February 2018—Like a pair of one-size-fits-all jeans, testing all head and neck carcinomas for human papillomavirus may have seemed like a good idea at one time. In many cases, in fact, HPV testing is just what treating clinicians and patients need. On the other hand, not every head and neck case requires it. Now, a newly published CAP guideline should help physicians figure out the right fit in multiple settings.
“There’s a lot of testing going on,” not all of it warranted, says William Faquin, MD, PhD, a coauthor and committee co-chair of the guideline (Lewis JS Jr., et al. Arch Pathol Lab Med. Epub ahead of print Dec. 18, 2017. doi:10.5858/arpa.2017-0286-CP).
“A significant amount of it seems to be driven by clinicians pushing for it,” adds guideline co-chair and lead author James Lewis Jr., MD. “There’s a lot of data. It’s confusing and complicated.” And pathologists don’t necessarily know how, or if, clinicians will use test results. “That’s where the pathologists start feeling pinched. It’s hard to say ‘no.’”
The guideline should help pathologists put their collective foot down as needed, say its authors, by giving clear guidance on the different scenarios for HPV testing.
Equally useful, they say, is that the guideline coincides with the eighth edition of the American Joint Committee on Cancer’s staging system, which took effect in January. The new edition stages patients specifically based on p16 or HPV-specific testing (depending on what approach institutions use). “It lines up perfectly,” Dr. Lewis says of the overlapping documents.
This, in turn, reinforces the need for laboratories to do appropriate testing, since the two groups (HPV positive and HPV negative) have different implications for patients and their management, says Dr. Faquin, professor of pathology, Harvard Medical School, director of head and neck pathology at the Massachusetts Eye and Ear hospital, Boston, and staff pathologist at Massachusetts General Hospital.
The timing of the guideline is fortuitous for another reason, says Dr. Faquin. In the past five or so years information about the role of high-risk HPV in head and neck cancers has grown tremendously. The literature on which the recommendations are based is robust—no small thing for this complicated and evolving field.
Like Orion’s Belt (to continue the vague sartorial analogy), the guideline contains three prominent points to help orient physicians, according to Dr. Lewis.
The first is statement No. 1 (there are 14 in total), which calls for routinely testing newly diagnosed oropharyngeal (tonsil and base of tongue) squamous cell carcinomas, of all histologic subtypes, from the primary tumor or from cervical nodal metastases using p16 immunohistochemistry, which is a prognostic marker and surrogate marker of high-risk HPV.
Second, the authors essentially state there’s no indication for routine testing for HPV or p16 in any other tumor type or at any other primary site in the head or neck. “We see testing being done for nonoropharyngeal sites from various institutions around the U.S.,” says Dr. Lewis, professor of pathology, Department of Pathology, Microbiology, and Immunology, Vanderbilt University Medical Center. This new guideline, he adds, “should give pathologists the strength to say no when asked for the test when it’s not indicated,” thereby cutting down on wasteful and potentially misleading testing.
Finally, Dr. Lewis says, the guideline offers much-needed direction on handling cases involving neck lymph nodes, where the risk of false-positives is higher. “The guideline will be very helpful, particularly with cancer of unknown primary, where you’ve got a squamous cell carcinoma in neck lymph nodes and it isn’t clear where it arose”—a fairly common scenario, he adds.
That first recommendation is critical, agrees Dr. Faquin. It’s a departure from what pathologists have traditionally done, which was to group all oral cavity and oropharyngeal carcinomas together. The guideline distinguishes between carcinomas from the oropharynx and those of the oral cavity.
Dr. Faquin also echoes Dr. Lewis in drawing attention to statement No. 5: Pathologists should routinely perform high-risk HPV testing on all patients with metastatic squamous cell carcinoma of unknown primary in a cervical upper- or mid-jugular chain lymph node (essentially levels II and III). Roughly 80 percent of patients with OPSCC develop metastases to their cervical lymph nodes. “If it’s positive for HPV, that helps the treating clinician to know this metastatic cancer is very likely coming from the oropharynx,” says Dr. Faquin.
