Real Time Screening Techniques In The Detection Of Cervical Cancer

Introduction

 


Cervical cancer is a largely preventable disease because of an easily recognisable and treatable premalignant stage. In countries that have implemented regular screening programmes to detect these lesions, cervical exfoliative cytology has been a successful screening tool. In many Western countries that have adopted such programmes, the incidence and mortality of cervical cancer has dropped dramatically. Eighty percent of the 466,000 cases of cervical cancer annually occur in the developing countries, where only 5% of the female population has had a Papanicolaou smear within 5 years1 and cervical cancer is the leading cause of cancer death2. In contrast, in the developed countries, where 85% have had at least one Papanicolaou smear in their life-time, it is only the tenth leading cause of cancer death2. Despite similar populations, Finland has lower incidence of cervical cancer compared to Norway; the former having a well-established cervical screening programme. Half of the women in the United States with newly diagnosed invasive cervical cancer have never had a Papanicolaou smear3, and another 10% have not had a smear in the 5 years before the diagnosis3. In the United Kingdom, after introduction of computerised call and recall system, the incidence of cervical cancer has decreased from 16.4 per 100,000 to 9.3 per 100,000 between 1988 and 1997. In the same period, mortality of cervical cancer has decreased from 7.0 to 3.7 per 100,000. Therefore, the best way to improve cervical cancer detection and reduce its incidence is by regular population-based screening.

Despite its success, exfoliative cytology is associated with certain drawbacks mainly in respect of the significant false negative rate4,5,6. An inherent disadvantage is the delay between the time the test is taken and when the results are eventually available. Consequently, alternative methods have been developed in an attempt to address this delay. Real time screening tests can reduce some of the anxiety associated with the cytological screening test by providing instantaneous results. Real time tests such as VIA (Visual Inspection with Acetic acid) and VILI (Visual Inspection with Lugol’s Iodine) have also been used in ‘one-stop cervical cancer prevention clinics’ in developing countries.


Currently available Real Time Screening Techniques

  • Technique based on electrical and optical properties of the cervix:
    • TruScreen (Polarprobe)
  • Techniques based on fluorescence from cervical tissues:
    • Medispectra
    • Lifespex
  • Visual Inspection based techniques:
    • VIA (Visual Inspection with Acetic acid)
    • VILI (Visual Inspection with Lugol’s Iodine)

Advantages of real time screening

  • Reduced anxiety
  • Improved compliance
  • Inadequate tests can be repeated immediately
  • ‘One stop’ screening, counselling and treatment thus avoiding repeated visits
  • Potential to be used for secondary screening as an adjunct or triage

TruScreen
The TruScreen is an optoelectronic probe-shaped device that algorithmically classifies cervical tissue types and provides a final patient result for each case. The TruScreen employs a real time approach to the detection of tissue abnormalities. Traditionally, screening and diagnostic tests rely on biochemical information or recognition of abnormalities based on cell morphology (e.g. cytology) or tissue structure (e.g. radiology). The TruScreen device is an in vivo system that uses the electrical and optical properties of cervical tissues to arrive at a diagnosis.

Table 1: Sensitivity and specificity of Truscreen12
Truscreen alone Cytology alone Truscreen + Cytology
Sensitivity for CIN2/3 70 % 69 % 93 %
Sensitivity for CIN1 67 % 45 % 87 %
Specificity for Normal 81 % 95 % 80 %
Figure 1: TruScreen system showing the handpiece, microcomputer console and the single use sheath.

The main components of the device (Figure 1) include a pen-shaped handpiece that contains the tissue stimulation and sensor elements. The handpiece is connected by a cable to a console that contains a microprocessor control module and a digital signal processor. The instrument uses a software-implemented tissue classifier, which provides the operator with instantaneous feedback in the form of a printed result.

Figure 2: Tip of the handpiece showing electrical and optical sensors

The handpiece is a 17cm long pen-shaped instrument with a 5mm diameter flat tip. The tip contains the tissue stimulator and sensor elements (Figure 2). These consist of three peripheral electrodes and four central light emitting diodes. The console contains a microprocessor module and a digital signal processor. The handpiece is covered by a single use sheath that is discarded after use to prevent cross-infection. The microprocessor handles the data to and from the digital signal processor that implements the complex algorithmic calculations necessary for operation of the tissue classifier.

