Your Federal Quarterly Tax Payments are due April 15th Get Help Now >>

Colposcopy and Selective Biopsy in Patients with Abnormal Cervical by mikeholy


									 47                                                    MEDlCAl        JOUR        AL                             10    eptember 1977

 13 caesarean sections in the study group. Twelve of these          oon a the djagnosis has been made and the paediatrician
 were performed for prolonged labour or disproportion.             informed beforehand. Antibiotics hould be administered
 In the control group 4 of the 6 caesarean ections were            to both mother and newborn. Although this has not been
 as ociated with these indications. Furthermore, the mean          well documented yet, it seems a if there i a rise in
 birth weight were lower than those described by Oden-             intra-uterine temperature before the onset of fetal tachy-
 daal and Crawford.' A possible explanation i that the             cardja and maternal pyrexia in ca es of intra-uterine in-
 two tudies were done in different population groups.              fection.
    The conclu ion drawn from this tudy is that a highly
 significant elevation in intra-uterine temperature occurs in         1 wi h to thank Dr H. J. Odendaal for his assistance in
                                                                   thi tudy and in the preparation of the article, as well a the
 cases of fetal tachycardia associated with intra-uterine          South African Medical Research Council for the statistical
 infection. Intra-uterine temperature mea urements could           analysis.
 be applied as a diagnostic aid to distinguish infection from                                  REFERENCE
 other causes of fetal tachycardia. Owing to the high              J. Beard, R. W., Filshie, G. M., Knight, C. A. et al. (1971): J. Obstet.
 neonatal morbidity and even mortality, as well as maternal           Gynaec. Brit. Cwlth, 78, 65.
                                                                   2. Walker, D., Walker, A. and Wood, C. (1969): Ibid., 76, 503.
 postpartum morbidity, the early diagnosis of intra-uterine        3. Wood, C. and Beard, R. W. (1964): Ibid., 71, 76 .
                                                                   4. Odendaal, H. J. and Crawford, J. W. (1975): S. AfT. med. J.,
 infection i mandatory. The fetu should be delivered a                49, 1 73.

 Colposcopy and Selective Biopsy in Patients with Abnormal
                    Cervical Cytology
                                      V. K. !U\T(JTZE       T     A. G. B. SHERWOOD

                         SUMMARY                                   Colposcopy was first introduced by Hinselman1 in 1925
                                                                   and was soon advocated as a primary method of early
  Patients with atypical or positive findings on cervical cyto-
                                                                   cancer detection. This technique has since gained wide-
  logy should be referred to a special colposcopy clinic
                                                                   spread acceptance in Europe,' but until recently has not
  as the next step in investigation. Colposcopy complements
                                                                   enjoyed a similar reputation in the UK or the USA.'
  cytology, and when combined with selective biopsy of             With improved technology and photographic equipment,
  the worst-affected area allows a high level of diagnostic
                                                                   and a better understanding of the role of colposcopy in
  accuracy (90,7%). The necessity for diagnostic conization
                                                                   the diagnosis of pre-invasive disease of the cervix, re-
  with its risks is markedly reduced. When all three
                                                                   newed interest has been shown over the last decade.'
  modalities were used in combination, only 0,7% of in-
                                                                      Cervical cytology remains the best screening method
  vasive cancers were missed.                                      for premalignant disease of the cervix. However, the
                                                                   increasing use of colposcopy as an adjuvant to cytology,
  S. Air. med. l., 52, 478 (1977).                                 and not as a competitor, is confirmed in this study, as
                                                                   it has been by others."·
                                                                      Selective biopsy is an integral part of colposcopic exam-
                                                                   ination, allowing the _wor t-affected areas to be sampled
                                                                   and ent for histological examination. The ample can be
                                                                   examined by conventional light microscopy, by electron
                                                                   microscopy, or by canning electron microscopyY It mu t
                                                                   be emphasized that cytological examination is a laboratory
                                                                   technique, whereas colposcopy is a clinical method, and
Date received: 12 May 1977.                                        each evaluates a different aspect of neopla ia"
    10 September 1977                          s        EDIE       E     TYD      KRIF                                                   479

