College of American Pathologists

  At LAST, one HPV diagnostic terminology


CAP Today




July 2012
Feature Story

Anne Paxton

“What’s in a name?” was the rhetorical question of Shakespeare’s Juliet. But the literal answer could apply as much to medicine as it does to Montagues and Capulets: Names and nomenclature can come freighted with implied allegiances and historical baggage that lead to serious miscommunication. Sometimes, it turns out, pathologists can be caught in the middle.

That’s one reason why the consensus recommendations of the CAP-ASCCP LAST project—the Lower Anogenital Squamous Terminology Standardization Project, proposing a unified terminology for human papillomavirus-associated disease—are such a noteworthy accomplishment. For more than two years, five interdisciplinary working groups of the LAST project, led by the CAP and the American Society for Colposcopy and Cervical Pathology, have deliberated on a consensus. They ultimately agreed on standard diagnostic terms for HPV-associated squamous intraepithelial lesions and superficially invasive squamous carcinomas of the lower anogenital tract, and appropriate use of specific biomarkers to clarify histologic interpretations and improve diagnostic accuracy.

The recommendations were published June 28 online ahead of print in the Archives of Pathology & Laboratory Medicine and the Journal of Lower Genital Tract Disease. Sessions on the LAST recommendations will take place at most of the major upcoming pathology society meetings.

HPV-associated disease has a long history of disparate diagnostic terms derived from multiple medical specialties, and pathologists have had a plethora of ways of signing out essentially the same disease process. “The different specialties over the course of time have named entities or diagnoses in particular areas of the body according to what their subspecialty interests were,” says David C. Wilbur, MD, director of clinical imaging, Massachusetts General Hospital, and professor of pathology, Harvard Medical School. “So you end up with different nomenclatures for the anus, cervix, vulva, and so on, and that’s led to miscommunication between pathologists and between pathologists and clinicians. But recent biology has shown that the disease processes all have the same etiology and pathogenesis.”

“Many of the terms that still linger in our practice were developed before we understood anything about HPV’s association with so many lesions,” says Mark H. Stoler, MD, associate director of surgical pathology and cytopathology, and professor of pathology, cytology and gynecology, University of Virginia Health System. This variation in terminology was a product of subspecialization. Dermatopathologists see a spectrum of biopsies, gynecologic pathologists see a different spectrum of biopsies, and GI pathologists see yet another spectrum. “But within the last 20 years or so, it’s been firmly established that it’s the interaction between HPV and the epithelium that drives most of the important pathology. So it makes sense that a precursor lesion caused by the same HPV type that looks morphologically identical under a microscope should have the same terminology, regardless of its body site—or the patient’s sex, for that matter.”

Unifying that terminology was the mission of the LAST consensus project’s five work groups. While Work Group 1 developed historical background, and Work Group 5 outlined an implementation plan, the core terminology and diagnostic recommendations were the focus of Work Group 2 (squamous intraepithelial lesions), Work Group 3 (superficially invasive squamous cell carcinoma), and Work Group 4 (biomarkers).

The LAST standardization process has been similar to that of the Bethesda Committee, which hammered out the Bethesda System standard nomenclature for cytology between 1988 and 2001, says Michael Henry, MD, director of cytopathology and associate professor of laboratory medicine and pathology at Mayo Clinic, Rochester, Minn. “The Bethesda System came about because there was a quagmire, if you want to call it that, of diagnoses that were being used for cervical cytology, and we wanted to standardize the terminology and make it clinically relevant.”

In histopathology of the lower anogenital tract, “the lack of standardization can result in varied treatments based on the different terminologies,” says Dr. Henry. “When a lot of these terms were being developed, we didn’t have a good understanding of what the disease process was in relation to HPV.”

For both the Bethesda System and LAST, clinical relevance is the bottom line, he emphasizes. “If the terminology is not clinically relevant, then it’s not going to be of use to the patient.”

The LAST project Work Group 2 produced three major recommendations: a unified histopathology nomenclature for HPV-associated preinvasive squamous lesions of the lower anogenital tract (LAT), a two-tiered nomenclature for noninvasive squamous proliferations of the LAT, and use of LSIL (low-grade squamous intraepithelial lesion) and HSIL (high-grade squamous intraepithelial lesion) terminology, which may be further qualified with the appropriate –IN terminology to indicate body site.

