More focus on atypical glandular cells in cervical screening: Risk of significant abnormalities and low histological follow-up rate

INTRODUCTION

With the wide application of cervical cancer screening, the incidence of squamous cancer and its precursors has been considerably reduced worldwide, whereas there is a rise in adenocarcinoma.^[1-3] It leads us to focus on recognition of the risk factors of adenocarcinoma. Atypical glandular cells (AGC) on Papanicolaou (Pap) smears are defined as glandular cells exhibit changes beyond those encountered in benign reactive processes, but lack the features of adenocarcinoma in situ or invasive adenocarcinoma on Pap smears.^[4]

Do AGC prove to be significant abnormalities? The result becomes significant for clinical management. It is reported that 18–83% of patients with AGC have significant abnormalities on follow-up histopathology.^[5-8] Notwithstanding, sometimes not enough attention is paid to women of AGC.

The objective of our study is to investigate the clinical significance of AGC. Our hospital is a large-scale specialized hospital in China. Only few patients, with gynecological diseases at their first visit or consultation to our hospital, have been lost to other hospitals in our follow-up. Therefore, the data are representative to some extent.

MATERIAL AND METHODS

RESULTS

Cytology

A total of 164,080 Pap smears were evaluated during the study period. Glandular cell abnormalities were encountered in 597 cases (0.4%) and AGC in 525 cases (0.3%). The number of cases with AGC-NOS was 314 and 211 with AGC-FN. The age of 525 women with AGC ranged from 22 to 78 years (median age: 46.7 years) at presentation. Incidence and prevalence of AGC was most common at the age of 41–50 followed by 51–60 and 31–40. About 84.4% of the 314 women with AGCNOS were aged between 31 and 60 years (AGC-FN, 78.2% of 211 women), whereas only 15.6% were aged below 31 years and above 60 (AGC-FN, 21.8%). About 46.0% (236 cases) of the women with AGC were postmenopausal. In Table 1, the age distribution is represented.

Close

Figure 1:

Study flowchart.

Export to PPT

Table 1: Age distribution of 525 women with AGC.

AGC	Age						Total
	21–30 (n=42) (%)	31–40 (n=118) (%)	41–50 (n=168) (%)	51–60 (n=154) (%)	61–70 (n=40) (%)	71–80 (n=13) (%)
NOS	33 (10.5)	71 (22.6)	102 (32.5)	92 (29.3)	13 (4.1)	3 (1)	314
FN	9 (4.3)	37 (17.5)	66 (31.3)	62 (29.4)	27 (12.8)	10 (4.7)	211

Pearson’s Chi-square test for women aged 31–60 years with AGC-NOS versus women aged 21–30 years or above 60 years with AGC-NOS: P=0.000. AGC: Atypical glandular cells, not otherwise specified (AGC-NOS), favor neoplastic (AGC-FN)

Cytohistological relation

Table 2 shows the histological follow-up after AGC. Only 310 women (310/525, 59.1%) with AGC had histological follow-up, 168 women (168/314, 53.5%) originally with AGC-NOS, and 142 (142/211, 67.3%) with AGC-FN [Figure 2]. Among 168 cases of histological follow-up, 102 cases were diagnosed of benign, 20 cases of significant squamous abnormalities, and 46 cases of significant glandular abnormalities after AGC-NOS. Among 142 cases, the data of the upper histological follow-up were 46, 20, and 76, respectively, after AGC-FN. Incidence and prevalence of significant abnormalities were highest when AGC was found at ages 51–60 followed by 41–50 [Table 3]. About 77.8% (126/162) of the women with significant abnormalities were postmenopausal. Benign histology mainly comprised inflammation, polyp, metaplasia, low-grade squamous intraepithelial (LSIL), and hyperplasia without atypia and so on. Significant abnormalities comprised significant squamous abnormalities and significant glandular abnormalities. Significant squamous abnormalities included high-grade squamous intraepithelial (HSIL) and squamous cell carcinoma. Significant glandular abnormalities included adenocarcinoma (originated from cervix, endometrial, or extra uterus), cervical high-grade cervical glandular intraepithelial neoplasia, and atypical endometrial hyperplasia.

Table 2: Histological follow-up in women with AGC.

Cytology	Histological follow-up			Total
	Benign (n=148) (%)	Significant squamous abnormalities (n=40) (%)	Significant glandular abnormalities (n=122) (%)
AGC-NOS	102 (60.7)	20 (11.9)	46 (27.4)	168
AGC-FN	46 (32.4)	20 (14.1)	76 (53.5)	142

Pearson’s Chi-square test for histological follow-up of AGC-NOS versus that of AGC-FN: P=0.000. AGC: Atypical glandular cells, not otherwise specified (AGC-NOS), favor neoplastic (AGC-FN)

Table 3: Age distribution of 162 women with significant abnormalities.

