LBC – Sink or Swim? Thames Valley Cytology Society Bedford 24th June 2004
NICE Guidance October 2003: NICE recommends that LBC is used as the primary means of processing samples in the cervical screening programme in England and Wales
NICE Guidance There is currently insufficient evidence to recommend one LBC product over another. The NHSCSP and SCW may wish to consider evaluating further different products as the method is introduced.
Conventional Smears False Negative Rate of up to 55% Sampling and interpretative errors Borderline Rates of up to 6.4% 70% are truly negative 30% represent more severe abnormality Inadequate Rates of 9.7%
Sources Of False Negatives Sampling issues (70%) cells not collected on the sampling device cells collected, but not transferred to the slide Interpretative issues (30%) abnormal cells present on slide but either not seen or misinterpreted
The Problem The Problem The Conventional Cervical Smear Non-randomized A cervical sample containing precancerous cells (red) Non-randomized portion of cells Over 80% of cells discarded Sample may not reflect patient’s actual condition Smear spray-fixed and sent to lab Missing cells, obscuring elements
The Solution The Solution The ThinPrep® Pap Test™ Virtually 100% of cells collected into ThinPrep vial A cervical sample containing precancerous cells (red) Cells immediately preserved and sent to lab Increased opportunity to detect early signs of abnormality Filtration process disperses, randomizes cells More representative and clear thin layer of cells
Evidence Base Clinical Validation ThinPrep® Improvement Patients LSIL+ HSIL+ S.Q. Split-Sample Lee et al. 6,747 65% 36% 54% Corkill et al. 1,583 110% 54% N/A Roberts et al. 81,754 26% 15% 91% Direct-to-Vial Scottish Pilot 30,228 92% 84% 76% Diaz-Rosario 56,339 72% 103% ??? Papillo et al. 8,574 52% 55% 52% Weintraub et al. 18,247 184% 94% 93% Bolick 10,694 181% 173% 70% Johnson et al. 902 71% 125% 41% Guidos/Selvaggi 9,583 267% 233% 97% Quddus 14,216 24% 35% 28% Hornish 30,431 N/A 83% N/A >60 peer reviewed international publications from 4 Continents
Does LBC increase sensitivity? Educational bias Lab / Smear takers Study design bias Split sample / Direct-to-Vial Population bias Self-selecting / Non-screening End point bias Dyskaryosis as ‘surrogate’ / lack of histological confirmation
Does LBC decrease inadequate rates? SurePath 5000 cells Thinprep 40000 cells Conventional 50000 cells?
Scottish Pilot Results Weighted Average (total 30,228) Conventional % ThinPrep % Unsatisfactory 7.63 1.84 Borderline 3.98 3.67 Mild 1.10 2.12 92%↑ Moderate 0.45 0.97 Severe 0.65 1.06 84% ↑
Scottish Pilot Conclusions No major storage problems Reduced workload for smear takers 349/350 smear takers preferred ThinPrep® Increased productivity in laboratories Sharp reduction in the unsatisfactory rate Less anxiety and discomfort for women More appropriate referrals to colposcopy Allows scope for further developments in the screening programme
HPV Triage Negative Borderline Mild dyskaryosis Moderate/severe Cancer COLPOSCOPY/ BIOPSY HPV Testing NO TREATMENT
Vial storage Modular storage units available (File-a-Vial) 150 vial capacity Example: 30,000 samples pa 4 week cycle 2,500 vials 17 units
ThinPrep® T2000 Processor
ThinPrep® 3000 Processor
SurePath Why I like the SurePath Liquid-Based PAP Test David R Bolick RPS Sandy, Utah
SurePath SurePath and ThinPrep Similarities: Diagnostic rates Detection rates of HPV from vial Histology/Cytology correlation rates
SurePath SurePath and ThinPrep Differences: Inadequate rates False negative fraction Detection of abnormalities in clusters
SurePath SurePath and ThinPrep Inadequate rates:
SurePath SurePath and ThinPrep False negative fraction: LSIL % HSIL % SurePath screened TWICE ; ThinPrep screened ONCE
SurePath SurePath and ThinPrep Abnormalities in clusters: AGUS Endom Adenoca SurePath 0.13 0.24 0.05 ThinPrep 0.09 0.13 0.02
LBC 5 year conversion timescale is too long LBC is on the way. Major retraining exercise for labs Benefits in terms of productivity and scope for future developments Good for smear takers and women Issues around sensitivity / specificity / inadequate rates will be answered in long run Expensive solution