There was plenty of discussion about metastases of unknown primary to a level II or III lymph node, Dr. Faquin recalls. “What do you do when it’s keratinizing?” he asks.
Nonkeratinizing SCC can be tested with p16. Standard nonkeratinizing morphology with positive p16 can be interpreted as HPV-positive SCC because both are very strong surrogate markers for transcriptionally active high-risk HPV. However, cases with keratinizing morphology and positive p16 are somewhat different. Is p16 testing sufficient, or is a confirmatory HPV-specific test needed? Eventually, the group decided a p16-positive test in that scenario should be followed with an HPV-specific test.
One way to think about that, Dr. Faquin explains, is to consider a subset of p16-positive skin SCC. “You can imagine a scenario where you have a metastatic, keratinizing skin squamous cell carcinoma to a cervical lymph node,” positive for p16 but HPV negative. That drives home another key point: “You shouldn’t interpret p16 immunohistochemistry positivity by itself as a marker of high-risk HPV outside of these specific recommendations in the guideline,” says Dr. Faquin. Positive results don’t carry the same meaning for tumors in other areas of the head and neck. “That’s something a lot of pathologists may not appreciate.”
It’s also important for physicians to know that, depending on the specifics of the case, p16 often can be used as a surrogate marker for high-risk HPV infection. “And then we also make a statement that additional HPV-specific testing can be done at the discretion of the pathologist or clinician,” Dr. Faquin says.
The guideline is far from simple, but it offers clarity in an area where, until now, confusion had been piloting the plane. Head and neck anatomy, with its wide range of tissues, is complex; likewise, a variety of cancers can arise not only in the oropharynx but in other areas of the head and neck, many of them HPV positive. This complexity creates a testing dilemma, Dr. Faquin says. “You want to perform a test where the result is going to mean something.”
Hence the exhaustive literature search that underpins the guideline. One intent was to identify situations in which testing could lead to a different prognosis. “But it’s even more than that,” Dr. Faquin says. “We found that it’s really only the [OPSCC] where being positive for high-risk HPV implies an improved prognosis.” These patients might be eligible for clinical trials involving deescalation. “In the past these patients have been treated the same way that patients with conventional squamous cell carcinomas were treated—very aggressively, which has a lot of associated morbidity.”
Dr. Faquin points to the algorithm (Fig. 1 in the published article) as one of the crown jewels in the guideline. “It looks very complicated”—it would be at home on a New York City subway map—“but it’s actually a nice algorithm, and fairly simple once you start going through it.” The top line provides four different scenarios: multisite tumor involving the oropharynx, cervical lymph node, nonoropharyngeal primary tumor, and oropharyngeal primary tumor.
The algorithm should help clarify for physicians when p16 IHC can be done alone, and when HPV-specific testing is needed. In standard scenarios of a nonkeratinizing SCC in the tonsil, for example, p16 alone can be used as an indicator of the cancer’s high-risk HPV status. But in other, nonclassic scenarios, p16 may be a useful screening test, followed by an HPV-specific marker.
The guideline is just as useful for telling physicians what not to do.
Some physicians might be surprised by statement No. 4, for example, which tells pathologists to not routinely perform high-risk HPV testing on patients with nonoropharyngeal primary tumors of the head and neck, though testing on a case-by-case basis can be done, of course. Dr. Faquin adds his voice to the dismayed chorus when he talks about inappropriate testing, including oral cavity cancers, benign tumors, and cancers in the larynx and sinus. “Some pathologists may find it surprising that there is no convincing evidence in the literature that supports testing of tumors outside the oropharynx.”
(One exception is the sinus tumor that resembles adenoid cystic carcinoma, says Dr. Faquin, but which is defined by HPV positivity. In this case HPV testing can be used for diagnostic purposes.)
It might also come as news to pathologists that they’re not supposed to provide a tumor grade for HPV-positive OPSCC (statement No. 13). “That’s important, because five or so years ago a lot of us were diagnosing these as poorly differentiated,” based on their nonkeratinizing basaloid microscopic appearance, Dr. Faquin says. “But they certainly don’t behave like that.”
