Pap smear accuracy for the diagnosis of cervical precancerous lesions

Abstract

This cross-sectional descriptive study, aimed at accessing the accuracy of Pap smear in diagnosing cervical precancerous lesions, was carried out between 3 January and 30 April 2017. All women screened for cervical dysplasia by means of Pap smear with biopsy done for confirmation were subsequently recruited. Data were analysed using SPSS 20.0. A total of 231 women were screened for cervical dysplasia using Pap smear with 75 biopsies performed. Cervical dysplasia was noticed in 54 cases. The sensitivity, specificity, positive predictive and negative predictive values of Pap smear were 55.5%, 75%, 88.2% and 33.3%, respectively. The sensitivity of Pap smear remains low. Therefore, biopsy should be done in cases of macroscopic cervical architectural changes irrespective of the result of the Pap smear. Moreover, to reduce the number of women with cervical precancerous lesions, the government should make available financial resources to set up HPV vaccination programmes rather than screening programmes.

Keywords

Cervical precancerous lesions negative predictive value Pap smear positive predictive value sensitivity specificity

Introduction

Cervical cancer is the second most frequently diagnosed gynaecological cancer worldwide.¹ An estimated 527,600 new cases of cervical cancer were reported worldwide in 2012, with around 265,700 deaths. Approximately 90% of cervical cancer deaths occurred in developing parts of the world.² It is the third leading cause of cancer death among women in less developed countries. The highest incidence and mortality rates are reported in Africa.²

The main risk factor for cancer of the cervix is human papilloma virus (HPV) infection. Cervical cancer may be prevented, or at least diagnosed at an early stage, given that the cervix is an organ easily accessible to clinical evaluation. Mortality from cancer of the cervix may be avoided, if not reduced, if precancerous lesions are diagnosed and treated. Cervical intraepithelial neoplasia (CIN) 1 needs around 10–15 years to evolve to the stage of cervical cancer.³ Immunosuppression may reduce this time interval.

Cervical dysplasia may be diagnosed using Pap smear. All epithelia shed and superficial cells may be collected and analysed. Pap smear consists of collecting shed superficial cells of the transformation zone and their subsequent examination by a cytopathologist.

The sensitivity of Pap smear is <70% in many studies.^4–6 Therefore, Pap smear might not be an appropriate test for the diagnosis of cervical precancerous lesions. Consequently, some cases of cervical dysplasia might unknown evolve to cervical cancer, especially when this method has been used for screening. The techniques for diagnosing cervical precancerous lesions have improved. Indeed, the liquid medium cytology technique has been recently used but with conflicting results.^7,8

No recent study evaluated the reliability of Pap smear in the diagnosis of cervical precancerous lesions in our environment, hence this study which aimed at evaluating its accuracy in the diagnosis of cervical dysplasia as well as looking at the sociodemographic profile of women affected by cervical dysplasia.

Methods

This cross-sectional descriptive study was carried out in two university teaching hospitals between 3 January and 30 April 2017. All women who were screened for cervical dysplasia by means of Pap smear were recruited. Biopsy of the cervix was performed in all women with abnormal results as well as in women with normal results, but with macroscopic architectural cervical changes. Women with cancer at biopsy were excluded when analysing the accuracy of screening of cervical precancerous lesions. When calculating the accuracy of Pap smear, women who refused cervical biopsy because they could not contribute financially to biopsy too were excluded. Informed consent was obtained from each patient. This study received approval from the national ethics committee.

The variables recorded on a questionnaire included the patient’s age, age at first sexual intercourse, age at first delivery, number of pregnancies, abortions and deliveries, cumulated number of sexual partners and tobacco consumption. These variables were chosen because we hypothesised that they could influence the occurrence of cervical dysplasia. Other variables studied were the macroscopic appearance of the cervix, the result of Pap smear and that of cervical biopsy.