                PATIENTS AND METHODS                                   When a latitude of one histological diagno tic grade wa
                                                                       allowed (e.g. cytology uggested severe dyspla ia and the
     In 1974 a colposcopy clinic was started at Groote chuur           final diagnosi was carcinoma in situ), a ha been u ed
    Hospital, and since 1975 colpograms have been utilized             by others·'· to indicate clinical accuracy, the correlation
    for the accurate recording of data. All patients with an           between the three techniques and the final histological
    atypical or positive cervical smear, or an abnormal cervix
                                                                       diagno is was 96,Oo~ (Table Il).
     with a negative cervical smear, are currently referred to
     this clinic.
        Patients with abnormal findings on cytology undergo            TABLE 11. CORRELATIONS BETWEEN CYTOLOGY, COLPO-
    colposcopy and selective biopsy, and receive definitive              SCOPY, SELECTIVE BIOPSY, AND FINAL DIAGNOSIS
    treatment according to the results of these. All colpo-
                                                                       Cytology correlated with:                 Number          Percentage
    scopic examinations are performed with a Zeiss Colpo-
    scope after the cervix has been cleansed with J% acetic              Colposcopy                              88 (134)*        58,7 (89,3)
    acid. Selected biopsy specimens are taken with an Eppen-             Selective biopsy                        91 (131)         60,7 (87,3)
    dorf forceps. All patients with a carcinoma in situ receive          Final diagnosis                        100 (134)         66,7 (89,3)
    definitive treatment, either hysterectomy or therapeutic           Colposcopy correlated with:
    conization of the cervix.                                            Selective biopsy                        93 (145)         62,0 (96,7)
       The results in the first 150 patients who have had                Final diagnosis                         96 (140)         64,0 (93,3)
    successful colposcopy, biopsy and definitive treatment,            Selective biopsy correlated with:
    and a final diagnosis based on these, form the material              Final diagnosis                        110 (144)         73,3 (96,0)
    for this study. All failed colposcopies (approximately 10%         • Clinical accuracy allowing latitude of one histological diagnosis.
    of patients seen at the clinic) have been excluded from
    this study. Failures included all patients in whom the
    squamocolumnar junction could not be visualized in                    A good correlation between cytology and colposcopy
    its entirety or in whom abnormal transformed epithelium            is shown in Fig. 1 (a), but with a tendency to a falsely
    extended up the endocervical canal out of the view of the          high recording of invasive carcinoma by cytology. The
    co Iposcopist.                                                     correlation between colposcopy and selective biopsy
       All cytology smears and histology reports were taken            (Fig. 1 (b)) is closer, but often in colposcopic evaluation
    from routine service reports and were not reviewed                 it was difficult to distinguish between severe dysplasia and
    by a single observer. However, all colposcopies were per-          carcinoma in situ.
    formed by one of us. All cytology smears, colposcopic                 The correlation between cytology and the final diag-
    evaluations, colposcopic biopsies and the final histological       nosis again shows the tendency to false positive recording
    examinations were graded into one of the following his-            of invasive carcinoma by cytology (Fig. 1 (c)). In Fig. 1 (d)
    tological groups: (i) mild dysplasia; (ii) moderate or             the tendency to underestimate carcinoma in situ by colpo-
    severe dysplasia; (iii) carcinoma in situ; (iv) carcinoma          scopy and designate it as severe dysplasia, as in Fig. I (b),
    in situ with micro-invasion; (v) invasive carcinoma.               is again shown. The correlation between colposcopic
                                                                       biopsy and the final diagnosis is good (Fig. 1 (e)).
                                                                          Evaluation of all diagnostic methods often depends on
                            RESULTS                                    the failure rates or on the ability not to miss serious
    Of the 150 patients reviewed, 17 had atypical Papanicolaou         lesions, rather than on the correlations with different
    smears (1 mild dysplasia; 16 moderate / severe dysplasia),         methods or the final diagnosis.
    and 133 had positive smears, suggesting either carcinoma              If clinical accuracy allowed one histological degree
    in situ (106), carcinoma in situ with micro-invasion (8),          difference, as used by Stafl and Mattingly: the false
    or invasive carcinoma (19).                                        negative rate was 2,7°{, for cytology, 2,7% for colposcopy,
       The final diagnosis, made after definitive treatment,           and 0,7% for colposcopic biopsy, when compared with
    was correctly predicted by cytology in 66,7,0{" by colpQ-          the final diagnosis. The false positive rate was 8,0% for
    scopy in 64,0% and by colposcopic biopsy in 73,3°{, of             cytology, 4,O°{, for colposcopy, and 3,3°(' for colposcopic
    patients (Table I). When cytology and colposcopy were              biopsy when compared with the final diagnosi (e.g. if
    combined, the final diagnosis was accurately predicted in          cytology recorded invasive carcinoma and the final diag-
    82,7°{, of patients, and when cytology, colposcopy and             nosis was carcinoma in situ with two histological degrees
    colposcopic biopsy were combined, in 90,1% of patients.            of difference this would be a false positive for cytology).
                                                                       The false positive rate for cytology is twice as high a
      TABLE I. ACCURACY OF PREDICTION BY CYTOLOGY,                     for colposcopy or selective biopsy.
    COLPOSCOPY, SELECTIVE BIOPSY AND COMBINATIONS                         More important than the false negative and false
                     IN 150 PATIENTS                                   positive rate are the cases in which invasive carcinoma
                                                                       was missed by one or more diagnostic method . In this
                                           Number    Percentage        study invasive carcinoma was mis ed by cytology in 4
    Cytology                                100         66,7           patients (2,7<>(,), by colposcopy in 3 patients (2,0%) and
    Colposcopy                               96         64,0           by selective biopsy in 2 patients (1,3 %). These were not
    Selective biopsy                        110         73,3           all the same patients and thus it is profitable to examine
    Cytology/colposcopy                     124         82,7           the 5 individual case report. a important lessons can be
    Cytology/colposcopy/selective biopsy    136         90,7           learnt from them.
 480                                                                     s.           MEDICAL                               JOUR                AL                                             10   eptember 1977

