HPV self-testing; and hpv prevalence changes

There was an updated meta-analysis on the efficacy of HPV self-testing in women (see hpv self-testing bmj2018 in dropbox, or doi.org/10.1136/bmj.k4823)

Details:

-- this meta-analysis assessed 2 aspects of HPV testing:

    -- accuracy review: comparing the accuracy of vaginal self-sampling by women vs clinician sampling for HPV, and verification of the presence of CIN 2+ or worse by colposcopy and biopsy in those with a positive test; 56 studies included

    -- participation review: women in the self-sampling group (intervention arm) were compared with those who were invited to undergo screening by a clinician (control arm); 25 trials included

-- they also evaluated the response rates comparing different invitation scenarios for self-sampling: mailing self-sampling kits to all, opt-in, community campaign, and door-to-door.

Results:

-- accuracy studies:

    -- HPV assays based on PCR: as sensitive by self-samples as clinician samples to detect CIN 2+ or more, pooled ratio 0.99 (0.97-1.02); specificity was 2% lower on self samples vs clinician samples. Sensitivity/specificity did vary by women's pretest risk (eg general population screening vs those at higher risk), but not by whether self-sampled vs clinician:

        -- primary routine screening: sensitivity 96%, specificity 79% for both self-screen and clinician

        -- testing of high-risk groups (but only one study): self-screen: sensitivity 100%/specificity 61%; clinician sensitivity 100%, specificity 64%

        -- follow-up of prior cervical abnormalities/ colposcopy clinic: self-screen: sensitivity 88%, specificity 51%; clinician: sensitivity 90%, specificity 48%

    -- HPV assays based on signal amplification: self-samples less sensitive by 15%, pooled ratio 0.85 (0.80-0.89); specificity was 4% lower on self-samples vs clinician samples

-- participation studies:

    -- means of inviting women to self-sample vs going to clinic for clinician sampling

        -- mailing self-sample kits to the women’s home addresses had a higher response rate than forthose invited to get tested by a clinician, with HR 2.33 (1.86-2.91)

        -- opt-in strategy: no significant difference for self-sampling vs clinician testing

        -- community campaign: (only 1 study): much higher rate for self-sampling, with HR 2.58 (1.67-3.99)

        -- door-to-door (community health workers delivering test kits directly to women): by far the highest response rate, with self-sampling twice as effective but not reaching statistical significance, HR 2.01 (0.66-6.15), but 4x that of mailing kits to all women (the second highest rate)

    -- direct offer of self-sampling devices to women in communities that were under-screened: high participation rates (> 75%)

    -- participation rates were lowest in women who had never had prior screening

-- sample adequacy: only 0.7% of self-samples were unsatisfactory for HPV testing

-- adherence to follow-up of positive self-samples: 81%

Commentary:

 -- cervical cancer rates in low-resourced countries are higher than in Western Europe, North America, Australia, and New Zealand; 85% cases of cervical cancer occur in less-developed countries (eg: an incidence rate of 35/100,000 women-years in eastern Africa).

    --in the United States there’s quite a variation in cervical cancer rates related to demographic and social disparities: incidence rates are higher among Hispanic women (8.9/100,000 women-years) vs black women (8.4/100,000 women-years) vs white women (7.4/100,000 women-years), with differences largely explained by access to screening

-- most cervical cancer cases occur in women who have never been screened or are not involved in routine screening

-- as noted in a prior blog, primary HPV screening is becoming the preferred mode of cervical cancer screening (see for the USPSTF new recommendations:  http://gmodestmedblogs.blogspot.com/2018/09/uspstf-guidelines-on-cervical-cancer.html ). One concern is that just doing cytology screening can lead to false negative test. And, HPV screening is done by vaginal (not need endocervical) sampling and therefore amenable to self-sampling

-- a prior meta-analysis found that self-sampling was less sensitive than in the above study, but the lower sensitivity studies included were based on the inferior test of signal amplification vs PCR

-- both Australia and the Netherlands have already introduced self-sampling into their national screening guidelines. The US, Canada, and some European countries are waiting for more vigorous comparative studies

-- there are some studies suggesting value to urine HPV testing, or even molecular testing of vital or human genes which promote carcinogenesis.  But, these are not-ready-for-prime-time yet

So, this study does bring up/reinforce some important issues:

--it is clearly best to decrease HPV infections and cancer (cervical, oropharyngeal, etc) through vaccination of kids. And, the earlier, the better. Though the FDA did just approve extending the age range to 45 (see http://gmodestmedblogs.blogspot.com/2018/10/vaccine-approved-to-age-45-tdap-best.html ), it is not covered by insurance yet (from my understanding, and the cost is $200-300 just for 1 shot of the) til the ACIP makes a formal decision. And this age extension should ultimately help control the increases in oropharyngeal and other HPV-related cancers (see below)

--this new HPV self-testing has been advanced as low-cost and more implementable in resource-poor countries, and can increase women's participation rate in screening by screening at home.  But seems to me to be quite appealing to use here. Can empower women to take a more active role in their care. Less invasive than a clinician doing this (of course, some women may still prefer the clinician option, but seems that the option of self-sampling should be available).

    --there is a downside, which I should mention: as we primary care clinicians do fewer and fewer vaginal exams (since we now often do urine tests for most STIs, and now possibly adding cervical cancer screening), we will become less practiced at doing the exams/deciphering abnormal from normal, and as a result there may be more clinical errors/reluctance to do vaginal exams when indicated.  It’s a little like trying to get a barium enema done now. Since we do colon cancer screening preferentially by non-radiologic procedures, fewer of the newer radiologists get enough training in barium enemas to do them even when indicated

    --that being said, I do feel that the benefits of self-screening to women outweigh this concern. But I thought it was important to articulate this issue

--and, in the US/other resource-rich countries, the participation rate for women in cervical cancer screening should be lots higher if they are given the option of self-testing at the time of their regular appointment with the clinician, or of doing so at home and bringing the swab to the lab

----------------------------------------------------------------------------------------

As a related issue, the MMWR reported  trends in HPV-associated cancer from 1999-2015 in the US, noting overall cervical/vaginal cancers decreased but oropharyngeal, vulvar and anal squamous cell cancers increased (see https://www.cdc.gov/mmwr/volumes/67/wr/mm6733a2.htm)

Details:
--CDC analyzed data from cancer registries covering 98% of the US population
--30,115 new cases of HPV-associated cancer were reported in 1999; 43,371 in 2015

Results:
--cervical cancer: 11,788 cases in 2015, decreased 1.6%/year from 1999
--vaginal squamous cell cancer (SCC): 809 cases in 2015, decreased 0.6%/yr
--oropharyngeal SCC: increased in men 2.7%/yr and women 0.8%/yr. this is now the most common HPV-associated cancer, with 15,479 cases in men and 3,438 in women
--anal SCC: 2236 in men in 2015, 4507 cases in women;  increased in men 2.1%/yr and women 2.9%/yr
--vulvar SCC: 3890 cases in 2015, increased 1.3%/yr
--penile SCC: 1224 cases in 2015, stable
--there still remain racial/ethnic differences: highest rate of cervical cancer in Hispanic women, vulvar cancer in white women, anal cancer in white women, anal cancer in males somewhat higher in black men, oropharyngeal cancer much higher in white men

Commentary:
--it is known that HPV is by far the most common STI, 90% of new cervical infections resolve spontaneously, and those that do progress to invasive cervical cancer take decades to do so.  BUT, the natural history of HPV infections at other sites is less clear
--even though smoking rates have decreased, and oropharyngeal cancer is related to smoking, the cancer incidence has been increasing, probably related to HPV transmission from increased oral sex exposures
--there is not consistent testing for HPV in these non-cervical cancers in the US, but overall HPV DNA has been found in the following cancers: 91% of cervical, 91% of anal, 75% of vaginal, 70% oropharyngeal, 69% of vulvar and 63% of penile

The authors say that changing sexual behaviors could account for the changes in HPV cancers. They note: "Population-based screening is recommended for only one HPV-associated cancer (cervical) at this time; however, HPV vaccination can prevent infection with the HPV types most strongly associated with cancer."

geoff

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@uphams.org

to get access to blogs since 8/15/17:

1. go to http://gmodestmedblogs.blogspot.com/ to see them in reverse chronological order

2. click on 3 parallel lines top left, if you want to see blogs by category, then click on "labels" and choose a category

3. or you can just type in a name in the search box and get all the blogs with that name in them

to access older blogs from the BMJ website, from October 2013 until 8/15/17: go to http://blogs.bmj.com/bmjebmspotlight/category/archive/ 

please feel free to circulate this to others. also, if you send me their emails, i can add them to the list