The cervix has by far the longest and most detailed experience with HPV-related lesions, and there was a tendency in the 1950s and 1960s to overtreat what was then called carcinoma in situ, notes Dr. Henry, who co-chaired the cervix subgroup of Work Group 2. “But new treatments have come along. Colposcopic exams improved, and new therapies were developed such as laser therapy, cryotherapy, and now what is most commonly used, LEEP. So therapy has evolved in the cervix, and we’re coming to some understanding about therapies to use in some of the other areas of the lower anogenital tract.”

There has been a perception that the pathogenesis of HPV-associated precancer is different at different body sites, but at the cellular level, it’s a mistaken perception, says Alan Waxman, MD, professor of obstetrics and gynecology at the University of New Mexico School of Medicine. “There are differences in progression, but the pathogenesis of the disease is essentially the same.”

Dr. Waxman, president of the ASCCP, was a member of Work Group 2. Of the clinical areas that show the need for a unified terminology, he says, “the two biggest areas that have affected my practice are vulvar and cervical disease.” In 2004, the International Society for the Study of Vulvovaginal Disease (ISSVD) replaced the previous three-grade classification system of vulvar intraepithelial neoplasia with the current single-grade system, in which only high-grade disease is classified as VIN, while condyloma (or genital warts) is a separate category. To him, that move made a lot of sense. “VIN1 is a condition of low malignant potential and we’ve generally not been treating it except when it appears as condyloma accuminata, but my pathologist colleagues have been slow to pick up that nomenclature and they still refer to VIN1, VIN2, and VIN3. As a clinician, I don’t know what to do with VIN2.”

The current terminology for the cervix—CIN1, CIN2, and CIN3—goes back to the late 1960s and early 1970s. “Our understanding of the pathophysiology of HPV-specific related disease has moved exponentially since then,” Dr. Waxman notes. “Now we’re becoming more and more comfortable with the concept of CIN1 being a productive viral infection and CIN3 being a precancerous condition, but CIN2 is this great confounder. Many of our high-grade diagnoses are CIN2. Different pathology labs use different criteria for diagnosing CIN2, and the reproducibility within a given lab, and even within a given pathologist looking at the same specimen over time, is very poor.”

For example, the rigorous definition of CIN3 is “lack of maturation that extends into the upper third of the cervical epithelium.” “Many labs don’t call it a CIN3, however, unless it becomes a full thickness change. So a lot of what would rigorously be called CIN3 is actually classified as CIN2. And some of the physicians that refer patients to me use labs that still use the term CIS [carcinoma in situ].”

“Many of us recognize CIN2 as a diagnosis that will include some lesions that behave like CIN1 and some that behave like CIN3. Which leads us, if we treat all of them, to overtreat a fair number of patients with CIN1 that would resolve.” Changing the nomenclature isn’t going to get rid of that gray area, Dr. Waxman admits. “But we hope by using an immunostain like p16 by a fairly rigorously established protocol, which is what LAST recommends, we’ll decrease the number of ambiguous diagnoses that we get.”

It was Work Group 4 that was charged with looking at what other testing could help standardize the diagnostic art and make it more reproducible, but Work Group 2 took those recommendations regarding the use of the p16 biomarker in certain circumstances and rolled them into its recommendations, Dr. Henry says. “There are certain areas where you’d want to use p16 staining to help in telling benign lesions from high-grade squamous intraepithelial lesions. It can help to delineate more specifically lesions where the histology diagnosis may be somewhat problematic.”

Work Group 2 had considerable discussion on whether to transition from the three-tiered system to a two-tiered system, and the solution was a “meeting in the middle,” Dr. Waxman says, through the addition of the –IN modifier. “The –IN modifier is something that worked with the Bethesda System, where we went from a three-tiered to a two-tiered nomenclature, and people seem very content with that.” In addition, he says, the strict definition of when to use the biomarker p16, and what constitutes a positive p16, went a long way toward the acceptance of the two-tiered system.

There are limitations on pathologists’ ability to make HPV-related diagnoses based on microscopic criteria, Dr. Henry notes. “What we are trying to do is go to a system that is more easily reproducible and more easily understood by most pathologists. The more tiers you put in a system, the less reproducible it is, but we had to balance that with the fact that use of the terms for intermediate lesions in younger individuals might be valuable information to the clinician to avoid overtreatment.”

