Re: Benign Endometrial cells on pap smear in Postmenopausal women.
From: Harrison Sheld (hsheld@anv.net)
Thu Oct 23 18:42:15 2003
I would think that "going backward" might be more informative if one is
trying to connect benign endometrial cells/foamy histiocytes as
indicators of neoplasia. Why not review the Pap histories on patients
with endometrial adenocarcinoma and determine the occurrence, in their
preclinical state, of benign endometrial cells and/or foamy histiocytes?
Terrence.Jones@kp.org wrote:
>
> Yes Harrison, I understood Your concerns WRT ascertainment bias,
> selection bias, methodology and statistical validation, when first
> mentioned. What puzzles Me, is why would neoplastic endometrium be
> assiociated with a greater exfoliation of normal endometrial cells?
> (Tho, as You've pointed out, we don't know the rate of endometrial Ca,
> in this study pop, with normal pap and no EM cells.) Our regional
> cytolopathologist described the sampling achieved by cytobrush
> (frequently can reach lower segment) - which seemed a reasonable
> explanation for EM cells on pap. But what is the biologic plausibility
> for the assoc with neoplasia? I think You're correct in implicating
> historic confounders. Increased BMI, peripheral conversion of A to E1,
> insulin resistance (and inc IGFBP1 effects on endometrium), would all
> be red flags for an endometrium at risk for both benign proliferation,
> as well as neoplastic. Clinically, it would be of important to kn! ow
> the effects of these variables on predictive value. And, as is the
> case with asymptomatic Pt on Tamoxifen: what is the relationship to
> final staging when the currently asymptomatic patient later becomes
> symptomatic. (I'm not referring to incidence, Dan, but the impact of
> delay in diagnosis.) Using the cervix as an analogy is not valid, as
> the symptom we are observing (irregular bleeding) is frequently
> present in the endometrium lesion, while still pre-invasive. One of
> our East Bay residents did a study on our regional pop database, as
> his research project. I'll try to find the numbers. I believe there
> was an intent to publish, but - who can keep up, nowadays?! And thanks
> for YOUR interest! tj
>
> Harrison Sheld
> <hsheld@anv.net> To: Multiple recipients
> Sent by: of list OB-GYN-L
> ob-gyn-l@obgyn.net <ob-gyn-l@dns.obgyn.net>
> cc:
> 10/23/2003 02:16 PM Subject: Re: Benign
> Please respond to ob-gyn-l Endometrial cells on pap smear in
> Postmenopausal women.
>
> There is the statement: "6 of the 10 patients with EH/CA did not have
> vaginal bleeding and were referred for evaluation based solely on
> their
> cytological abnormality showing benign endometrial cells (four
> patients)
> or foamy histiocytes (two patients.)" They did exclude 10 patients who
> had atypical glandular cells.
>
> It is hard for me to accept that that the sole reason for referral was
> the cytology when there is no statement that a complete medical
> history
> and physical examination was performed on all reported patients; we
> don't know what confounding factors those patients had.
>
> Neither do we know what in the 34/159 patients' medical histories or
> physical examinations that may have been responsible for no followup
> and
> their exclusion from the study.
>
> There are no details about how the specimens were obtained or by whom.
> Neither is there indication on who obtained the information about
> bleeding and HRT. Could have been done by secretarial help in the
> office
> checking or not checking a box on the Pap requisition. Thanks for your
> interest.
>
> Terrence.Jones@kp.org wrote:
> >
> > Thanks Harrison! I hadn't time to hunt down the article, so delayed
> > comment. Wonder if there is mention of the cytologic appearance of
> the
> > endometrial cells in the 4 Pts with adenoca? Were they 'benign
> > appearing' in every case? Also, any discussion regarding cytobrush,
> vs
> > spatula alone, WRT specificity? tj
> >
> > Harrison Sheld
> > <hsheld@anv.net> To: Multiple recipients
> > Sent by: of list OB-GYN-L
> > ob-gyn-l@obgyn.net <ob-gyn-l@dns.obgyn.net>
> > cc:
> > 10/22/2003 04:44 PM Subject: Re: Benign
> > Please respond to ob-gyn-l Endometrial cells on pap smear in
> > Postmenopausal women.
> >
> > I have reviewed the article in the JLGTD. I believe there are
> several
> > problems with it.
> >
> > First, the patients were drawn from the population of a tertiary
> care
> > medical center and may not be representative of the general
> > population.
> >
> > Second, aside from hormone therapy, there are no historical data as
> to
> > reproductive history, previous HRT, current medications including
> > those
> > considered alternative, concurrent medical conditions, previous
> > abnormal
> > Paps or previous GYN surgery; all possible confounders in the
> > analysis.
> >
> > Third, the number patients in that hospital's catchment area who did
> > not
> > have endometrial cells or histiocytes but who had "asymptomatic"
> > endometrial carcinoma is not known. Knowing that the sensitivity of
> > the
> > Pap smear with regard to detecting endometrial pathology is not
> nearly
> > 100%, it may be that the "negative" population also had a 4% or
> > greater
> > incidence of endometrial cancer. In other words one cannot assume
> that
> > all the patients who did not have endometrial cells or histiocytes
> on
> > their Pap did not have endometrial cancer.
> >
> > Fourth, 12 patients on followup had insufficient tissue on sampling
> > and
> > were not included in the analysis. Since a common finding in
> > endometrial
> > sampling of postmenopausal patients is insufficient tissue commonly
> > from
> > an atrophic endometrium, these patients could have been included.
> > Their
> > inclusion would not have substantially changed the findings.
> >
> > Finally, the Bethesda system as recently revised does not include
> > endometrial cells as an epithelial abnormality in women over 40. I
> > would
> > be hard pressed absent other risks for endometrial cancer to justify
> > cell sampling every postmenopausal patient who had one Pap that
> showed
> > histiocytes and/or normal endometrial cells on the basis of this
> study
> > (nor did the authors make that suggestion.)
> >
> > "Braun, R. Daniel" wrote:
> > >
> > > They were assymptomatic and there was no relationship to their
> being
> > on
> > > estrogen or not according to the study. How many negative
> samplings
> > do
> > > you get for every 4 malignancies when you sample 40 yr olds with
> > > menomenorrhagia?? I'll bet it is a lot more than 96, still it is
> the
> > > standard of care.
> > >
> > > Dan
> > >
> > > -----Original Message-----
> > > From: ob-gyn-l@obgyn.net [mailto:ob-gyn-l@obgyn.net] On Behalf Of
> > > Harrison Sheld
> > > Sent: Tuesday, October 21, 2003 12:18 PM
> > > To: Multiple recipients of list OB-GYN-L
> > > Subject: Re: Benign Endometrial cells on pap smear in
> Postmenopausal
> > > women.
> > >
> > > I have not read the article in the ASCCP Journal but there are
> other
> > > articles that have a different statistics. I would advise the
> > patient of
> > > the statistics (some studies report a only 1% malignancy level in
> > > asymptomatic patients), and would let her make the decision. That
> > said,
> > > patients who are on estrogen, symptomatic, or with risk factors
> for
> > > endometrial CA should be sampled. Doing endometrial sampling on
> > > menopausal patients is often times difficult and uncomfortable.
> > > According to the article cited, and I don't know if the patients
> > were at
> > > risk for CA, on estrogen, or symptomatic, 96 out of 100 samplings
> > would
> > > be unnecessary. If the patient chose not to be sampled I would
> > inform
> > > her to report in a timely fashion symptoms if they should occur
> and
> > > return for a Pap in 4 months. Just my opinion and I could be
> wrong.
|
|
