AUTORES
Prof. MsC José Queiroz Filho a
Prof. PhD Janaina Cristiana de Oliveira Crispim Freitas b
Prof. Dr Ricardo Cobucci e
Prof. Dr. José Eleutério Júnior c Prof. PhD Paulo César Giraldo d
Prof PhD Ana Katherine Gonçalves a –e
a) Pos Graduate Program in Health Sciences - Federal University of Rio Grande do Norte - Rio Grande do Norte, Brazil
b) Department of Clinical Analysis, School of Pharmacy - Federal University of Rio Grande do Norte - Rio Grande do Norte, Brazil
c) Department of Gynecology and Obstetrics, Faculty of Medicine - Federal University of Ceará, Brazil
d) Department of Gynecology and Obstetrics, Faculty of Medicine from State University of Campinas - Sao Paulo, Brazil
e) Department of Gynecology and Obstetrics, Faculty of Medicine - Federal University of Rio Grande do Norte, Brazil
Correspondence José Queiroz Filho
Av Airton Senna, 1000 Nova Parnamirim. CEP:59151 600 Parnamirm (RN) Brazil
Phone (84) 98135 0909, fax (84) 3203 5039 Email: [email protected]
Abstract
Objective: To evaluate 100% rapid review (100%RR) as a useful tool to detect FN (false
37
and negative results were undergone to 100%RR, concordant results were considered final diagnosis, while the discordant were debated in a consensus meeting to conclude. The positive results were examined by two cytologists, and the concordant results were considered final diagnosis, while the discordant were revised in a consensus. The data were entered into SAS statistical software, and the agreement of the 100%-RR results with the final diagnosis was tested with the weighted kappa statistic. Results: After ending review, there was a significant increase of unsatisfactory results from 348 to 1927 and positive results from 174 to 349. On the other hand, there was a substantial decrease in negative results from 8155 to 6401. Assessing the relative risk (RR) of false negative in smears that were not undergone to quality control (100%RR), was found: 2.93 (ASC-US), 2.72 (LSIL), 2.25 (HSIL-ASC-H). Evaluating by age group, the higher risk for LSIL (4.90) and ASCUS (3.85) was observed in patients under 25 years. Whereas patients between 25 and 64 and over 64 presented a higher risk for HSIL and ASC-H, 2.46 and 2.75 respectively. Conclusion: 100%RR is an effective screening tool for FN results in countries where molecular tests for DNA-HPV and prophylactic vaccines are not available in cervical cancer screening programs.
Keywords
Quality control, Diagnosis, Accuracy, Cytology
Introduction
Currently, the research shows the association between human papillomavirus (HPV)
and cervical what has well established the role of HPV in cervical carcinogenesis when
associated with a high oncogenic virus related to other factors [1]. The use of molecular tests
for HPV and vaccines as prophylactic forms of infection in prevention programs provides
success of programs in countries with high economic power [2 ].
The incidence has decreased considerably in western countries due to organized
screening programs based on tests that provide good accuracy. The effectiveness of these
38
4, 5]. The lower incidence and mortality of cervical cancer in developed countries can be
explained by well-structured cytological screening programs and better medical infrastructure,
which are not frequently found in low-and-middle-income countries [6 ].
Although the cytopathological examination is the most used method for the cancer of
the uterine cervix since the early 1980s, there have been some criticisms related to the high
proportion of false-negative results, ranging from 2% to 62%. The leading causes are material
collection, scrutiny of the smear or in the interpretation of diagnosis [ 7, 8, 9 ].
Various methods of internal quality monitoring were developed with the purpose of
reducing errors of scrutiny and analysis of the cytology diagnosis, among them the rapid review
of 100% (100%-RR) of negative smears [8, 10].
The 100%-RR consists of quickly scrutinize all smears considered negative. After this
rapid review (RR) al suspect smears are subjected to a detailed review to define a final
diagnosis [11]. This method was known to be used as a tool in internal quality control (IQC) in
the United Kingdom, countries in Europe, Australia, and North America 100%-RR is also used
in quality assurance in cytopathology. [11, 12].
Other works comparing review methods show that RR is an efficient method that has
good relation cost benefit as an internal quality assurance when compared different means of
review, detecting significantly more false negative (FN) [13, 14].