In the whole LAST consensus process, the biggest disagreement was from gynecologists who were concerned about the loss of CIN2 nomenclature, Dr. Wilbur says. “That’s why you can still see that nomenclature retained. The option for saying whether something is CIN1, CIN2, or CIN3 wasn’t one of the original recommendations; it was put back in based on input from multiple parties in the gynecologic community.” Their concerns relate to “kind of a hodge-podge of cases, but most likely those with very tiny lesions, particularly in young women. Gynecologists were concerned they would be forced to treat all those women rather than simply follow them.” He does not agree that the option to not treat would be lessened, “but they felt strongly that it’s ingrained in practice out in the trenches, that if all of a sudden CIN2 disappeared, physicians might have a knee-jerk response to treat.”

Some clinician members of Work Group 2 expressed concern about being left without clear guidance in the event of equivocal diagnoses like CIN2 and CIN2,3. “Our clinical guidelines allow more conservative management for young women with CIN2 than those with CIN3. Lumping them both as ‘HSIL’ could muddy the waters in terms of who we treat and who we follow,” Dr. Waxman says.

However, Dr. Waxman is hopeful that the main outcome of Work Group 2 recommendations will be fewer lesions called high-grade that would go to treatment. “If you’re taking part of the CIN2s out of the high-grade mix, then you have a more homogeneous and smaller pool of patients whose cervixes need to be treated.”

Unfortunately, there is still too much overtreatment of cervical lesions that would probably resolve spontaneously, Dr. Waxman says. “We know that overtreatment of young reproductive-age women can lead to increased risk of pre-term ruptured membranes and pre-term deliveries. But we as clinicians are leery of leaving a potentially precancerous lesion untreated, so CIN2 was lumped in among the precancerous lesions. I’m hoping with more uniform use of p16, we’ll start seeing some change in that now.”

The main thing the LAST recommendations should accomplish, in his view, “is they should help clinicians better appreciate the prognosis for the diagnosis they get from the pathologist.”

Work Group 3’s task was to focus on very early cancer in the lower anogenital tract body sites, says Teresa M. Darragh, MD, professor of clinical pathology and obstetrics, gynecology, and reproductive sciences at the University of California San Francisco/Mt. Zion Medical Center. The work group recommended unifying superficially invasive squamous cell carcinoma under the abbreviation SISCCA, stating in pathology reports whether or not invasive tumors exceed the defined dimensions for a SISCCA, including parameters on lymph-vascular invasion and independent multifocal carcinomas in the pathology report, and specifically defining SISCCA of the cervix, the anal canal, and the perianus.

“The cervix is clearly where the vast majority of the literature has been, and also where some of the confusion has been, because there’s been a lot of different variations in the staging system and what is used varies across the U.S. and between the U.S. and other countries,” Dr. Darragh says.

The cervix subgroup synthesized the data and made a major step forward by eliminating the term “microinvasive cancer,” she says. “It is used inappropriately by pathologists for these very superficially invasive cancers in other body sites, where there is an understanding that the patient has a very low risk of metastatic disease, and could potentially be treated with a conservative surgical incision, for example.”

Her role was co-chair of the anus/perianus subgroup. UCSF, she explains, has one of the largest anal neoplasia clinics in the country and is involved in improving diagnosis of anal cancers. “Up until a couple of decades ago, the treatment for squamous cancer of the anus was a very, very big surgery called abdominal-perineal resection, basically taking out all the anus and rectum and sealing things. But along came radiation and then radiation combined with chemotherapy, and they found that the combined modality therapy had even better outcomes than surgery. But it’s still pretty tough because we’re hitting people with essentially sledgehammers, and not all patients can tolerate the course of treatment.”

Now clinicians are diagnosing more of the superficially invasive cancers, and the potential for limited surgical excision has been renewed. The improved LAST terminology, she hopes, will aid in the collection of consistent data prospectively as the incidence of anal cancer continues to increase. “So, if a patient and the health care provider opt for limited surgical excision, we will have defined parameters to be able to say, oh this was a success, there’s no recurrence or no metastasis given these parameters. Or no, this treatment failed the patient.”

Improvements in the treatment of HIV disease are part of the reason anal cancer is increasing, particularly in large urban centers like San Francisco, Chicago, and New York. “Patients with HIV disease are living longer, and now we’ve allowed the natural history of HPV to happen, which takes place over years to decades,” Dr. Darragh points out.

But while screening has helped reduce the incidence of cervical cancer by about 70 percent since the introduction of Pap testing—from 40 to 50 per 100,000 women per year down to eight to 10 per 100,000—the rate of anal cancer for the far smaller targeted risk group of HIV-positive men with sex with men (MSM) now runs at about 70 to 130 per 100,000.

