Endometrial Secrets Unmasked through Cytology Analysis

Abstract

Background:

Endometrial aspiration (EA) is an economic, painless technique on an outpatient basis. Present study aimed at cytological evaluation of EA for (a) cellular yield and morphology and (b) utility of cell block (CB) and cytocentrifuge (CC) techniques.

Materials and Methods:

EA samples were divided into two aliquots. Colorless samples were processed (1000 rpm × 6 min) for conventional smear (CS) and CC, both stained by Papanicolaou. Hemorrhagic samples were processed for CS and CB (paraffin embedding, hematoxylin–eosin stain).

Results:

Endometrial aspirates from January 2021 to January 2022 were included. Indications comprised abnormal uterine bleeding (AUB; 87), prolapse (eight), and infertility (two). Among 77 hemorrhagic aspirates, the yield was 85.7% by CS and 90.9% by CB. Among 20 colorless aspirates, the yield was 55% by CS and 65% by CC. The yield was 85.7%, 84.4%, and 83.3% with endometrial thickness 1–5, 6–10, and 11–15 mm, respectively. The yield was 83.9%, 50%, and 0% in AUB, prolapse, and infertility, respectively. CS morphology showed the categories of benign (93.5%) and atypia (6.5%). All cases with benign morphology correlated with CB and CC. CB offered architectural evaluation, while CC had a shorter turnaround time.

Conclusion:

Focusing on menorrhagia cases in secretory phase, nondilution of EA samples, and simultaneous endometrial biopsy can enhance cytology evaluation. In an era where “less should convey more,” EA shows potential as a screening technique vis-à-vis invasive “dilatation–curettage” technique.

INTRODUCTION

Endometrial aspirate (EA) is an economic, painless, and ambulatory technique for endometrial sampling on an outpatient department (OPD) basis. Cytological analysis of EA is useful for screening.[1-3] Cytoconcentration of EA sample is possible by cytocentrifuge and cell block. The present study involves a comparison among conventional smear (CS), cytocentrifuge (CC), and cell block (CB) techniques.

Aim

This study aimed towards 1) evaluation of EA in terms of cellular yield and cytomorphology 2) comparative utility of CB and CC.

MATERIALS AND METHODS

This prospective, analytical study was conducted in a tertiary health center. EA samples were collected at Gynecology OPD during the 13-month study period (January 2021–January 2022). After due clinical consent was taken, infant-feeding canula was inserted via cervical os. EA was collected by negative suction and flushed with 2–5 ml saline into a sterile container.

On receipt at the cytology department, each sample was equally divided into two aliquots. Samples were classified as hemorrhagic or colorless. Colorless samples were processed for CS at 1000 rpm × 6 min, and in Shandon CC at 1500 rpm × 6 min. Both smears were stained by Papanicolaou. Hemorrhagic samples were processed for CS and CB. The latter involved post-centrifugation fixation in 10% neutral formalin, paraffin embedding, and hematoxylin–eosin staining.

All slides were evaluated for yield adequacy (minimum cluster of 15 endometrial epithelial cells) and cytomorphology.[1-4]

Data was entered in Microsoft Excel. Cellular yield in terms of sample type, clinical indication, and endometrial thickness (ET) was analyzed using unpaired t-test using Statistical Package for the Social Sciences (SPSS) software. P-value <0.05 was considered statistically significant.

RESULTS

A total of 104 EA samples were received during the 13-month study. Seven cases received as only air-dried smears were excluded. Among the 97 final cases, 59.8% were of age 41–50 years (range 22–56 years).

Clinical presentation: Indications for EA are shown in Table 1. Ultrasonography was done in 81 cases, of which 55 cases (67.9%) showed an organic lesion like leiomyoma 32, adenomyosis 10, ovarian cyst/mass eight, endometrial growth three, and endometrial hyperplasia two. Cases (26, 32.1%) without any organic cause qualified as dysfunctional uterine bleeding (DUB).[1-3]

Table 1.