Molecular tests and prophylactic vaccines for HPV are procedures used in developed
countries associated with cytology which ensures the success of cervical cancer screening
programs, which for high financial value does not happen in developing countries. Therefore,
the 100% RR is an efficient alternative that can be implanted in these nations favoring the
39
Material and Methods
This study was conducted at the Laboratory of Cytology of the Department of Pharmacy
at the Federal University of Rio Grande do Norte, Brazil, was approved the Committee on
Ethics and Research in Human (CAAE17535013.90000.5296). The cervical smears obtained
by conventional collect were donated by different cytology laboratories participating in the
program of the Brazilian Ministry of Health. Cervical samples were collected from patients by
primary care professionals in the National Health System.
In this study, the smears considered as 174 altered, 349 unsatisfactory and 8155
negative released at the first screening were reviewed as follows. The altered smears were
examined by two cytologists, and the concordant results were considered final diagnosis, while
the discordant were for consensus.
In this study smears considered negative and unsatisfactory underwent 100%-RR. Two
cytologists were responsible for implementing the quick scrutiny. Two other cytologists were
accountable for the detailed review of smears considered suspect. Three different cytologists
were also responsible for the review of all discordant cases to complete the final diagnosis
considered the gold standard. These cytologists participated in the consensus meetings to
conclude the diagnosis of discordant results.
Reviewers underwent prior training before performing the 100%-RR technique. The
instruction was given in the form of a pilot study for 60 days in which cytologists had the
opportunity to practice the pre-quick scrutiny and define the operational characteristics of the
study to ensure that all methodological processes established would be fulfilled.
To avoid any error due to lack of concentration, or other factors that could affect the final
result of this study, we followed the methodology proposed by Tavares and colleagues [14].
40
limited to the task of scrutinizing 40 smears daily. It was determined that each count should be
done within one minute.
Study procedures were performed in the following sequence: All smears considered
altered, unsatisfactory and negative in the first ballot were subjected to 100%-RR and were
classified as suspicious, negative or unsatisfactory and were recorded in a spreadsheet:
Reviewers did not have access to clinical data of patients and smears showed no marked field.
The results classified as suspect were subjected to a detailed review and the diagnoses were
recorded in a spreadsheet. Smears whose results were discordant in the in-depth review were
recorded in a spreadsheet.
When two reviewers issued concordant results in this detailed review, this was
considered the final diagnosis, but the results deemed discordant were analyzed in a
consensus meeting to set the final diagnosis. Reviewers who participated in this stage to
complete the definitive diagnosis did not have access to the results previously issued, except
in the cases submitted to consensus meeting, as shown in the flowchart (Figure 1).
Results
In this study, 8677 cervical cytology smears were analyzed, which in the first scrutiny
presented the following cytological diagnoses; 348 smears considered unsatisfactory, 8155
negative smears and 174 positive smears with 73(0.84) cases of ASC-US, 05 (0.06) of ASC-
H, 54 (0.62) of LSIL, 17 (0.19) of HSIL and 25(0.50) of AGC. The smears considered
unsatisfactory and negative in the total of 8503 were submitted to a 100%RR and then a
detailed review, while those with positive diagnoses at the first scrutiny were subjected to
detailed review. The final result after monitoring was 1927 unsatisfactory smears, 6401
negative smears and 349 positive smears with the following diagnoses; 171 (1.96) cases of
41
squamous carcinoma and adenocarcinoma 01 (0.01) case respectively, As shown in table 01.
When assessing the relative risk (RR) of the results considered positive in the scrutiny first in
relation to the diagnosis final, the following results were found for ASC-US 2.93 (95%CI 2.23 - 3.84), for LSIL 2.72 (95%CI 1.97 – 3.76), for HSIL-ASC-H 2,25 (95%CI 1.33 – 3.79), as shown in table 2. In the present study, the following results were obtained for the age group of less than years 25 to calculate the RR for ASC-US 3.85 (95% CI 2,19 – 6,76), LSIL 4.90 (95% CI 2.52 -9.49), HSIL-ASC-H 1.07 (95% CI 0,29 – 4,01). In the age group of years 25 to 64the following results; for ASC-US 2,52 (95%CI 1,84 -3,46), LSIL 2,19 ( 95%CI 1,50 – 3.19), HSIL- ASC-H 2,46 (95%CI 1,38 – 4,41). For the age group above years 64, it was possible to calculate the RR for HSIL-ASC-H in this is 2.75 ( 95%CI 1,16 – 6,51 ), as shown in table 3.