                                                                                                                  ';W Ic. o!I c. 0 IINlS
                                                                         ~                            1lI1~~
                                                                                                                                                 tl        r~                                  1WICo 0
                         MILD IIDD               C. D C.                                                                                                                                                  C.O   C.
                           DYS    m Coo           (1lI1 r.IlS       9    Ol-
                                                                                                          D        DYS        11lI)
                                                                                                                                     c.                                       NEG -.D
                                                                                                                                                                                   DYS DYS                1.)   1-    ?
                                                                                                                        I                                                                  I
    NEGATIVE                                                        0    NEGATIVE
                                                                                                                        I              I         'i        .GATlVE

    IlIILD DYS                              I                       I    IlIILD DYS                                                              1     I   IlIILD DU                            I                     1
    IlIOD&sEV~                    1        1       I            15       MDO&SEVovJ                       2        1        30    2    I         42        lIDO &SEV on,                        4    11    I         1I

    C.O                           31       6li     1     )      101      C.O                              I        I        11    1             1)1        C. 0
                                                                                                                                                                                   I   I       13   12     1    2
    C. D IIll)
                                       I    2      I            I   1    c. 0 11lI)
                                                                                                                             I    I        2     14        c.O IIlII
                                                                                                                                                                          I                I
                                                                                                                                                                                                     1     I         I!
    C.INVAS                       4    I    1            1      I!       C.INVAS                              I    I         I             !
                                                                                                                                                 "l        C.I....AS
                                                                                                                                                                                   I            I   10     I    1    I!

    TDTAL          0      0       42       13     14     11     150      TOTAL                   0        3        15       105   II       11   1nl        TOTAL               I       I       I!   110    10   !    150

         IO!      NEG (1lIILD I ~ c. 0
                           DY DYS
                                                 IIlII 'NC:J        ¥I   r.:.
                                                                          ~           ill
                                                                                                NEG IlIIL~-m C.O 11lI) IIII/Al 0
                                                                                                                 c. D C.
                                                                                                                               T                           D       Ah...... Co"ololion

                                       I                        I0       NEGATIVE
                                                                                                                                       I         0

                                                                                                                                                                   Cli"iC81 Accuracy
    MILD OYS                                                        0    IlIILD DYS                                3                             3                 I Histological latitude
   MOO & SEVDYS           I       13       21      2            42       MOO &SEVDYl                      I        I

                                                                                                                             I                   15
                                                                                                                                                           D       No Co".lotio.
   C.O             I      I
                                  I        11      3      I     13       C. D
                                                                                            I I
                                                                                                 I                          I!    5        I     104

   C.OIMI                                   1      I      I     14       c.O (Ml                          1                 10    I        I     II

   C.INVAS                                 4       I     I          11   C. INVAS
                                                                                            I                               3              1     ,0
                                                                                            I         I
   TOTAL           I I 2          I!       101    12     1      15D      TDTAL                   I        2        I!       101   11       !    150

    Fig. 1. Correlations between cytology, colposcopic examination, selective biopsy and final diagnosis. (Dys.                                                                                      =    dysplasia;
    Ca. 0 = carcinoma in situ; (M) = micro-invasion; Ca. In vas. = invasive carcinoma.)

