Pathology Associates Of Lexington, P.A.
Pathology Associates Of Lexington, P.A.
Pathology Associates Of Lexington, P.A.
 Home | Pathology Group MembersOur Hospital  Search This Website:
        HPV test methods comparison

[disclaimer] As of 2000, our lab began to deal with liquid-based Pap samples; and we had SurePath "in-house" in 2001. As HPV requests were received, we forwarded samples to Molecular Pathology. In early 2003, responding to clinical demand, we brought the Digene "Hybrid capture 2" (hc2 or HC 2) HPV test "in house".

I like Dr. Plotz' comment14 about HC 2 (and it goes for PCR also): such a test " unable to tell the difference between a cervical sample that's negative for HPV and a vial of water." It is highly important that one always be aware of potential confounders of poor samples or mixed-up samples. Visually interpreted samples give the possibility of concordance correlation.

Through the oversight of our hospital's Institutional Review Board (IRB), we began to contribute sample residuals to an investigation of the "Inform HPV" (an instrument-based CISH method). In November 2003 (having become aware of the poor PPV of hc2), we decided that we ought to "take a look" at the Inform HPV system if we could do so at no direct cost to our group or hospital. There will be great debate as to efficacy of different methods for HPV detection. A method can be "valid" either by FDA approval or ASR (analyte specific reagent) protocol. A real advantage of CISH is that it is a morphologically interpreted test (viral positivity is seen as either in, or not in, abnormal cells), lets one know whether the virus is not integrated or integrated ("irreversible") in the nuclear chromatin, and can be tested "real time" so that a Pap ordered with "reflex HPV" can be reported as a single report with HPV results correlated into the original Pap report.

Since the chart below was constructed, TriPath's (TriPath bought out by BD in late 2006) "ProEx C" brought out its IHC marker for the abnormal S phase correlated with high risk HPV; other IHC markers proposed for HPV affected cell "marking" are p16, p21, p27, telomerase, and cyclin E.

Non-Pap samples: We have scraped the cut surfaces of a few HPV-suspicious lesions into SurePath vials and tested with the PCR methods with some success; it is easy to do this and hold the specimen pending the histology [L13-2101].

None of these methods test for immunoserological immunocompetence. They test for either viral elements or viral-related effect in the epithelial cells of the cervix. Gynecologists13 are aware of many patients with squamous dysplasia which goes away after biopsy. Did the biopsy inadvertently inject some viral elements into that patient's blood during the biopsy process? That is, did a therapeutic outcome become triggered by a diagnostic procedure? Do young patients with HPV-affected cervix epithelium "clear" the infection by age 30 because tiny traumas to their cervix over the years induce "vaccination"?

Caution: sales reps, clinical doctors, and many technologists & pathologists fail to keep their eyes on the precise characteristics of: the (1) clinical strategy for using the test in question (Pap smear...and, for example, never miss an invasive cancer vs. the greatest sensitivity for LGSIL); (2) therefore, the particular "gold standard" for looking at sensitivity, specificity, and predictive values (biopsy vs. PCR, etc.); (3) and therefore, the prevalence of the particular disease of most importance (ca/HGSIL vs. ASCUS/LGSIL). Some would claim, for example, that a high false positive rate represented advantageous "sensitivity"...we would disagree. The HCII is FDA approved as an adjunct to the Pap smear for women over 30 y/o.
 Pap SmearPCRDigene HC2 Ventana Inform CISH ProEx C
basic test cell recognition with Pap stain & reflects all HPV types that cause koilocytosis target DNA amplification (Roche to come out with PCR based manual system2) against any types of HPV for which there are probes 13 high risk HPVs; type-specific RNA probes that select HPV DNA single strands in solution & the target RNA:DNA Ag hybrids  "capture" to corresponding anti-RNA:DNA Abs anchored to microtiter well walls & signal amplification when subtrate added with chemiluminescent tagged anti-RNA:DNA that attaches to the well wall bound targets non-amplification chromogenic insitu hybridization & stain with cell recognition; tags 13 high risk HPVs.  
test readingslide based, visual morphologic diagnosis by pathologisttest tube based, system resultstest tube based, system results slide based, visual morphologic diagnosis [*see below] by pathologist; uses least amount of specimen.  
viral load needed to "trigger" we need several typical abnormal cells to make the diagnosis no way to consider adequacy of specimen cellularity no way to consider adequacy of specimen cellularity can observe and factor in specimen cellularity  
cross reactions ??yes, HR with LR5 no; but some artifacts may seem "positive"  
FDA approved?conventional, nono3yespending 3  
"market approval status"by common practiceASRASR then FDAASR  
lab must validate instrument & perform QC with all runs?n/ayesyesyes  
correlations with cytology & histologyyesnonoyes (reaction seen through microscope in the cell)  
PPV (% proven positive when test was positive) 15-41%719%1;21%948%1;52%9  
NPV (% proven negative when test was negative)[not real good]
best report 97.5%8
sensitivity (% of true positives detected)(conventional Pap regularly repeated because of low sensitivity8; 47 % for ASCUS or higher & 51% for just LGSIL)...liquid possibly better5 791; 87%9;
78-96% for CIN2-35
specificity (% of true negatives that test negative)(conventional Pap 98% for ASCUS or higher & 95% for just LGSIL)...liquid possibly better5 561861; 89%9  
false positive7 rate  39%112%1  
false negative rate  1.6%1; 25%100.4%1  
detection of latent HPV (<age 30 high rate of spontaneous regression6)      
detection of extraneous HPV from site other than cervix (vulva, semen, penile epithelium6)  yes & one has no chance of discerning as contaminantyes & one has a good morpho-specific chance of discerning as contaminant  
tech time factorshigh & manual but conventional & in widespread use high & manual & very specializedhigh & manual & quite specializedlow..."walk away"  
other cost factorsstandardcostly lab design & reagentsmust use whole plate or strips of wells & therefore batch; a "run" mistake requiring repeat is costlyreagents metered so that almost same cost per slide for 1 or 20 cases  