Discussion
The high percentage of unsatisfactory results is attributed to obscuring factors such as;
inflammation, cell overlap, cytolysis, blood, and desiccation. Queiroz Filho et al. [15] justify the
high the index of unsatisfactory to pre-analytical errors such as those cited. Studies using
conventional and liquid-based cytology demonstrate the influence of preanalytical factors on
the suitability of samples. The use of lubricating gel, blood, and factors associated with an
inadequate collect such as cell overlap, inflammation, and low cellularity favors unsatisfactory
sample [16, 17]. Although there may be limitations and unsatisfactory, a markedly cytological
method is still used in clinical practice to attempt to reduce cancer incidence and mortality. For
it to be effective, a high percentage de unsatisfactory, among the factors involved in the context
of the Brazilian screening program, such as economic, social, political and cultural
developments should be noted, which suggests the need to identify the failures and
42
In the first scrutiny, 174 positive smears were detected, but in the final diagnosis 349
positive smears, which demonstrates the importance of an internal quality control, monitoring
tool to identify both FN and PF results. Cytology as a screening method for cervical lesions
may show moderate sensitivity and depends or the interpreter`s interpretation, with HSIL being
more easily traced which favors FN [18]. Another factor that influences the false negative
results is the volume of smears examined daily by a professional that leads to the error of
smears considered negative at the first scrutiny not submitted to a new evaluation before
release [19]. Studies performed with 6000 Papanicolaou smears whose abnormal results were
analyzed histology showed a positive correlation 0f 96/152 (63.2%) and a discrepancy of
55/152 (36.80%) suggesting the need for 100%RR in cytology services[20].
It was observed that women whose cytological smears only submitted to routine scrutiny
has the possibility of being carriers of lesions and that these lesions were not detected in the
first scrutiny.
In this study we also analyzed the relative risk of positive results being considered NB in the
first survey, which demonstrates significant risk mainly in not tracking borderline lesions such
as ASC-Us and LSIL in women whose smears are not submitted to quality internal control and
the studies demonstrate the effectiveness of 100% RR in detecting false negative results.
Additionally, we analyzed the relative risk of positive results being considered NF in the first
screening, which demonstrates significant risk mainly in not tracking limited lesions such as
ASC-US and LSIL in women whose smears are not submitted to IQC and studies show the
efficacy of 100% RR in trackback NF results. Studies using RR-100% as IQC using different
counting times demonstrate the efficacy of this methodology for the screening of cervical
cancer. A study of 5235 smears considered negative at the first screening using RR-100% in
43
diagnosis 45 cases were confirmed independently of times used [21]. Another study with 2500
cytological tests considered negative after RR-100% detected 310 suspected cases com 140
confirmed after histopathological analysis [22] can be used as internal quality control. However
the studies show that RR-100% is one of the most efficient methods, all the results considered
negative in the first scrutiny are scrutinized in a minimum time, and those considered suspects
undergo a detailed analysis for final diagnosis. This methodology allows exposing the tellers
to constant training that raise the quality of the programs of prevention cervix. What decreases
the risks of women with cellular atypia not being screened, increasing the sensitivity of cytology
as a screening method for cervical cancer in countries of economic transition that have the
cytology method used in the screening of this pathology, where the molecular tests for DNA-
HPV and as prophylactic vaccines that in developing countries like those of Latin America,
Africa and some Asian countries, whose programs of prevention of cervical cancer having a
cytology like method of screening are not yet consolidated.
