicine, College of Medicine, Division of Gastroenterology and Hepatology, Mayo Clinic. Moreover, the overall incidence of cholangiocarcinoma has increased over the past few decades.
Seen from other angles, however, the numbers are more cheering. Take liver transplantation. The latest approach uses neoadjuvant chemoradiation prior to transplant. A 12-center study (Darwish Murad S, et al. Gastroenterology. 2012;143:88–98.e3) that looked at patients with perihilar cholangiocarcinoma showed recurrence-free survival rates of 78 percent and 65 percent at two and five years, respectively, post-transplant. (Survival following liver transplant alone, without neoadjuvant therapy, is about 20 percent, says Dr. Rizvi, citing data from the 1990s.)
“Everybody sees the prognosis for this tumor,” Dr. Izzy says, “so everybody is eager to try to find something new,” both diagnostically and therapeutically. When the challenge is large, it’s often best to get everyone kicking in the same direction, like the Rockettes.The first step is to master the basics of cholangio-carcinoma. Although it’s the most common biliary malignancy and the second most common hepatic malignancy, the numbers are still low enough for it to be considered rare. It’s hard to gain mastery over a problem that’s rarely encountered, says Rondell Graham, MBBS, GI molecular pathologist, Mayo Clinic, noting there are about 3,000 new cases total in the United States annually.
That number alone doesn’t tell the entire story. The need for diagnostic testing—including laboratory-based modalities—is high for certain subsets of patients, who will need more intensive monitoring, including multiple FISH assays. “So even though incidence is low, the number of patients being tested is significantly higher,” says Benjamin Kipp, PhD, section head, genomics laboratory, solid tumor testing, Mayo Clinic.
The low incidence is one reason cholangiocarcinoma remains somewhat misunderstood. In fact, “The genetic profile of these are so different,” says Dr. Kipp, “that it really represents three different diseases.”
- Perihilar cholangiocarcinoma arises between the second-order bile duct and the insertion of the cystic duct. These account for approximately half of cholangiocarcinomas. These are often de novo, but they can also be associated with primary sclerosing cholangitis, or PSC, a chronic inflammatory condition of the biliary tree. Perihilar cholangiocarcinoma occurs in about 10 percent of PSC patients at some point in their lifetime.
- Distal cholangiocarcinoma arises below the insertion of the cystic duct and accounts for about 30 percent of cases.
- Intrahepatic cholangiocarcinoma arises above the second-order bile duct. This is usually found incidentally, in patients who may have underlying chronic liver disease or cirrhosis and undergo a routine surveillance CAT scan or ultrasound, which reveals an intrahepatic mass lesion. These account for 20 percent of cholangiocarcinomas.
The first two types are the focus of laboratory-based diagnostic modalities, including biliary cytology and FISH. Diagnosis can be a challenge, says Dr. Graham, “because there is no specific affirmative marker for cholangiocarcinoma. We rely heavily on morphology.”
Very high levels of CA 19-9 are associated with metastatic cholangiocarcinoma, Dr. Izzy says. “So it gives an impression, a clue, about the extent of the disease when it’s very high.” But some cancers—the subset called Lewis antigen-negative—don’t secrete the marker, so normal levels can’t be used to rule out disease, he cautions.
It’s not unusual, says Dr. Rizvi, for someone to be told they have bile duct cancer based on suspicious cytology. “They’ll come to us, but on a repeat ERCP we don’t see suspicious cytology and/or FISH polysomy, and we can’t confirm it or find other confirmatory criteria. That is a situation we occasionally encounter.”
Perihilar and distal cholangiocarcinomas are the most diagnostically challenging, Dr. Izzy says. “You need multiple modalities in addition to imaging,” including endoscopic biopsy and both cytology and FISH. “Even then, you might not be able to confirm or exclude the diagnosis,” Dr. Izzy says.
It’s also important, he says, to rule out IgG4 cholangiopathy in perihilar and distal biliary strictures. This is a type of autoimmune phenomenon that affects the biliary tree and can be confused with cholangiocarcinoma.
As noted, intrahepatic cholangiocarcinoma is usually detected during radiologic surveillance in patients with cirrhosis. “When we do every-six-months imaging,” says Dr. Izzy, “we see sometimes atypical lesions—atypical in the sense that they don’t look like hepatocellular carcinoma.” These are the simplest cases to address diagnostically, he says, which can be done with an intralesional biopsy that reveals features of intrahepatic cholangiocarcinoma rather than hepatocellular carcinoma.
Diagnosing perihilar and distal cholangiocarcinomas requires cumulative evidence from the multiple modalities.
“Let’s start with cytology,” suggests Dr. Izzy. The problem is that cholangiocarcinoma is a desmoplastic tumor—the fibrotic surface can make it difficult to discern any malignant cells that are obtained from the brushings. In other cases, samples can be paucicellular.
The rather subjective nature of cytology can also bedevil clinicians, despite pathologists’ best efforts to classify a sample as normal, atypical, suspicious, or positive.