   Patient I, aged 39 years, was referred by the rational                                                         picked up by selective colposcopic biopsy. Clinically this
 Cancer As ociation      CA). Cytology was positive and                                                           represents an occult invasive carcinoma.
  uggested carcinoma in situ. The smear was assessed with                                                            Patient 4, aged 53 years, presented with a cytology smear
great difficulty owing to a messy background. Colposcopy                                                          suggesting invasive carcinoma. Colposcopy revealed only
showed frank invasive carcinoma, which was confirmed                                                              carcinoma in situ, but biopsy of the worst area showed
by selective biopsy. The final hysterectomy showed no                                                             invasive carcinoma, as did the hysterectomy specimen.
residual invasive carcinoma but extensive carcinoma in                                                               Conclusion: This invasive carcinoma was missed by
situ.                                                                                                             colposcopy but not by the other methods.
   Conclusion: Cytology can miss a small early lesion,                                                               Patient 5, aged 43 years, presented when she was 36
possibly because of incorrect taking or fixing of the                                                             weeks pregnant with a cytology smear suggesting carcinoma
smear." This invasive carcinoma was totally excised by                                                            in situ. Colposcopic examination revealed a small area of
colposcopic biopsy.                                                                                               invasive carcinoma and this was confirmed by selective
   Patient 2, aged 54 years, was referred by the      CA.                                                         biopsy and by caesarean/ hysterectomy.
Cytology was positive, suggesting a non-keratinizing                                                                 Conclusion: Cytology smears in late pregnancy are
squamous carcinoma in situ. Colposcopy confirmed this,                                                            difficult to take adequately and to interpret. Colposcopy
as did the selective biopsy. The hysterectomy showed                                                              has a special place in pregnancy. This carcinoma was
carcinoma in situ, a small focus of invasive carcinoma in                                                         missed by cytology.
one area.                                                                                                            Tn the 5 patients described, only 2 invasive carcinomas
   Conclusion: This very small invasive carcinoma was                                                             were missed by all three methods (in patient 2 a very
missed by all three methods and was clinically stage                                                              small carcinoma and in patient 3 an occult small-cell
la carcinoma of the cervix.                                                                                       carcinoma). Since the former was preventable and the
   Patient 3. aged 39 years, was referred by the      CA.                                                         latter not, the true 'missed' invasive carcinoma rate is
Cytology was positive, suggesting large-cell, non-kera-                                                           0,7°£ (i.e. 1: 150).
tinizing squamous carcinoma in situ. Colposcopy confirmed
this and selective biopsy showed carcinoma in situ ex-                                                                                                     DISCUSSIO
tending deep into gland. The hysterectomy confirmed
extensive carcinoma in siru, and in one area deep into                                                            Cytology is still the best screening method for the detection
the epithelium an invasive small-cell carcinoma which                                                             of premalignant and malignant disease of the cervix.···
approached the epithelium but did not reach it.                                                                   Cytology should have a low false negative rate (2,7%
   Conclusion: This invasive carcinoma will always be                                                             in this study) and a low 'missed' invasive cancer rate
missed by both cytology and colposcopy. It may be                                                                 (2.7°£ in this study), but it is acceptable to have a high
10 September 1977                          SA    MEDIESE         TYDSKRIF                                                            4 1