Medical centers like UCSF are evaluating whether it makes sense to screen for the precursor lesions and potentially treat them or to find anal cancers in the very earliest stages. But as it is still an uncommon cancer, it probably doesn’t make sense to have a broad-scale screening program, Dr. Darragh says.

Another complication is that even though HPV16 is known to cause a large percentage of anal cancers in the highest-risk population—HIV-positive MSM—about 90 percent of that population have HPV in their anus even if they don’t have the disease, she notes. So the test may not be useful as a screening test or adjunct to screening.

With anal/perianal lesions, the importance of the LAST recommendations—especially the SISCCA term, which implies an early cancer with a low risk of spread—is somewhat different from its importance in the other areas. The work group hopes LAST will help the increasing number of research projects that are underway or contemplated. “These are just suggested definitions so we can go forward and collect some consistent data,” Dr. Darragh says. “The majority of data is about the cervix, the vulva comes in next, and there is precious little about the other body sites.”

The LAST Project Work Group 4 recommended the use of biomarker p16 immunohistochemistry for differential diagnoses between precancer and a mimic of precancer, to help clarify certain diagnoses, as an adjudication tool for cases where there is a professional disagreement in histologic specimen interpretation, and as an adjunct for negative biopsy specimens at high risk for missed high-grade disease. But the work group recommended against p16 as a routine adjunct for interpretation of all biopsy specimens.

One factor that complicated the task of Work Group 4 on biomarkers was the need to systematically review the literature on the large number of biomarkers that have been linked to HPV, says work group co-chair Dr. Stoler. “We put forward every biomarker we could think of that would be possibly useful, and captured all the literature, and let the chips fall where they may. And it was very clear that p16 was felt to be the best biomarker. There was much less evidence for the other biomarkers.”

Among the other biomarkers that Work Group 4 considered were ki-67 and ProEx C, says co-chair Dr. Wilbur. “They have a fairly significant body of literature, but it’s not even close to the body of literature out there for p16.” There were comments proposed to the effect that if p16 is technically inadequate or has some issue of interpretation, then ki-67 or ProEx C could be done as alternatives because of similar trending data, he notes. “But there were just anecdotal reports and essentially no clinical trials with any degree of rigor about these biomarkers.”

A corollary to the singling out of p16 was the recommendation against using multiple markers. “There was concern throughout the process that we would be promoting a more expensive workup of patients, or that the cost considerations of following the recommendations would not really support the medical benefit to the patient,” Dr. Stoler says. “Immunohistochemistry costs money, and we specifically looked at papers where multiple stains were used. And the benefit of doing two stains, which of course doubles the cost, versus one stain, was only marginal.”

The work group also considered whether, given the interpretive variability among pathologists, there is a really good biomarker that should be done for every single biopsy. “But because of the limitations of the evidence, we couldn’t find one that should be done.”

Like Work Group 2, Work Group 4 debated the merits of a two-tiered versus a three-tiered system. “Our charge was to look at the patterns of staining of all the biomarkers and to see if there would be patterns that would correlate with either system, and we could find no paper that used these biomarkers that correlated with the morphology for three categories, so by default they supported a two-tiered system,” Dr. Stoler says.

The issue of interpretive variability, he points out, is just a reality of pathology practice. “The average clinician with the average patient thinks the pathologist looks at a slide and whatever is said is accurate, but our paper and many others show that there is interpretive variability in many areas of pathology. And in this particular area, it is fortunate we have good biomarkers.”

The Work Group 4 recommendation that p16 biomarker be used to adjudicate in certain cases in which there is professional disagreement was also based on available evidence, Dr. Stoler says. “We looked at papers describing in-depth adjudication, which means expert pathologists all looked at the same slide without knowledge of the biomarker outcome and classified the tissue as CIN1 or CIN3 or whatever, and we know from many papers that adjudicated readings correlate better with biology. But in real practice you can’t do that; people are too busy to have every slide read by more than one pathologist. So what some of the papers demonstrated is by adding a good biomarker, you can make an individual pathologist read like an adjudicated panel of pathologists.”

Professional disagreement is not that prevalent in these cases, says Dr. Wilbur. He notes that the LAST recommendations include a chart estimating the amount of p16 that might be ordered or the percentage of biopsies that might have p16 applied, and it is one percent. “In terms of a percentage, we think that is pretty low.”