Likewise, statement No. 9 says pathologists should not routinely perform low-risk HPV testing on head and neck carcinomas. That’s important, Dr. Faquin says, because low-risk HPV in general is not associated with transformation of tissues into carcinomas; rather, it’s associated with squamous papillomas.
In many cases the literature was not robust enough to make definitive recommendations, in part because other types of head and neck cancers are so rare. For small cell carcinomas of the oropharynx or sinonasal undifferentiated carcinomas (SNUC), a subset of each will be HPV positive. Regardless of HPV status, however, these tumors tend to have a bad prognosis.
Dr. Faquin predicts that the key will eventually turn out to be whether HPV is transcriptionally active and thus likely driving the tumor. The opposite may be true for the aforementioned SNUCs, where “the HPV may just be along for the ride. It may or may not have gotten the tumor started,” he says, “but it’s not necessarily transcriptionally active and driving the tumor.”
The guideline is pathologist driven, though treating clinicians (a phrase Dr. Faquin favors) contributed as well, and it has been endorsed by the American Society of Clinical Oncology. It’s aimed at those inside the lab and out, across all types of practices.
Nevertheless, it may prove particularly valuable to those who practice outside of large academic centers.
At Vanderbilt, Dr. Lewis and his colleagues see a steady stream of referrals. It’s not unusual, he says, to see cases of oral tongue SCC that have undergone p16 immunohistochemistry or HPV in situ hybridization. Dr. Lewis recalls being at a meeting and speaking with a pathologist from a lab that regularly refers patients to Vanderbilt, who said, “We understand and agree with you that there isn’t any established significance to it in oral squamous cell carcinoma. But our clinicians absolutely insist on it, so we grudgingly do it and report it.”
Why has unwarranted HPV testing become so widespread?
The motivations of treating physicians are good, but, as coauthor Beth M. Beadle, MD, PhD, puts it, “Some people take an idea and run with it.” If HPV is important, then we should do HPV on everyone, goes the thinking.
Another culprit: I don’t want to miss anything. “Exactly,” says Dr. Beadle, who is associate professor of radiation oncology and director of head and neck radiation oncology, Stanford University. She sees a similar path followed by some dentists—patients with a head and neck cancer require a dental evaluation before radiation treatment. “If there are any problems with dentition, then we want the teeth removed,” she says. For most patients, no or minimal intervention is needed. Sometimes, however, well-meaning but overly enthusiastic dentists “will suggest pulling all the teeth, because they’ve read radiation impacts teeth.”
Dr. Lewis says unneeded testing also gets launched by the understandable confusion stirred up by tumors outside the tonsil and oropharynx, including those in the larynx, oral cavity, nasal cavity, nasopharynx, and, occasionally, even the hypopharynx. A significant minority of these tumors are driven by high-risk HPV, but it varies by type. In the nasal cavity, the number may be three to five percent; in the nasopharynx, it may be 15 to 20 percent, he says.
These situations are familiar to most physicians, Dr. Lewis says. Take the case of a young, nonsmoker patient with larynx cancer. In such a setting, the near-automatic response might be to assume it’s HPV related. “And maybe it is,” says Dr. Lewis. “But we don’t know what it means for the patient, the prognosis, or what anyone would then do about it. We really don’t have any idea. Often it doesn’t change anything at all, and you would never want to change how they were treated or managed, because you might [do so] in a way that lessens the chance of a cure, or using a therapy that we know is effective.”
Dr. Lewis suspects it is the tendency to extrapolate. “You can’t do that,” he says.
Pathologists aren’t the only ones excited by the guideline. Dr. Beadle is blunt: “I cannot decide how to stage a patient and provide a treatment recommendation until I have HPV status.”
She describes her role in the guideline’s creation as providing relevance from a clinical perspective “and also some reality checks,” including limits on tissue procurement and integrating pathology into a comprehensive treatment plan.