Our minimal sample size of 44 patients was calculated using the following formula⁹ for descriptive studies

N = P (1 - P) Z α^{2} / D^{2}

where Zα = 1.96 corresponds to a confidence level of 0.05, D = 0.08 is the degree of precision and the prevalence of cervical dysplasia (P) was 7.9% in our milieu.¹⁰

The variables of women who presented dysplastic lesions were compared to those who did not present any dysplastic lesions. Data were analysed using SPSS 20.0. The t-test was used to compare continuous variables and Fisher’s exact test to compare categorical variables. P < 0.05 was considered statistically significant. Moreover, the accuracy of cervical Pap smear in diagnosing cervical dysplasia was assessed.

Results

During the study period, 231 women (186 having macroscopically normal cervices and 45 slight macroscopic cervical architectural changes) were screened for cervical dysplasia using Pap smear. Of the 186 women with a macroscopic normal cervix, 79 had cervical dysplastic cells, but only 30 accepted cervical biopsies. Of these 30 women, cervical dysplasia was found in 23 cases. In seven cases, there was no dysplasia.

All the 45 women with macroscopic cervical architectural changes (43 erythematous cervices and two small polypoid lesions) had colposcopy directed cervical biopsy. Among these women, Pap smear was normal in 35 cases, and revealed the presence of dysplastic cells in 10 cases. Among these 10 women, biopsy showed four cervical dysplasia, two cases without dysplasia and four cases of microinvasion. Among the 35 women who presented macroscopic cervical architectural changes without dysplastic cells found at Pap smear, biopsy showed 27 cervical dysplasia, seven cases without dysplasia and one case of microinvasion (Figure 1).

Figure 1.

Flowchart showing screening sequences.

In summary, 75 biopsies were carried out with 54 showing cervical dysplasia (54/231 or 23.4%), 16 non-dysplastic lesions as well as five micro-invasive lesions (2.1%) that were excluded when analysing the accuracy of screening of cervical precancerous lesions. The sociodemographic and obstetrical variables of the study population is presented in Table 1.

Table 1.

Sociodemographic and obstetrical variables among the population under study.

Variables	Women with cervical dysplasia (n = 54)	Women without cervical dysplasia (n = 16)	P value
Patient age (years)	43.5 ± 9.8 (24–63)	43.0 ± 11.0 (26–63)	0.862
Age at 1st intercourse	16.7 ± 2.7 (11–25)	16.0 ± 3.1 (12–23)	0.381
Age at 1st delivery	20.0 ± 4.3 (14–28)	19.1 ± 2.1 (16–23)	0.423
Abortions (n)	1.3 ± 1.3 (0–7)	2.0 ± 1.0 (0–4)	0.051
Deliveries (n)	4.9 ± 2.3 (0–10)	6.1 ± 3.2 (2–12)	0.099
Sexual partners (n)	7.2 ± 2.6 (1–12)	6.0 ± 2.1 (1–10)	0.096
Tobacco consumption* (n (%))	3/54 (5.6)	0 (0)	0.581
Previous cancer screening (n (%))	25/54 (46.3)	7/16 (43.7)	1

Values are presented as mean ± SD (range) unless specified otherwise.

One active and two passive tobacco consumptions.

Results of Pap smear, presented according to Bethesda system, was abnormal (cells showing abnormal nucleus activity) in 34 cases (14.7%). They included 11 (32.3%) high-grade squamous intraepithelial lesions (HSIL), 13 (38.2%) low-grade squamous intraepithelial lesions (LSIL) and 10 (29.4%) atypical squamous cell of undetermined significance (ASCUS). Furthermore, 22 with chronic cervicitis were observed. In 14 cases, no abnormality was found. Of the 231 participants, 107 women (46.3%) had Pap negative smears a few years earlier, while 124 women (53.7%) had never been screened before.