Clinical indications and cellular yield for endometrial aspiration (n=97)

Clinical indication	Number of cases	Aspirates with adequate cellular yield No. (%)
Abnormal uterine bleeding	87 (89.7%)	73 (83.9%)
-Menorrhagia	- 66
- Polymenorrhea	- 3
- Dysmenorrhea	- 3
- Postmenopausal bleeding	- 7
- Irregular menses	- 5
- Postcoital bleeding	- 3
Prolapse	8 (8.2%)	4 (50%)
Infertility	2 (2.1%)	0 (0%)
Total	97 (100)	77 (79.4%)

Epithelial cellular yield was as per the following:

• a) Sample type [Table 2]: Hemorrhagic samples had higher yield (85.7%) than colorless samples (55%) on CS (P < 0.05).

• b) ET was documented in 51 (53.5%) cases. There were seven, 32, and 12 cases having ET 1–5, 6–10, and 11–15 mm, respectively. The yield was 85.7%, 84.4%, and 83.3%, respectively (P > 0.05).

• c) Clinical presentation: Cases with abnormal uterine bleed (AUB) demonstrated highest yield (P < 0.05) [Table 1].

Table 2.

Comparison between hemorrhagic and colorless endometrial aspirates (n=97)

	Hemorrhagic aspirate	Colorless aspirate
Number of cases (%)	77 (79.4%)	20 (20.6%)
Mean age in years	53	42
Cellular yield on CS	66 (85.7%)	11 (55%)
On CC	Not applicable	13 (65%)
On CB	70 (90.9%)	Not applicable
Turnaround time (days)
CS	1	1
CC	Not applicable	1
CB	3 to 5	Not applicable
Concordance between CS and CC	Not applicable	100%
Concordance between CS and CB	93.5%	Not applicable

CB=cell block, CC=cytocentrifuge, CS=conventional smear

Morphology on CS

The architectural patterns included clusters (38), sheets (22), three-dimensional exodus (11), and tubules (six cases) [Figure 1a–d]. Epithelial cells were categorized as benign (72, 93.5%), atypia (5, 6.5%), and malignant (0%).[5] Benign features included uniform, round nuclei with smooth nuclear membrane [Figure 1a and b]. Cellular “microbiopsy,” nucleomegaly, and irregular nuclear membrane were labeled as atypia [Figure 2a and b]. Ectocervical squames (86), endocervical cells (23) [Figure 2c], lymphocytic infiltrate (three), and actinomycotic colonies (one) [Figure 2d] were additional observations.

Figure 1.

Architectural patterns of endometrial epithelial cells on conventional smear showing (a) tight clusters, (b) monolayered sheets, (c) three-dimensional balls, and (d) cohesive tubules (photomicrograph, 100×, Papanicolaou stain)

Figure 2.

(a and b) Cellular, branching “microbiopsy” with nucleomegaly and nuclear overlap suggesting atypical endometrial epithelia (photomicrograph, 100×, Papanicolaou stain). (c) Ectocervical squamous cells (red arrow) and endocervical glandular cells (blue arrow) (photomicrograph, 100×, Papanicolaou stain). (d) Actinomycotic filaments (photomicrograph, 400×, Papanicolaou stain)

CB versus CC

CB had higher yield and longer turnaround time (TAT) than CC [Table 2]. Architectural details such as proliferative phase (PP; 63 including five cases labeled as atypia on CS), secretory phase (SP; three), menstrual phase (MP; three), and pill endometrium (one) could be discerned on CB. Based on CB findings, we reviewed CS for hormonal evaluation.[1,4] On CS, epithelial cells with scant cytoplasm, round nuclei, fine chromatin, and stromal “streaming fish” pattern corresponded to PP [Figures 1a and 3a]. Epithelial cytoplasmic vacuolation, vesicular nuclei, and stromal decidualization corresponded to SP [Figure 3b and c].