References & notations:

  1. Qureshi MN, RR Tubbs, Layfield LJ, et. al. [Columbia U., Cleveland Clinic, ARUP Lab] Role of HPV DNA Testing in Predicting Cervical Intraepithelial Lesions: Comparison of HC HPV and ISH HPV, Diagnostic Cytopathology 29(3):149-155, September 2003. [10861 women had liquid based Pap smears...762 ASCUS or LGSIL reflex tested for HPV...250 acceptable HC2 positives & residual specimen submitted for CISH...72 were younger than age 30 and 178 older...accuracy determined by biopsy gold standard for "truth"]
  2. Blair Holladay, PhD, comparative methods investigator & director of MUSC School of Cytotechnology, Charleston, S. C. [personal...EBS... on-site interview and Ventana slides review 17 November 2003]. *Positive nuclei are quickly/easily, slam-dunk visible at Pap screening magnification; "positive is positive" so that even one positive nucleus is "callable"; one can easily discern two different nuclear positives: (a) episomal positive nuclear pattern(a diffuse even nuclear positivity indicating viral presence and intracellular replication but not yet inserted/integrated into chromosomes (especially under age 30, a high % of these low risk "infections" are cleared by an immunocompetant patient...almost all low risk cases are this pattern and some high risk & "highs" progress to an integrated form; and (b) integrated positive nuclear pattern(discrete nuclear dots...virus is in the chromosomal DNA & now able effect the genetic aberration...almost always a marker of a high risk HPV).
  3. Though not FDA approved, PCR was the reference test method for "truth" for A. L. T. S. and for the original sensitivity & specificity performance data for Digene's hybrid capture methodology. The FDA approval process is very expensive and beyond the budget of many companies, especially when the market is not yet settled on a test's utilization. Does the FDA approve your medical education, your medical or surgical skills, or your legitimacy for medical staff privileges in your hospital? 
  4. ASR stands for  "analyte specific reagent". This means that there has been plenty of "prior art" method utilization for other specific analytes than, say, HPV. Medical labs would grind to a halt if only FDA approved tests were legitimate (much as medications with off-label uses). Must validate the use in "your" lab against a reference method & determine that it is OK.
  5. ACOG Practice Bulletin #45, August 2003 (replaces Committee Opinion # 152 of March 1995...79 references). (a) "Although liquid-based thin-layer cervical cytology appears to have increased sensitivity for detecting cancer precursor lesions over the conventional method, the degree to which sensitivity is increased is unknown. Equally important, the difference in specificity between the liquid-based and conventional tests has not been determined." (b) "...many women harbor the [HPV] virus in their lower genital tracts without showing cytologic or histologic changes."
  6. D. A. Baunoch, PhD. [of US Labs, Irvine, Calif.], "In Search of a Paradigm", [HPV review] Advance Newsmagazine for Administrators of the Laboratory, 1 May 2001, p69-72. [studies have as yet failed to clearly prove that there is a molecular test with the advantage of required sensitivity plus morphologic identification].
  7. RM Richart, MD [panel discussion] "Combined HPV & Pap Testing: Advances in Risk Assessment", a supplement to Contemporary OB/Gyn, April 2003. [a pos. HPV test with negative colposcopy is lesion negative but not risk careful how "false positive" is used]
  8. RM Richart, MD, "Human Papilloma Virus DNA Testing as an Adjunct to Cytology in Cervical Screening Programs", Archives of Pathology and Lab. Medicine 127(8):959-968, August 2003.
  9. Qureshi MN, RR Tubbs, Layfield LJ, et. al. [Columbia U., Cleveland Clinic, ARUP Lab] "Predicting Cervical Lesions by Inform HPV and Hybrid Capture II", Modern Pathology  16(1), January 2003.
  10. NM Lonky, et. al., "Triage of Atypical Cells of Undetermined Significance With Hybrid Capture II: Colposcopy and Histologic Human Papillomavirus Correlation", Obstetrics and Gynecology 101(3(::481-489, March 2003.
  11. Wright TC, Cox JT, ASCCP monograph/booklet: "Clinical Uses of Human Papilloma Virus (HPV) DNA Testing", 2004 (EBS's office).
  12. Mohrota S & Wojcik EM, "Human Papilloma Virus- A Brief Synopsis of Methods of Detection", The ACS Bulletin, Nov. 2006.
  13. Expert practitioner comments.
  14. Plotz RD, of Harvard Vanguard Medical Associates, Forbes magazine letters, 25 Feb. 2008 page 12.

(posted 29 October 2003; latest addition 19 February 2013)

© Copyright 1999 - 2006, all rights reserved, Pathology Associates Of Lexington, P.A.