The results of cervical biopsies are presented in Table 2. Biopsy showed 54 cervical intraepithelial neoplasia (CIN) (20 CIN 1, 21 CIN 2 and 13 CIN 3) and no dysplasia in 16 cases (Table 3). Biopsy among the 10 women with ASCUS revealed two cervical dysplasia (20%), including one CIN 1 and one CIN 2. The eight other women with ASCUS had either chronic inflammation (three cases) or normal epithelium (five cases).

Table 2.

Cervical non-dysplastic abnormalities after biopsy.

Abnormality	n	%
Inflammation	13	81.2
Polyp	3	18.8
Total	16	100

Table 3.

Accuracy of Pap smear in diagnosing cervical dysplasia.

Pap smear	Cervical dysplasia on biopsy specimen
Pap smear	Present	Absent	Total
Positive	30 (a)	4 (b)	34 (a + b)
Negative	24 (c)	12 (d)	36 (c + d)
Total	54 (a + c)	16 (b + d)	70 (a + b + c + d)

Sensitivity: a/a + c = 30/54 = 55.5%; specificity: d/b + d = 12/16 = 75%; positive predictive value: a/a + b = 30/34 = 88.2%; negative predictive value: d/c + d = 12/36 = 33.3%.

Table 3 also shows the accuracy of Pap smear in the diagnosis of cervical dysplasia.

All the women screened collected their results (reports of Pap smear and biopsy). All patients with cervical precancerous lesions at biopsy had either cryotherapy (36 cases) or loop electrosurgical excision procedures (18 cases).

Discussion

Our prevalence of abnormal Pap smear (14.7%) is higher than the 6.0% found in Thailand among pregnant women.¹¹ The most common abnormal Pap smear was LSIL in our study (38.2%). LSIL was also the most common Pap smear abnormality observed in Thailand (44%).¹¹ Our data showed that cervical dysplasia is frequent among women screened for cervical precancerous lesions. Our hospital-based prevalence of cervical dysplasia (23.4%) is higher than that of 7.9% observed in our environment in a previous study.¹⁰ Our high prevalence might be attributed to the fact that it is the prevalence among women who consulted in our unit and not among the general population.

The mean age of patients observed in our study (43.5 years) was higher than that of 35.7 years observed in Pakistan.¹² This can be explained by the fact that their youngest patient was 19 years old, while the youngest patient in our study was 24 years old. According to some societies, screening should concern women aged 25–69 years.^13,14 Our patient aged 24 years first had sexual intercourse at the age of 11 years. She might have presented cervical dysplasia many months before, given that it was the first time she was screened. We think that screening should start earlier, for instance at 21 years, as done by Ryu et al. in the Korea University Medical Center, Seoul, Republic of Korea,¹⁵ especially among women whose first sexual intercourse occurred before the age of 12 years. In our study, another patient first had sexual intercourse at the age of 12 years. Age at first intercourse is decreasing, with some adolescents abusing children at the age of eight or nine years.^16,17

No significant difference was observed between women with cervical dysplasia and those with non-dysplastic lesions as concerns maternal age (P = 0.862), tobacco consumption (P = 0.581), age at first sexual intercourse (P = 0.381), abortion number (whether spontaneous or induced) (P = 0.051), age at first delivery (P = 0.423), delivery number (P = 0.099) and number of sexual partners (P = 0.096). This suggests that the previously described risk factors for cervical dysplasia are only co-factors.¹⁸ The main risk factor, not investigated in this series, is HPV infection that is mostly spread by sexual intercourse.

The presence of cervical invasion or micro-invasion among some women is explained by the fact that the majority of our women (53.7%) never had cervical screening combined with adequate treatment. Moreover, the diagnosis might have been missed in a previous screening among women who had a prior Pap smear. Indeed, 107 women (46.3%) were screened negative for Pap smear a few years before. Women should be encouraged to perform Pap smear or other screening methods more often. They should also be prepared for immediate treatment of positive cases.