Dr. Beadle predicts the guideline will help put a reality check on her clinical colleagues as well. “Some radiation oncologists, some surgeons, some medical oncologists ask for tests that are not appropriate. And in this day and age, every test that we ask for that’s not useful costs money and burdens people who are already busy with work that isn’t indicated.”
She plans to publicize the guideline heavily with her multidisciplinary colleagues. “I think it will be welcomed,” she says. Clinicians will know what to order and be reassured they’re not missing something critical; pathologists will waste less time doing testing on low-yield cases.
In addition, the guideline lets her remind pathologists how heavily she and her clinical colleagues rely on testing neck specimens of unknown primary. Those that turn out to be HPV-associated cancer “will significantly alter my treatment plan in terms of the fields and the dose that I treat with radiation.”
Dr. Beadle says the guideline (including future incarnations) will only become more important in the years ahead. “Community practices that typically had not seen a lot of head and neck cancers are seeing more. We know that this cancer is not going away. At all. Even if all children are vaccinated tomorrow, which isn’t going to happen, all the modeling suggests we will see the same number of oropharynx cancer patients—which, again, is growing—until 2060, because of the impact of previous infection of these patients from decades ago.”
Physicians are also excited about how the guideline ties in with changing treatment options. Deescalation, in particular, could be a welcome strategy for HPV-related OPSCC, depending on patients’ smoking status and disease stage. Patients and physicians would like to use less-intensive therapy, though Dr. Lewis says there are still too many unknowns to proceed with gusto. There’s a fair amount of indirect data to support deescalation, but quality prospective clinical data are not yet available, he says.
As Dr. Beadle confirms, “We have a lot of questions on how to manage these patients once we know they have HPV-associated cancers.” All evidence suggests that patients have excellent prognoses and better outcomes with standard treatment; several ongoing and recently completed trials are looking at how to balance optimal outcomes and minimal toxicity. “We have more evidence that treatment deescalation is going to be the answer, but we don’t know how low we can safely go.” Interestingly, she adds, this patient population appears to be quite engaged in their care. “The clinical trials testing deescalation accrued at record rates and closed much more quickly than anyone anticipated.”
Nevertheless, says Dr. Lewis, “Some radiation oncologists, surgeons, and medical oncologists are already tailoring their treatments quite a bit, based on the p16/HPV status. Others are more conservative and recommend the full standard of care until clinical trial data accrues.” It’s also not unusual, he says, for patients to be quite involved in discussions about their treatment. “I’ve not been in those conversations,” he says. “I’m sure it’s agonizing, because you want to be cured of the cancer without incurring major side effects from treatment, such as fibrotic neck tissues, dry mouth, or needing a feeding tube. It’s a tricky discussion right now.”
Pathologists can help calm the waters with a more consistent approach to HPV testing, Dr. Lewis suggests. “We need to provide more consistent information to doctors and patients.”
The guideline—for good reasons—commits to agnosticism on the matter of HPV-specific testing. “I do think that will be a little bit of a disappointment to some physicians,” Dr. Lewis says, “that we don’t recommend anything more than p16 for primary tumors in the oropharynx.”
Crowning one test as champion would be controversial, he continues, which is why the authors left a bit of daylight and left testing method to the discretion of the pathologists and treating clinicians for HPV-specific testing. “Because some people feel very strongly that you should do both tests—the p16 is positive, and then you do a [confirmatory] HPV-specific test. But we don’t all agree on that, and not all recommending bodies are doing that.” Furthermore, he notes, the guideline authors wanted to ensure they recommended a prognostic marker available to all labs.
Guideline coauthor James Rocco, MD, PhD, recalls this being one of the more lively discussions as the guideline came together. As the authors culled the considerable literature, they found only a handful of papers that compared different techniques.