The cell membranes and varying internal structures of any tissue form a complex of capacitors and resistors. When an electrical voltage is applied to tissue and then turned off abruptly, the tissue behaves like a decaying battery, lasting for a fraction of a second. Because both the decay time and the waveform will differ among different types of tissue, the voltage decay waveform can provide a dynamic signature of the tissue that assists in its classification7-10.

Figure 3: Electrical decay curve for tissues

TruScreen applies low-voltage pulses (0.8V of 260µs duration) to the cervix via a combination of its three electrodes, and the resultant electrical decay curve is measured and analysed. (Figure 3)

Analysis of electrical properties alone is unable to distinguish unambiguously among the different cervical tissue types, because of the degree of overlap. Additional parameters are therefore required to provide complementary information. Optical properties are used for this purpose in the form of selective wavelength spectroscopy. Four light-emitting diodes transmit light at specific wavelengths within the electromagnetic spectrum – two red diodes (600nm), a green diode (525nm), and an infrared diode (940nm). These are activated in sequence, and the tissue response is detected by a detector photodiode at the excitation as well as at the off-excitation frequency, thus producing a relatively broad-band spectrum for analysis.

By combining the electrical decay and spectroscopic information from a particular area on the cervix, TruScreen is able, by means of a classification algorithm, to categorise the tissue7,11.

The system incorporates a handpiece with a tip designed to take measurements from 5mm diameter areas of the cervical tissue. During the examination, ‘stop/go’ lights on the handpiece guide the operator to move the tip of the probe to a tissue spot, stop for the measurement to be performed, and then move to the next tissue spot. This sequence is repeated until the area of the cervical transformation zone has been covered. After the operator has signalled the completion of the examination by pressing a button on the handpiece, the screening result is calculated and printed out from the console.

The TruScreen is capable of classifying approximately 17 different basic cervical tissue types as well as junctions between different tissue types. The expert system has been ‘trained’ to recognise various normal and abnormal cervical tissue types using a previously obtained database of over 1,500 patients collected from a geographically diverse range of centres. These specific tissue type classifications are grouped in a manner useful for cervical screening, and hence the “worst case” tissue type seen on the cervix determines the final device output. The initial model of the TruScreen returns one of two possible final patient screening results – “normal” (normal squamous epithelium, columnar epithelium, physiologic metaplasia) or “abnormal” (CIN I-III, invasive cancer).

Used as a primary screening test, TruScreen has been shown in a recent study of over 700 women either referred with an abnormal smear or attending for routine screening to have a sensitivity of 70% for CIN2 and above. The specificity was 81%. In this study, a smear was obtained at the time of probing and the combined sensitivity increases to 95% without a further decline in specificity.

Lifespex (Cerviscan)
Cervical tissue fluoresces when excited at spectral regions in the Ultraviolet and visible region. This fluorescence response for normal and abnormal cervical tissues varies in intensity and spectral contents. These optical properties are assessed by Cerviscan which is a multispectral tissue fluorescence imaging system for detecting cervical intraepithelial lesions. The instrument is in its early stage of development and initial studies on pap positive population suggest that it may be a useful adjunct to colposcopy. This light-based system can provide the colposcopist with an objective fluorescence map of the visible cervix and indicate areas of abnormality for directed biopsy13. Further population-based studies are needed before the device can be used as a primary screening tool.

Medispectra
Medispectra is a light-based system, which uses a combination of laser-induced fluorescence, white light backscatter technology and a multivariate algorithm to classify normal and abnormal cervical tissues. Initial studies have shown this to be a useful adjunct to colposcopy14.

Visual Inspection based techniques (VIA and VILI)
The difficulties in implementing cytology-based screening programmes in developing countries have led to the investigation of screening tests based on visual examination of the uterine cervix. Among these tests, visual inspection with 3-5% acetic acid (VIA) involves non-magnified visualisation of uterine cervix after application of 3-5% dilute acetic acid. Visual inspection with Lugol’s iodine (VILI) involves naked eye examination of the cervix after application of Lugol’s iodine solution.

The results of test performance in cross-sectional study settings (Table 2) indicate that the sensitivity of VIA to detect high-grade precancerous lesions ranged from

Categories: ARTICLES

About Author