The presence of dysplasia and cancer among women with macroscopic cervical architectural changes negative for Pap smear reminds us that biopsy should be directly carried out in such cases, i.e. when the cervix looks abnormal.

The sensitivity, for women who had also a biopsy, of a Pap smear in our study (55.5%) was similar to that of 52% found by Bhattacharyya et al. in India,⁴ but higher than the 18.2% observed in Iran⁶ and lower than the 63.2% found by Bobdey et al. in India.⁵ This shows that biopsy should be preferred in certain cases such as those with macroscopic cervical architectural changes. This low sensitivity also reveals that Pap smear is ineffective in reducing mortality from cancer of cervix. The government should look for financial resources to put an emphasis on HPV vaccination rather than on screening.

The specificity of Pap smear noticed in our study (75%) was similar to the 76% observed in Turkey.¹⁹ The negative predictive value of Pap smear observed in our series (33.3%) was lower than the 71.3% found in Iran⁶ and the 92% noticed in Turkey.¹⁹

For cases with slight macroscopic cervical architectural changes, colposcopy-directed biopsy might be associated or substituted to Pap smear to reduce the false-negative rate, as suggested by some authors.¹⁹ Moreover, this attitude shortens the proceedings.

The presence of ASCUS was observed in 10 women in our series. Some societies believe that women with ASCUS or LSIL could just be observed and followed up every six months, and colposcopy-directed biopsy is advised after three consecutive abnormal smears.^14,20 In low-resource countries, for financial or cultural reasons, many women do not routinely attend hospital. If they do, some of them are lost to follow-up.¹³ We think that women from low-resource countries with ASCUS should be immediately investigated (colposcopy-directed biopsy), as it has been associated in our study with 20% of cervical dysplasia. This point of view is shared by some authors in Japan who observed cases of CIN 3 and carcinoma in-situ among women diagnosed with ASCUS.²⁰ Screen-and-treat approaches for cervical cancer prevention have been proven effective in low-resource settings.²¹ World Health Organization (WHO) guidelines for screening and treatment of precancerous lesions for cervical cancer prevention recommend colposcopy with or without biopsy among women with ASCUS living in high HIV endemic regions.²²

The limitation of our study is the absence of tests for diagnosing HPV as this might have shown its association with cervical dysplasia. Moreover, we could not be certain of the reliability of the answers given by women, especially when it concerned the number of sexual partners. Finally, cervical biopsies were not performed in all cases with abnormal Pap smear, because some women could not contribute financially to both Pap smear and biopsy, as biopsies might have increased the size of our population with cervical dysplasia. Indeed, women with low socioeconomic status are at highest risk of cervical cancer.

Conclusion

The accuracy of Pap smear in diagnosing cervical precancerous lesions remain low. To reduce the rate of undiagnosed cervical dysplasia, colposcopy-directed biopsy should replace or be associated to Pap smear among women with macroscopic cervical architectural changes, as suggested by some authors.¹⁹ Screening might start 10 years after the first sexual contact. Finally, given that the sensitivity of Pap smear remains low, the authors think that after two normal Pap smear results, it could be repeated yearly or every two years rather than every three years, as recommended,¹⁴ if resources are available. Preferably, HPV vaccination might be the method of choice, especially in resource-limited countries, as it will be more cost-effective in preventing cervical precancerous lesions and, therefore, cervical cancer.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Parkin

Bray

. The burden of HPV related cancers. Vaccine 2006; 24(Suppl. 3): S11–S25.

Torre

Bray

Siegel

et al.

Global Cancer Statistics, 2012. CA Cancer J Clin 2015; 65: 87–108.

Moscicki

Schiffman

Kjaer

et al.

Updating the natural history of HPV and anogenital cancer. Vaccine 2006; 24(Suppl. 3): S3/42–S3/51.

Bhattacharyya

Nath

Deka

. Comparative study between pap smear and visual inspection with acetic acid (VIA) in screening of CIN and early cervical cancer. J Midlife Health 2015; 6: 53–8.