With all those challenges, cytology becomes a very specific but relatively insensitive test. One study showed cytology sensitivity can be as low as 18 percent, says Dr. Izzy; that same study showed that FISH revealing polysomy was 47 percent sensitive (Moreno Luna LE, et al. Gastroenterology. 2006;131:1064–1072). Another study showed sensitivity of 15 percent and 34 percent, respectively (Kipp BR, et al. Am J Gastroenterol. 2004;99:1675–1681). “So there’s an increased yield for FISH when it reveals polysomy,” Dr. Izzy says. “With advanced FISH probes, you can improve sensitivity to more than 60 percent” (Barr Fritcher EG, et al. Gastroenterology. 2015;149:1813–1824.e1).Despite the need for cumulative evidence, clinical practice can fall short of that ideal. “Sometimes all you have is just one of these,” Dr. Izzy says—imaging suggesting a malignant stricture, but nothing else. “There may be no supporting evidence whatsoever, whether histologically or biochemically.” In those cases, physicians are compelled to keep obtaining follow-up specimens to prove or disprove cancer.
Primary sclerosing cholangitis adds to the diagnostic difficulties. Unlike most other cancers, Dr. Rizvi explains, cholangiocarcinoma cells have an affinity for bile, so they tend to grow longitudinally, along the bile ducts. It can be tricky, on imaging, to differentiate between a stricture related to PSC and one that is cancer in a patient who also has PSC.
The first step from a laboratory perspective would be to obtain biliary brushings from endoscopic retrograde cholangiopancreatography, which are sent for cytologic evaluation and FISH.
Biliary cytology can be as high as 95 percent specific. That’s the good news. The bad news: “Sensitivity is terrible,” says Dr. Rizvi. Though estimates vary depending on the study (as Dr. Izzy noted), she puts it at about 30 percent. “Cholangiocarcinomas are, histologically, very dense, desmoplastic tumors. They have an abundant tumor microenvironment, but they’re also paucicellular.”
Adding to the challenge, the biliary specimen can be difficult to access. Nevertheless, because of its specificity, biliary cytology remains the gold standard.
Dr. Rizvi notes that some cytology diagnoses are trickier for clinicians to deal with than others. Atypical cytology is a frequent diagnosis in PSC, and by itself should not raise concern for the presence of cholangiocarcinoma. “When we see it, we consider it in the same way we would consider normal cytology. So we don’t react to it, especially in PSC patients.”
In terms of suspicious cytology, she continues, “That’s more concerning than atypical,” especially if there’s FISH polysomy and perhaps a dominant stricture. About 30 to 40 percent of PSC patients who have a suspicious cytology but don’t have a mass lesion ultimately end up being diagnosed with cholangiocarcinoma down the line.
Mayo Clinic has been at the forefront of developing and using FISH to diagnose cholangiocarcinoma, including a specific pancreatobiliary FISH probe set (Barr Fritcher EG, et al. Gastroenterology. 2015;149:1813–1824.e1) that Dr. Rizvi and her colleagues use. Cytology and FISH are done simultaneously, after an MRI/MRCP to look for a mass lesion or dominant stricture. Dr. Rizvi says that Mayo is one of the few institutions in the country that offers FISH analysis on biliary brushings. “That has been essential in our ability to diagnose these patients,” she says.
It has four locus-specific probes (1q21, 7p12, 8q24, and 9p21) against loci that are known to be abnormal in cholangiocarcinoma. A normal finding is two copies of a chromosome. FISH polysomy is the abnormal finding. Sensitivity with this set is about 65 percent. Neither trisomy nor disomy indicates cholangiocarcinoma.
This approach “has proven very useful and valuable, especially in those challenging, difficult-to-diagnose cases,” Dr. Rizvi says. Because PSC is a chronic inflammatory condition, however, “We have to be very careful if we notice FISH polysomy.” In these patients, FISH polysomy alone would not confirm cholangiocarcinoma. “We have to be careful not to overdiagnose.”
Although Mayo obviously sees its fair share of referrals, it also performs its FISH analysis as a sendout test for other centers. “I do think if there’s any lack of clarity in terms of the diagnosis, it’s beneficial to send the patient to a tertiary referral center, just because it’s such a challenging malignancy to diagnose,” Dr. Rizvi says.
A good management protocol is essential, she continues, and she credits Mayo’s for helping colleagues sidestep the field’s many knowledge gaps (Rizvi S, et al. Clin Gastroenterol Hepatol. 2015;13:2152–2165). If a patient—particularly one with PSC—has a first-time FISH polysomy result, with nothing else to indicate possible cholangiocarcinoma, a repeat ERCP is done in three to four months to ensure the finding is true. If a patient has serial polysomy FISH, “we keep them under intensive surveillance,” she says.
Should a repeat ERCP show positive cytology, that, coupled with serial FISH polysomy, would result in a diagnosis of cholangiocarcinoma. “We’re done,” says Dr. Rizvi. But if a repeat ERCP shows normal cytology, and if the FISH result is now also normal, “then we would probably just do an MRI/MRCP in about six months to make sure if anything’s evolving, we can capture it.”