REFERENCES
1.Vacarella S, Laversanne M, Ferlay J, Bray F, Cervical cancer Africa, Latin America and the Caribbean and Asia: Regional Inequalities and Changing trends. Int. J. Cancer. 2017; 141(10) 1997 – 2001
2. Pilozo BQCH, Magues LGM, Garcia ME, Sanchez M, Valdiverezo RJ, Molina D. Molecular epidemiology and phylogenetic analysis of Human papillomavirus infection in women with cervical lesions and cancer from the Coostal region of Ecuador. Rev Argent Microbiolog , 2017; S0325-7541 (17): 30137-2
3.Elmajjaoul S, Ismaili N, El Kaceni H, Kebdani T, Sifat H, Benjaafar N, Epidemiology and outcome of cervical câncer in National Institute of Morocco. BMC Womens Health. 2016; 16(62): 2-8
44
4.Samson KK, Haynatzki G, Soliman AS, Valerianova Z, Temporal changes in the cervical cancer burden in Bulgaria: implications for European countries going through transition. Cancer Epidemiol. 2016; 44: 154-160
5.Derchain S, Teixeira JC, Zeferino LC, Organized population-based cervical câncer screening program: it would be a good time for Brazil now. Rev Bras Ginecol Obstet. 2016;38: 161-163
6.Mathew A, George PS, Trends in incidence and mortality rates of squamous cell carcinoma and adenocarcinoma of cervix - worldwide. Asian Pac J Cancer Prev. 2009;10(4): 645-50
7.Amaral, R, Marinque E, Guimarães J, Souza P, Mignoli J, Xavier A, Oliveira A, Influência da adequabilidade da amostra sobre a detecção das lesões precursoras do câncer cervical. Rev Bras Ginecol Obstet. 2008; 30(11) :556-560
8.Gullo C, Dami A, Barbosa A, Marques A, Palmejani M, Lima L, et al., Results of a control quality strategy in cervical cytology. Einstein. 2012: 10(1): 86-91
9.Branca M, Longatto Filho, A. Recommendations on Quality Control and Quality Assurance in Cervical Cytology, Acta Cytolog. 2015; 59(5): 361-369
10. Franco R, Amaral RG, Montemor EBL, Montis DM, Morais SS, Zeferino LC, Fatores associados a resultados falso-negativos de exames citopatológicos, Rev Bras Ginecol Obstet, 2006:28(8): 479-485
11. Ministério da Saúde, Instituto Nacional do Câncer José de Alencar Gomes da Silva; Coordenação de Prevenção e Vigilância; Divisão de Detecção Precoce e apoio à organização de Rede. Manual de Gestão de Qualidade para Laboratórios de Citopatologia. RJ, INCA. 2016
12. Manrique EJ, Souza NL, Tavares SB, Albuquerque ZB, Zeferino LC, Amaral RG, Analysis of the performance of 100% rapid review using na average time of 1 and 2 minutes according to the quality of cervical cytology pecimens. Cytopathology.2011;22(3):195-201
13. Tavares SB, Souza NL, Manrique EJ, Azara CZ, Silveira EA, Amaral RG, Internal quality control for cervical cytopathology: comparison of potential false-negatives detected at rapid prescreening anda t 100% rapid review. Acta Cytolog 2014;58(5);439-445
45
14.Tavares SB, Souza NL, Manrique EJ, Albuquerque ZB, Zeferino LC, Amaral RG, Improvement in the routine screening of cervical smears: A study using prescreening and 100% rapid review as internal quality control methods. Cancer Cytopathology. 2011;119(6): 367-376
15. Queiroz Filho J, Freitas JC, Pessoa DC, Eleutério Junior, J, Giraldo PC, Gonçalves AK, Assessment of 100% Rapid Review as na effective Tool for Internal Quality Control in Cytopathological Services. Acta Cytolog, 2017; 61(3): 207-213
16.Mirzamani N, Chau K, Rafael O, Shergill U, Sajjan S,Sumskaya I, et al. Quality assessment and improvement of Unsatisfactory liquid-based cervicalvaginal papanicolaou smears. Diagn Cytopathol. 2017; 45 (10)873-877
17.Rosa M, Pragasam P, Saremian J, Aoalin A, Graf W, Mohammadi A, The unsatisfactory ThinPrep Pap Test: analysis recommendations for improvement. Diagn Cytopathol. 2013; 41(7): 588-594
18.Sorbye SW, Suhrke P, Reva BW, Berland J, Maurseth RJ, Al-Shibli K, Accuracy of cervical cytology: comparison of diagnoses of 100 Pap smears read by four pathologists at three hospitals in Norway. BMC Clin Pathol, 2017; 29: 17-18
19. Longatto-Filho A,Shmitt FC. Cytology education in the 21 st century: living the past or crossing the Rubicon? Acta Cytol, 2010;54: 654-656
20.Alwahaibi NY, Sulimi SK, Bai UR, Cytohistological correlation and discrepancy of conventional Papanicolaou smear test with corresponding histopathology: a retrospective study over a 5-year period. East Mediterr Health J. 2015; 21(8): 579-583
21. Marinque EJ, Souza NL, Tavares SB, Albuquerque ZB, Zeferino LC, Amaral RG, Analysis of the performance of 100% rapid review using na average time of 1 and 2 minutes according to the quality of cervical cytology specimens. Cytopathology, 2011; 22(3): 195-201
.22. Sood N, Singh V, Evaluation 0f 100% rapid rescreening of cervical smears. Indian J Pathol Microbiol. 2009, 52(4): 495-497
46
FIGURE 1: Flowchart of the methodology
1° SCRUTINY