The work group recommends against use of p16 as a routine adjunct to histologic assessment with interpretations of negative, –IN1, or –IN3. “It’s very clear that p16 is not an absolute determiner, and in and of itself does not determine the presence or absence of high-grade disease,” Dr. Wilbur says. “That’s exactly the reason it should not be used in obvious low-grade disease. There is some data suggesting those low-grade lesions may be at increased risk for becoming high-grade lesions or harboring high-grade lesions, but the literature is incomplete at this point.”

But there was one circumstance that was considered an important exception. Says Dr. Stoler: “This is a special case where you’re sitting at your microscope and you get a cervical biopsy you call normal—but then the patient has an HSIL Pap test. We know biopsy misses a lot. There’s a paper showing 30 percent or more of the time the first round of colposcopy misses the CIN3 that is actually present. When you go back and look at slides that were called normal by the pathologist, a fraction of them really aren’t normal. So we try to save the patient a second colposcopy and get the right diagnosis, and one way is if a normal biopsy is associated with an HSIL Pap test, consider doing a p16 stain. A significant percentage of the time, this will highlight an area of missed CIN2/3 and will save the patient a second biopsy.”

This recommendation goes somewhat against tradition, he notes. “Historically, the perspective has been that histology was probably a better standard than cytology, but we know that the current scientific literature completely refutes that. The idea that histology doesn’t necessarily trump cytology is an important concept. In the most important example, a patient with unequivocal high-grade cytology who has a normal biopsy, statistically it’s overwhelmingly the case the biopsy has missed the lesion, and it still needs to be found and treated.”

How much additional use of p16 is likely now that the LAST recommendations have been made? “We’re hoping that p16 won’t be overused for things we don’t consider to be indications as we’ve defined them in the recommendations,” says Dr. Wilbur. “We estimate that less than 20 percent of cervical biopsies should have p16.”

The main focus of the LAST consensus, Dr. Stoler adds, is to address the variables that can affect an accurate diagnosis from a slide. “Given the literature and the known issues of interpretive variability, when pathologists have too many categories or when there are shades of gray that really represent a biological continuum, we think the approach in this paper leads to more accurate diagnoses.”

“Every time we make a diagnosis, clinicians respond with a management choice for the patient: Do I have to do another biopsy? Do I have to do surgery to treat the patient? And those choices are driven by words on the paper, and the motivation here is to more clearly communicate what the words on the paper mean so the clinician understands, and to improve how well the words on the paper reflect an accurate diagnosis.”

He expects that more clarity will be the result now that the LAST recommendations are out and being widely disseminated. “Some individuals are already doing a lot of what’s in the recommendations,” Dr. Stoler says. “For others, there will be changes in practice based on the terminology.” He cautions that while the LAST recommendations are exactly analogous to the recommendations of the Bethesda System for cytology, there is one important exception. “Historically, government payers and other payers said we’re not going to reimburse unless the diagnosis is put in TBS terms, and that really helped drive adoption.”

The LAST recommendations, by contrast, were developed with funding from the CAP and ASCCP, without a government agency taking the lead, and they affect only the fraction of the 60 million Paps that are done that get biopsies. Nevertheless, he hopes that as with the Bethesda System, there will be global adoption of the LAST standards. “To me that would be a wonderful thing, because a CIN3 in France should be the same thing as a CIN3 in the U.S.”

Now that the LAST consensus recommendations are public, Dr. Waxman thinks they have opened up an important research opportunity. “That would be to see if –IN2 as defined by the presence of a strong block p16 staining behaves similarly to –IN3. It will be interesting to find if there is uniformity of biological behavior in addition to a high consistency of interobserver diagnosis.”

The LAST consensus recommendations will help pathologists give much more consistent diagnoses, allowing clinical colleagues to be able to rely more on pathologists’ interpretations, Dr. Darragh believes. “Pathologists who may not be able to spend 90 percent of their professional time on this area will now have guidance on how to move forward and evaluate things.”

“Using all the tools that pathologists have in their toolbox,” she says, “be it simple light microscopy and ocular micrometers to measure depth of invasion in a consistent fashion, or the new molecular tests that we have, and, specifically for LAST, things like the IHC test to improve upon what we can do just by H&E morphology alone, that’s the pathologist moving into the 21st century. There’s so much change happening in medicine and pathology that having the guidance of these recommendations, I think, is really a step forward.”

Anne Paxton is a writer in Seattle.

Related Links Related Links