Bobdey

Balasubramanium

Kumar

et al.

Cancer screening: should cancer screening be essential component of primary health care in developing countries?

Int J Prev Med 2015; 6: 56–56.

Karimi-Zarchi

Zanbagh

Shafii

et al.

Comparison of Pap smear and colposcopy in screening for cervical cancer in patients with secondary immunodeficiency. Electron Physician 2015; 7: 1542–1548.

Lertvutivivat

Chanthasenanont

et al.

Silent high grade cervical intraepithelial neoplasia in atypical smears from liquid based cervical cytology - three years experience in Thammasat University Hospital. Asian Pac J Cancer Prev 2016; 17: 4353–4356.

Latsuzbaia

Hebette

Fischer

et al.

Introduction of liquid-based cytology and human papillomavirus testing in cervical cancer screening in Luxembourg. Diagn Cytopathol 2017; 45: 384–390.

Kieser

Friede

. Sample size in internal pilot study designs with control of the type I error rate. Statist Med 2000; 19: 901–911.

10.

Tebeu

Sandjong

Nkele

et al.

[Cervical precancerous lesions in a rural area] [article in French]. Med Afr Noire 2005; 52: 27–31.

11.

Lertcharernrit

Sananpanichkul

Suknikhom

et al.

Prevalence and risk assessment of cervical cancer screening by Papanicolaou smear and visual inspection with acetic acid for pregnant women at a Thai provincial hospital. Asian Pac J Cancer Prev 2016; 17: 4163–4167.

12.

Khan

Sultana

Jabeen

et al.

Visual inspection of cervix with acetic acid: a good alternative to pap smear for cervical cancer screening in resource-limited setting. J Pak Med Assoc 2015; 65: 192–195.

13.

Dumont

Bessières

Benbassa

et al.

[Cervical cancer screening in rural Madagascar: Feasibility, coverage and incidence] [article in French]. J Gynecol Obstet Hum Reprod 2017; 46: 327–332.

14.

Nygård

Skare

Thoresen

SØ

. The cervical cancer screening programme in Norway, 1992–2000: changes in Pap smear coverage and incidence of cervical cancer. J Med Screen 2002; 9: 86–91.

15.

Ryu

Lee

Min

et al.

Management of atypical squamous cells of undetermined significance or low-grade squamous intraepithelial lesions of the uterine cervix with human papilloma virus infection among young women aged less than 25 years. Diagn Cytopathol 2016; 44: 959–963.

16.

Pitche

. [Child sexual abuse and sexually transmitted infections in sub-Saharan Africa] [article in French]. Med Trop 2005; 65: 570–574.

17.

Foumane

Chiabi

Kamdem

et al.

Sexual activity of adolescent school girls in an urban secondary school in Cameroon. J Reprod Infertil 2013; 14: 85–89.

18.

Opoku

Browne

Spangenberg

et al.

Perception and risk factors for cervical cancer among women in northern Ghana. Ghana Med J 2016; 50: 84–89.

19.

Barut

Kale

Kuyumcuoğlu

et al.

Analysis of sensitivity, specificity, and positive and negative predictive values of smear and colposcopy in diagnosis of premalignant and malignant cervical lesions. Med Sci Monit 2015; 21: 3860–3867.

20.

Nishimura

Miyatake

Nakashima

et al.

Clinical significance of atypical squamous cells of undetermined significance among patients undergoing cervical conization. Asian Pac J Cancer Prev 2015; 16: 8145–8147.

21.

Denny

Kuhn

De Souza

et al.

Screen-and-treat approaches for cervical cancer prevention in low-resource settings: a randomized controlled trial. JAMA 2005; 294: 2173–2181.

22.

World Health Organization. WHO guidelines for screening and treatment of precancerous lesions for cervical cancer prevention, Geneva: WHO, 2013.