Histology	Age (n=162)
	21–30	31–40	41–50	51–60	61–70	71–80
Significant abnormalities	8	26	50	59	16	3

Patients aged 41–60 years with AGC are at higher risk for significant abnormalities. AGC: Atypical glandular cells

Close

Figure 2:

Histological follow-up of AGC-NOS and AGC-FN. AGC: Atypical glandular cells, not otherwise specified (AGC-NOS), favor neoplastic (AGC-FN).

Export to PPT

AGC-FN (96/142, 67.61%) was more likely to be associated with a clinically significant abnormalities (P < 0.001) compared to AGC-NOS (66/168, 39.29%). The diagnosis of AGC on Pap smears was associated with significant glandular abnormalities compared with significant squamous abnormalities on histopathology.

DISCUSSION

The incidence of AGC is reported to range from 0.1 to 2.1% in the literature.^[12-15] It was 0.32% in our study. From the data in this study, the peak occurrence age of both cytological AGC and significant pathological abnormality was from 41 to 60. The rate of significant pathological abnormalities was higher if the women with AGC were aged ranged 41–60, especially in postmenopausal state. Age was a risk factor.

AGC on Pap smears can be associated with benign, premalignant, and malignant conditions on followed-up histopathology. Some studies reported squamous epithelial abnormalities on histopathology were most common.^[4,5,15,16] In our study, non-significant abnormalities were separated from significant abnormalities on histopathology to avoid further active clinical procedures. AGC proved to be significant glandular abnormalities in this study (occupying up to 39.4%), compared with significant squamous abnormalities (12.9%). The diagnosis of AGC in our study was vigilant, which needed a consensus by three experienced cytologists. It indicated patients with AGC carried a significant risk for having a diagnosis of significant abnormalities, especially significant glandular abnormalities on histopathology.

AGC-FN was more closely associated with significant abnormalities than AGC-NOS, found to have statistically significant difference (P < 0.05). There is a considerable variation of AGC in association with significant abnormalities. To subdivide AGC into AGC-NOS and AGC-FN is necessary. In the literature, the detection rates of malignant or premalignant abnormalities ranged from 15% to 43% for AGC-NOS, and the corresponding range was 41–100% for AGC-FN.^[7,17-20] LSIL and hyperplasia without atypia were excluded from significant abnormalities in our study, with which patients usually received conservative treatment, resulted in decrease of positive rate on histopathology in our study. It was worth mentioning that three cases of AGC-FN on cytology: (1) Case of the cervical biopsies was diagnosed of LSIL and (2) were negative fractional curettage for the 1^st time. By reviewing the Pap smears, all the three patients underwent fractional curettage for the 2^nd time 4 months later. The histological results were endometrial carcinoma or HSIL, respectively. Clinicians and patients should attach more attention to AGC-FN on cytology. If cytologists highly suspect severe abnormality and histopathology diagnosis is negative otherwise, clinicians had better communicate with cytologists and take further examination. Sixty-six cases of AGN-NOS (66/168, 39.3%) were diagnosed of significant abnormalities on histopathology. AGC-NOS should not be ignored as benign lesions either.

In our study, 525 cases were diagnosed of AGC, yet only 310 cases had subsequent histopathology (AGC: 310/525, 59.1%; AGC-NOS: 168/314, 53.5%; and AGN-FN: 67.3%). A nationwide audit found that one-third of women with AGC lacked histological follow-up.^[21] This lack of follow-up might be associated with very high risks for cancer. The major hurdles are as follows: Clinicians have not enough knowledge of AGC. Cytological diagnosis of AGC is not accurate. Poor compliance of some patients also leads to low follow-up of AGC. It needs increased awareness of the necessity for histological follow-up after AGC. We would recommend that all smears reported as ACG should be followed-up with biopsy and curettage for confirmation of the diagnosis.

Forty cases of significant squamous abnormalities (including HSIL and squamous cell cancer) on histopathology were originally diagnosed of AGC on Pap smears. In our study, 148 cases of benign lesions followed up, severe chronic endocervicitis, polyps and submucosal leiomyomas, and microglandular hyperplasia were most common, 67, 46, and 18, respectively. These benign lesions led to reactive glandular hyperplasia/hypertrophy, which mimicked glandular abnormalities on cytology. These findings led us to focus on the cytologic characteristics of AGC in an effort to identify features that may be associated with underdiagnosis or overdiagnosis of this lesion, but some cases continue to be problematic when evaluated based on cytologic features.

CONCLUSION

Patients aged 41–60 years with AGC are at higher risk for significant abnormalities, especially for glandular abnormalities. To subdivide AGC into AGC-NOS and AGC-FN is necessary, since AGC-FN is more closely associated with significant abnormalities than AGC-NOS. Clinicians should carry out aggressive procedure in patients with AGC, especially with AGC-FN, to enhance the histological follow-up rate. When the corresponding histopathology after AGC-FN is negative, clinicians should thoroughly evaluate it. If necessary, a further clinical procedure should be carried out. The increase incidence of adenocarcinoma and precursors calls for a focused effort to be awareness of AGC.