Preoperative endometrial tissue sampling is essential before proceeding to

endometrial tissue sampling is essential before proceeding to hysterectomy.
Curettage is the traditional routinely used tool to diagnose and sometimes treat
women with different clinical pictures in need for hysterectomy for decades. Thus,
it was described as “the most expensive test in the whole medical field” by
Mengert and Slate. 9

In this
study our hypothesis was to determine the accuracy of the curettage specimen
pathological examination and its similarity to the more definitive hysterectomy
specimen. We did not test the curettage procedure against any other diagnostic
modality such as hysteroscopy, 3 Pippelle endometrial sampling 10
or endometrial brush cytology 11 as theses are not available in
every medical centre worldwide especially in the developing countries where the
curettage seems to be the cheap alternative to an expensive machines or
disposable catheters.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!

order now

The efficiency
of curettage was tested and evaluated since 1950s. Multiple studies
revealed the inaccuracy of curettage procedure and they considered D&C as a
poor diagnostic tool for endometrial cancer, as presented with high frequency
of false-negative findings. 6, 8 On the contrast, others reported
acceptable sensitivity, specificity and accuracy for the use of preoperative
D&C biopsies. 12-15

Although D&C
is considered effective for detecting endometrial lesions, it seems poor tool
for detection of its focal lesions. 12 Endometrial polyps,
submucous myomas and focal endometrial malignancy can be missed providing false-negative
results. 7 Inadequate uterine cavity evaluation and improper sampling
of targeted endometrial tissues are the most concerning points with obtaining
endometrial samples blindly. Thus, focal premalignant or malignant endometrial
pathologies can be frequently missed.

In our
study the frequently missed endometrial lesions were endometrial polyps for the
whole study group in addition to atrophic endometrium for the postmenopausal
ones. These findings were previously reported by other researchers. 7,
14 Other authors reported that D&C missed endometrial hyperplasia,
adenocarcinoma as well as cervical pathologies. 2, 17 Stovall
et al. reported that throughout 5.7% of their study group of 407 women, endometrial
hyperplasia or carcinoma were missed. 2 Furthermore, D&C missed
two women with cervical intraepithelial neoplasia and one woman with endometrial
carcinoma in 411 women as reported by Möller and Berget. 17

pitfall that can occur with endometrial sampling is the inability to obtain
adequate tissue sample sizable enough for optimum pathological examination or
what is called “insufficient sample”. In a recent meta-analysis, authors
included 12 reports over 1029 postmenopausal women and reported a weighted
endometrial sampling failure rate of 11%, while “insufficient samples” were
detected in 31% of cases. An endometrial cancer or premalignant lesions were
found in 7% of these women with insufficient or failed samples. 16

“insufficient samples” in our study group, no premalignant or malignant
endometrial pathology was encountered for the premenopausal group during the
hysterectomy specimens’ evaluation; while five premalignant (3.4%) and two
malignant (1.4%) pathologies were detected in postmenopausal women. This
obstacle seems to be more common and misleading among postmenopausal women due
to atrophic inactive endometrium. Thus, it is mandatory to investigate these
women using more advanced detailed techniques. 14, 16

calculation of D sensitivity, specificity and accuracy from most of
published reports cannot be considered conclusive owing to inadequate
information as well as the wide heterogeneity of choosing different sample
sizes, techniques, procedures and outcomes. Some used the missed diagnosed samples
without sub-grouping them as benign or malignant or sub-classifying their study
groups according to the age or symptoms. Thus, the usefulness and validity of
performing any routine D before hysterectomy remains questionable.

comparing D with hysterectomy are not entirely accurate, as invasive
techniques such as D may alter the uterine cavity. If endometrial polyps,
for example, are totally removed during curettage, the rate of false-positive
results may be overestimated. 18

We reported
that D, apart from low sensitivity for detection of premalignant lesions,
has high sensitivity, specificity, PPV, NPV and accuracy in detecting different
endometrial lesions for the whole study group. Also, the weighted kappa
coefficient proved a good agreement between D and hysterectomy diagnoses
regarding among the whole study population and post-menopausal women and
moderate agreement in the premenopausal women group. The ability of D to
diagnose the malignant lesions effectively was demonstrated.

accordance with our results, Sayg?l? reported positively correlated
preoperative D endometrial pathology findings when compared with the
postoperative hysterectomy pathology reports. 12 Also, Yarandi
et al. concluded that D had a high accuracy (92.1%) for endometrial
hyperplasia and carcinoma. 13 The weighted endometrial sampling
sensitivity using D as a reference for the diagnosis of endometrial
cancer and atypical hyperplasia were 100% and 92% respectively. 16

contrary, Ceci et al. evaluated the D accuracy based on the
histological findings or symptoms persistence in a cohort of 397 women. 6
In 248 cases (62.5%) of their study group, D was unable to determine
intrauterine lesions, having a 46% sensitivity of, 100% specificity, 100%
positive predictive value of, and 7.1% negative predictive value. Furthermore, Bettocchi
et al. reported 5.6 and 11.9% as false-positive and false-negative rates
respectively. 8

D is
not a satisfactory therapeutic procedure as by comparing D and
hysterectomy histological findings, the entire endometrial lesions could still
be identified in the removed uterus. However, Bettocchi et al.
considered D as a rare therapeutic option as they were still able to
recognize all the endometrial disorders after hysterectomy. Also, they reported
that neither endometrial hyperplasia (simple, complex or atypical) nor
endometrial carcinomas were entirely curetted. 8

To our
knowledge, this is the first study which calculated the sensitivity,
specificity, PPV, NPV and accuracy of each type of endometrial pathologies that
could be frequently encountered during management of women undergoing
hysterectomy for any indication.


conclusion, D remains an efficient diagnostic tool for preoperative
diagnosis of women undergoing hysterectomy for different indications in the
absence of other advanced diagnostic techniques. Preoperative D biopsy
offers acceptable accuracy regarding the diagnosis of most of endometrial
disorders, especially the premalignant and malignant lesions in postmenopausal
women, when compared to hysterectomy diagnoses.