Figure 3.

(a) Streaming pattern of benign fusiform stromal cells in proliferative phase (photomicrograph, 100×, Papanicolaou stain). (b) Secretory phase epithelia with vacuolated cytoplasm, vesicular nuclei, and micronucleoli (photomicrograph, 400×, Papanicolaou stain). (c) Stromal polygonal decidual cell (red arrow) with abundant cytoplasm, vesicular nucleus, and nucleolus. Blue arrow shows columnar epithelial cell (photomicrograph, 400×, Papanicolaou stain)

DISCUSSION

AUB means abnormal volume, abnormal frequency, or lasting beyond 6 months.[1,3]

Age group and clinical features

Majority of AUB cases occur in the fourth to fifth decade.[1-3,6] AUB can have organic causes (leiomyoma, polyp, hyperplasia, malignancy), while DUB is unassociated with an organic cause.[3]

Sampling device

EA with Mi-Mark helix, Isaacs sampler, Karman’s canula, Vabra aspirator, intracath canula, etc., shows variable cellular yield from 59% to 97%.[1,2,6,7]

Sample type

In our study, hemorrhagic samples showed higher yield. Saline flushing may have led to diluted, low yield, colorless samples. Handa et al.[3] used a similar saline technique with 89% yield. Studies using undiluted EA material to make direct smears onto glass slides documented 90%–100% yield.[1,7] Atrophic endometrium and technical errors are known factors for inadequate cellularity.[8]

ET and Last menstrual period (LMP)

Anovulatory cycles with unopposed estrogenic effect in DUB lead to endometrial hyperplasia with ET >1 mm.[9] Zhang et al.[2] used >4 and >20 mm as the criteria for thickened endometrium in postmenopausal and menopausal age, respectively. There is higher rate of sample adequacy with increasing ET.[10] We found uniform yield despite variable ET.

Cytoarchitectural criteria for hormonal evaluation demonstrated 97.6% accuracy rate.[1,3,7] Late SP is best suited to diagnose anovulatory cycles.[3] Our study had menstrual cycle documented in only 48 cases, of which 23 were PP and 15 were SP by date. Only six cases showed SP on review of CSs, while no case showed hyperplasia or malignancy.

Comparison with CB and CC

Concordance between EA and biopsy ranges from 90% to 96% in literature.[1,3,7] Endometrial biopsy was not conducted in our study. Instead, cytology–histology comparison was made between CS and CB for hemorrhagic samples. Atypia in 6.5% CS cases did not reflect in CB, but all benign CS cases showed concordance with CB. Thus, CS demonstrated 92.4% sensitivity and 100% positive predictive value for benign morphology, with 100% negative predictive value for atypia. Cellular atypia in atrophic endometrium and endometritis is a known diagnostic pitfall.[1]

Zhang et al.[2] and Kyroudi et al.[11] inserted EA material into fixative for liquid-based cytology (LBC). Residual sample was converted into CB.[2,11] They reported higher diagnostic accuracy using combined LBC with CB than LBC alone.[2,11] However, due to use of residual sample with cell loss in LBC preparation, CB had more inadequacy (22.2%) than LBC (7.1%).[2] We circumvented this drawback by dividing each sample at receipt into two equal aliquots, to be used for CS and CC/CB, respectively. Dharan[12] also documented the diagnostic superiority of CB.

Highlights and limitations

The present study found excellent cellular yield, short TAT, and high concordance. Constructive feedback (focus on AUB/DUB cases, sample during late SP, non-dilution of EA sample) has been conveyed to the gynecologist. Concurrent endometrial biopsy will further sharpen the diagnostic skills of histopathologist.

CONCLUSION

Ambulatory health services are increasingly in demand. EA procedure is painless and cost-effective. A fruitful symbiosis between gynecologist–pathologist team will enhance its screening value. In an era where “lesser tissue should convey more information,” EA cytology is an exciting prospect.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.