Though PSC clouds the diagnostic process, it offers one advantage: These patients are being followed closely, increasing the odds of earlier detection. Cases of de novo perihilar cholangiocarcinoma are difficult to catch early.
Says Dr. Kipp of Mayo’s genomics laboratory: “The best way to cure this disease is to detect it earlier.” But laboratorians also need to keep an eye on the literature, “so we can assure ourselves, when there is a companion diagnostic or a new targeted therapy, that we have the assays in-house to predict whether patients will respond.”A correct diagnosis of perihilar cholangiocarcinoma can offer hope to some patients. If the tumor is less than 3 cm and is unresectable, otherwise healthy patients might qualify for neoadjuvant chemoradiation and liver transplantation. Mayo has done more than 200 transplants for this indication, Dr. Rizvi says.
Patients with PSC will qualify for this treatment even if their tumor is resectable, assuming they meet the other criteria. “PSC patients have what we call a field defect,” Dr. Rizvi says, placing them at risk for developing cholangiocarcinoma—a risk that would remain even if half the liver were removed.
She urges caution on one more point. With the emphasis on making the correct diagnosis, some providers might be tempted to perform a confirming biopsy on a mass lesion present on MRI/MRCP. But Dr. Rizvi and colleagues have found that for patients who’ve undergone a liver transplantation who first underwent a percutaneous biopsy—in other words, not an endoscopic biliary biopsy—“the risk of seeding along that needle track was incredibly high,” Dr. Rizvi says, “and the vast majority of patients developed recurrent or metastatic disease after liver transplantation.” Even biopsying via endoscopic ultrasound is a contraindication for liver transplant. “We have had instances where patients who otherwise qualified for liver transplant were, unfortunately, not candidates because the mass had been biopsied percutaneously.”
For intrahepatic and distal cholangiocarcinoma, patients might be candidates for surgical resection, which is the current standard of care in early disease. Emerging data show a potential benefit for transplanting in very early cases of intrahepatic cholangiocarcinoma. “It’s still under investigation,” cautions Dr. Izzy.
If surgery is not an option in these types of cases, the treatment is systemic therapy. In some cases of intrahepatic cholangiocarcinoma, Dr. Izzy adds, locoregional therapy can be considered, which involves transarterial embolization or radiotherapy; a combination of locoregional therapy and chemotherapy might also be an option. Dr. Izzy pronounces himself “hopeful” that future approaches will offer better survival and outcome.The incidence of cholangiocarcinoma overall is rising, in part due to improved diagnostic modalities. A recent study looking at tumor of unknown primary in the liver showed that a good percentage were, in fact, cholangiocarcinoma (Hainsworth JD, et al. J Clin Oncol. 2013;31:217–223). In addition, some risk factors that contribute to the development of cholangiocarcinoma, such as obesity and alcoholic liver disease, are on the rise.
Many, including Dr. Izzy, are holding out hope for cell-free DNA or circulating tumor DNA. One study involving cases of predominantly perihilar cholangiocarcinoma has shown that circulating cfDNA with certain differentially methylated regions can detect the presence of cholangiocarcinoma. The sensitivity exceeded 80 percent in the same study, presented in 2017 at Digestive Disease Week, Dr. Izzy says, though he hastens to add that more studies are needed.
Dr. Rizvi and her colleagues recently reviewed emerging technologies for diagnosing perihilar cholangiocarcinoma (Rizvi S, et al. Semin Liver Dis. 2018;38:160–169). Next-generation sequencing is being compared to FISH, and is in the looks-promising-but . . . category.
Some groups have looked at extracellular vesicles, which are membrane-bound vesicles that are intimately involved in cell-to-cell communication, carrying possibly cancer-promoting proteins such as microRNAs. Looking at the cargo carried by extracellular vesicles, and comparing the difference between patients with and without cholangiocarcinoma, might be a future diagnostic modality, Dr. Rizvi says.
Another possible diagnostic: proteomic analysis by mass cytometry, which has been used to look at a variety of specimens from bile duct cancer patients, including urine, serum, bile, and stool. Different peptide panels have been reported on, with variable sensitivity and specificity.
Of all the options Dr. Rizvi has seen so far, circulating tumor cells and cell-free DNA look to be the most promising, she says. All are novel biomarkers and could aid in earlier diagnosis.Progress continues at a polite pace, but no one is sounding discouraged. As Dr. Izzy notes, the obstinate nature of the tumor, and not of physicians, is the real challenge, and what seems to draw many to this field. Like singing “Bohemian Rhapsody” at a karaoke bar, tackling cholangiocarcinoma is not for the shy.
“As a transplant hepatologist, you see that this is one of the most difficult topics we deal with in our specialty,” says Dr. Izzy. “Difficult topics are stimulating topics—and you want to help patients.”
Dr. Graham likewise enjoys the twists and turns in the road, especially with molecular genetics and biomarkers. “There’s been so much to explore. We are just getting started.”
Karen Titus is CAP TODAY contributing editor and co-managing editor.