After long negotiations, striking a bargain on AIP criteria

Autoimmune pancreatitis (AIP) is a rare form of pancreatitis that can also affect the bile ducts, salivary glands, kidneys and lymph nodes. The most common presenting symptom is obstructive jaundice with or without a pancreatic mass. AIP closely resembles pancreatic cancer and only recently has been recognized as a distinct clinical entity.

Distinguishing AIP from cancer is essential because the prognoses and therapies for the two disorders are vastly different. Yet diagnosing AIP is particularly challenging, and until recently, the difficulty was compounded by the lack of universally accepted diagnostic criteria.

East vs. West

The inability to reach consensus resulted from very different practices in the East and West, says Suresh T. Chari, M.D., of Mayo Clinic in Rochester, Minn. Dr. Chari is a principal author of Mayo's AIP diagnostic criteria known as HISORt, the mnemonic for histology, imaging, serology, other organ involvement and response to therapy.

"For example, in Japan, ERP is routinely used for investigating obstructive jaundice and is part of the criteria for AIP. In the West, diagnostic ERPs are rarely performed because of the risk of pancreatitis," he says. "What's more, very early on, the Japanese described changes related to AIP on pancreatogram, whereas we really didn't know what to look for."

The Japanese also chose to exclude other organ involvement and histology from their criteria, even though histology alone is often diagnostic. But Dr. Chari acknowledges that obtaining enough tissue to evaluate characteristic histopathologic features of AIP is difficult — a particularly important consideration in type 2 AIP, which can only be diagnosed on biopsy.

"Our endosonographers have an uncommon ability to get core pancreatic tissue using EUS. A histologic diagnosis of AIP requires core tissue rather than cells obtained by the more-commonly used fine-needle aspiration (FNA) biopsy. Also, our pathologists have a keen eye for and great experience reading EUS core biopsies, and their calls are almost always accurate. But not all centers have pathologists experienced in making sense of these bits and pieces of tissue," he says.

Serology was another sticking point, with the lack of specificity in the blood markers used by the Japanese a particular concern.


Many of these differences were finally resolved in 2010, when an international panel of experts developed consensus criteria for AIP. The criteria take into account worldwide differences in clinical practice while aiming to avoid misdiagnosis of either AIP or pancreatic cancer.

"The history has been somewhat tortuous," Dr. Chari says, "but we've finally arrived at a point where we can agree. We looked at various practices around the world, and if the approach was appropriate, we adopted it. This allows people to use different strategies, which is somewhat cumbersome, but we know the criteria are valid."

Like HISORt, the consensus criteria rely on five main features of AIP:

  • Histology
  • CT or MRI imaging or ERCP
  • Serology
  • Other organ involvement
  • Optionally, response to steroid therapy — usually a two-week steroid trial

"Diagnosing AIP is like solving a jigsaw puzzle, and when all the pieces come together, the diagnosis is certain," Dr. Chari says. "And in an effort to unite East and West, we came up with four or five different practice patterns that can be tailored for use in different countries and institutions."

The patterns are 100 percent specific, but have to be applied to the letter, he points out. "And even then, a small subset of patients with AIP — about 15 percent — will end up going to surgery. "Right now, we're trying to understand who these patients are and why we miss them."

Another problem is how to maintain long-term remission. Although the majority of patients experience significant improvement in symptoms two to three weeks after starting corticosteroid therapy, approximately 30 to 50 percent relapse when treatment is withdrawn.

"The Japanese put patients on small doses of steroids for two to three years," Dr. Chari says, but long-term steroid treatment, especially for elderly people, is controversial. So we are studying alternative approaches to long-term management of AIP. We are constantly learning from every patient and tailoring our approach."