false positive rate (8,0% in this study). The false negative    of view of the colposcope.'· Allowing for clinical accuracy.
rate recorded in this study only represents the under-          the false negative colposcopic biop y rate wa only 0,7%
prediction rate in a group with abnormal cytology, and          in this study. Under these circum tances, colpo copy i
thus may not be representative of the true false negative       a failed procedure and an endocervical curettage i em-
rate if the entire population were to be screened.              ployed to aid accurate diagnosis, a ha been advocated
   Colposcopy is a useful adjuv,ant to cytology'·'u and         by others."'>
should be the next step in the investigation of a patient          The histological groupings used in thi tudy are not
with an atypical or positive cytological smear or an ab-        ideal; a better grouping to classify premalignant lesion
normal cervix. A similar programme is used by others."·'u       of the cervix would be: (i) mild/moderate dy pia ia;
It is impractical to use colposcopy on every patient being      (h) severe dysplasia/ carcinoma in situ; (ih) micro-invasive/
screened for premalignant or malignant disease of the           invasive carcinoma. All diagnostic modalitie - cytology,
cervix, since it is a skilled procedure requiring specific      colposcopy, selective biopsy and the final histology -
training"·'» and the equipment is costly.' We believe all       could be related to one of these three group. Thi would
gynaecologists should understand the principles of tbis         allow a more rational approach to treatment and eliminate
procedure but not necessarily be able to perform colpo-         many of the minor difficulties, such as the colposcopic
scopy (the analogy with ultrasound is not inappropriate).       difference between a severe dysplasia and an early carci-
Others disagree.,,1. In the use of any scientific instrument,   noma in situ. The treatment for each of these groups
the accuracy of the method is directly related to the           differs markedly.
expertise of the person who uses it.' When colposcopy is           The ability to rule out the presence of inva ive carci-
combined with cytology, the final diagnosis is predicted        noma is the single most important factor in the selection
more accurately than by either technique used alone; tbis       of cytology, colposcopy and selective biopsy.'
is well shown in tbis study and in others."'··,l.                  The employment of a programme whereby all patients
   Cytology, colposcopy and colposcopic biopsy are com-         with an abnormal cytology smear or abnormal cervix
plementary, and produced a combined accuracy in final           are referred to a colposcopy clinic for examination and
diagnosis of 90,7% in this study. Ostergard and Gondos'         selective biopsy of the worst area should allow a very
found that by using a combination of colposcopy, directed       accurate degree of diagnosis, provided the extent of the
biopsy and endocervical curettage, the possibility of in·       abnormal epithelium can be adequately visualized.
vasive carcinoma of the cervix could be ruled out in
85% of patients. Others·,13 show that combined cytology and
colposcopy produce an accurate diagnosis in over 85%.
   Selective colposcopic biopsy of the worst area is an          J. Hinselman, H. (1925): Miinch. med. Wschr.• 72. 1733.
                                                                 2. Sonek. M. and     ewton. M. in Taylor. M. L. and Green. T. H .•
integral part of the colposcopic technique and should               eds (1975): Progress in Gynaecology, vol. VI, pp. 217 - 233. New
always be performed;' it complements the colposcopic                York: Grune & Stratton.
                                                                 3. Scott. J. W .• Brass. P. and Seekinger. D. (196): Amer. J. Obstet.
examination, markedly reducing the need for diagnostic              Gynec.. 103. 925.
                                                                 4. Stall. A. and Mattingly, R. F. (1973): Obstet. and Gynec.• 41. 168.
conization of the cervix with its inherent risks.,,··13." The    5. Ostergard. D. R. and Gondos. B. (1973): Amer. J, Obstet. Gynec.•
poor correlation between colposcopic biopsy and the cone            115. 783.
                                                                 6. Benedet. J. L.. Boyes, D. A .• Nichols. T. M. et al. (1976): Brit. J.
biopsy bistology found by Crapanzano" was due to his                Obstet. Gynaec., 83, 177.
                                                                 7. Murphy. J. F .. Jordan. J . A .• Alien. J. 1. et al. (1974): J. Obstet.
failure to allow for the limitations of colposcopic exam-           Gynec. Brit. Cwlth. 81, 236.
ination and biopsy. If strict criteria are used and failed          Murphy. J. F., Allen. J. M., Jordan, J. A. et al. (1975): Brit. J.
                                                                    Obstet. Gynaec.. 82. 44.
colposcopies excluded, excellent correlation can be ob-          9. The Editor (1975): S. Afr. Cancer Bull .• 19. 173.
tained, as shown in tbis study and in others..,1' Definitive    10. Fritsches. H. G. and Busch. W. E. (1977): Acta cytol. (philad.).
                                                                    21, 10.
therapy can be instituted without the need for healing          11. Towsend. D. E.• Ostergard. D. R., Mishell. D. R. et al. (1970):
                                                                    Amer. J. Obstet. Gynec.. IOS. 429.
of the conized area, especially in invasive carcinoma of the    12. Ortiz, R..    ewton. M. and Langlois. P. L. (1969): Obstet. and
cervix.                                                             Gynec.• 34. 303.
                                                                13. Claman. A. D. and Lee. . (1974): Amer. J. Obstet. Gynec.• 120, 124.
   The false positive colposcopic biopsy rate compared          14. Davis. R. M., Cooke. J. K. and Kirk. R. F. (1972): Obstet. and
                                                                    Gynec., 40, 23.
with the final bistological result was due to an excision       15. Creasman. W. T. and Parker. R. T, (1975): Clin. Obstet. Gynec.,
biopsy being performed. A false negative colposcopic                18. 233.
                                                                16. Krumholz. B. A. and Knapp. R. C. (1972): Obstet. and Gynec..
biopsy may occur because the area has been missed or                39, 22.
                                                                17. Crapanzano. J. T. (1972): Amer. J. Obstet. Gynec.• 113, 967.
because the lesion extends up the endocervical canal out        18. Donohue. L. R. and Meriwether. W. (1972): Ibid., 113, 107.

To top