“And then there were experts there who had direct experience with some of these tests,” Dr. Rocco continues, “and in their hands these tests were very, very accurate.” In other hands, accuracy proved harder to achieve. “If someone at a high-volume academic center calls a tumor p16 positive, that could mean something different than someone in the community calling it positive,” says Dr. Rocco, chair of the Department of Otolaryngology, Ohio State University Wexner Medical Center, and professor and John and Mary Alford chair of head and neck surgery, OSU Comprehensive Cancer Center. (Hence the guideline recommendation calling for 70 percent cutoffs for nuclear and cytoplasmic expression with moderate to strong intensity. “There’s a small percentage of skin cancers that can score as p16 positive, but if you use those nuclear/cytoplasmic cutoffs, they would often score as negative on subsequent HPV-specific testing,” Dr. Rocco says.)
The data on test choice are immature, Dr. Lewis says, but he adds that the guideline does note the authors’ strong support for HPV RNA ISH. It’s not available across the board in clinical practice, however, and it’s only now appearing on clinical automated platforms, which limits its use to a handful of large academic centers and reference labs.
There is some scuttlebutt about the future availability of p16 IHC. The data suggest one particular antibody works slightly better than others. Dr. Lewis calls it of “very marginal benefit, but the one or two percent performance difference is one or a handful of patients.”
But if p16 IHC follows prognostic and potentially theranostic routes—along the lines of ER/PR as breast biomarkers—standardization will be in order. “Our hair is not on fire about that,” Dr. Lewis says, given that most of the antibodies currently available work very well and most pathologists can reproducibly determine positive versus negative results. “But we all see that coming.”
Though the guideline will not mean a 180-degree turn for those who practice in larger, academic settings, it will still have an impact.
Dr. Lewis says pathology practices at Vanderbilt have been “tweaked a bit” by the guideline, particularly in the workup of neck cancers. Prior to this, “We were doing this in all squamous cell [cancers] in the neck, regardless of where it was located.” Now, he and his colleagues will likely move to site-appropriate HPV-specific testing in the neck.
At Massachusetts General Hospital, the guideline “supports what we’re already doing,” says Dr. Faquin.
Dr. Rocco says the guideline reinforces what many physicians are already doing, especially at high-volume centers. But for community hospitals, or at academic centers where there’s less buy-in on current standards, “the guidelines are very compelling,” he says.
Dr. Rocco says he wore two hats as a member of the guideline committee, the first being to bring additional clinical perspective to see how the recommendations would fit into the flow of clinical care. Not that the pathologists in the group were divorced from those considerations, he hastens to add. But some 85 to 90 percent of cases are initially worked up at the instigation of an otolaryngologist or head and neck surgeon.
Dr. Rocco is excited about the likelihood that the guideline will be used as part of clinical trials, including deintensification studies. To be productive, those trials need a common mechanism for determining HPV status. Of course, most of these trials are done at large academic centers that already tend to follow the guideline’s recommendations, he says. “But the little quirks and differences are now standardized, and that’s valuable.”
On a personal level, he says he’s thrilled that refining HPV testing will save not only money but time, for him and his patients. It’s not uncommon, he says, to see “patients who come to me as kind of a ‘prepackaged’ HPV-related head and neck cancer, based on inappropriate HPV testing,” done with either the wrong technique or in the wrong type of scenario. He then has to untangle the case for these patients, a sometimes lengthy process. Wrong tests and wrong conclusions can make patients think physicians don’t know what they’re doing, he says.
As the authors point out, of course, not every question has an answer yet, a truth that came up again and again during their discussions. The literature was too thin to establish a consensus on certain aspects of pathology workflow, such as tissue fixation, unmasking of antigens, and the like. Similarly, it wasn’t possible to make a recommendation about whether to test a rebiopsy in the case of a recurrence. And even though the guideline supports p16 testing, many felt RNA in situ hybridization against E6 and E7 would be the test of the future. “Technology is just exploding,” Dr. Rocco says, acknowledging the bane of every guideline writer’s existence.
Nevertheless, Dr. Rocco sounds excited that the guideline will help put an end to clinically irrelevant HPV testing. “I can stand up at my head and neck tumor board and say, ‘That’s according to CAP guidelines.’”
Karen Titus is CAP TODAY contributing editor and co-managing editor.