PART TWO
Diagnostic Cytology
Fig. 19.75 Smear of peritoneal effusion depicting a tiny Curschmann
spiral that at one end appears to blend with a streak of mucin. The effusion
also contained metastatic gastric adenocarcinoma cells (Papanicolaou x HP).
Reproduced with permission from Naylor B: Curschmann's spirals in pleural
and peritoneal fluids.
Acta Cytol
1990;34:474-478.
Otherwise, endometrial cells, either from foci of endometrio-
sis or from tubal reflux, are most likely to be seen in specimens
obtained by culdocentesis or in peritoneal washings.9,117 But-
ler and Stanbridge described cells consistent with endometrial
stroma and epithelium in aspirates obtained from the cul-de-
sac (pouch of Douglas); the presence of these cells was believed
to be due to reflux menstrual fluid. The cells believed to be of
endometrial stromal origin were very similar to the histiocyte-
like cells seen in postmenstrual cervical smears.
The diagnosis of endometriosis from cells in spontaneously
occurring serous effusions, culdocentesis specimens, or perito-
neal washings is fraught with difficulty. Individual endometrial
stromal cells may be similar to lymphoid cells in shape and size,
or they may closely resemble macrophages or small mesothe-
lial cells. Furthermore, endometrial cells of epithelial type in
a serous effusion or peritoneal washing may, unless their epi-
thelial architecture is extremely well preserved, resemble small
fragments of mesothelium. In fact, in the three reported cases
of endometriotic cells in spontaneously occurring effusions
referred to previously, there is no record of a cytologic diagnosis
of endometriosis being reported before biopsy; it appears that
the cytologic presentation of endometriosis was recognized ret-
rospectively. Probably the most reliable method of diagnosing
endometriosis in a serous effusion is to demonstrate endome-
triotic fragments in a cell-block preparation, which would allow
identification of both epithelial and stroma components.
Key features of endometriosis in serous effusions
• Diagnosis fraught with difficulty: endometrial cells
simulated by other cells; and
• Best evidence of endometriosis: cell block demonstrat-
ing endometrial fragments.
Curschmann Spirals
Curschmann spirals are common in cytologic specimens from
the respiratory tract and have also been recorded in cervical
smears. In these situations, their presence could be explained
by their being formed from mucus normally present in the envi-
ronment. However, a paradoxical source of Curschmann spirals
Fig. 19.76 Smear of empyema effusion depicting a long Curschmann spiral.
The patient had no evidence of fistula or carcinoma (Papanicolaou x LP).
Reproduced with permission from Naylor B: Curschmann's spirals in pleural
and peritoneal fluids.
Acta Cytol
1990;34:474-478.
Fig. 19.77 Cell block of empyema effusion depicting a coiled Curschmann
spiral. This is the same specimen as in Fig. 19.75 (H&E x LP).
was reported by Wahl, who found them in peritoneal washings,
pleural effusions, and peritoneal dialysate.165
We have now observed six examples of Curschmann spirals
in smears and cell blocks of spontaneously occurring pleural
and peritoneal effusions of six patients. On the whole, these spi-
rals were much shorter than those usually seen in respiratory or
cervical secretions (Fig. 19.75), although some were similar in
size to those seen in sputum (Figs 19.76 and 19.77).166 Some
spirals were rather poorly formed and were regarded as incipient
spirals. Three of the six patients had serous effusions containing
mucus-secreting carcinoma cells; in fact, one patient had pseu-
domyxoma peritonei, characterized by mucinous ascitic fluid.
Of the other three patients, two had pleural empyema, and the
other possibly had spontaneous bacterial peritonitis associated
with hepatic cirrhosis.
Based on the observation that Curschmann spirals in the
female genital tract have been recorded only in patients who had
a uterus in situ, thereby providing a source of mucin, we believe
that the presence of mucin in serous effusions can account
for spiral formation. However, where there were no mucin-
producing epithelial cells to give rise to spiral formation, as in
548
previous page 542 ComprehensiveCytopathology 1104p 2008 read online next page 544 ComprehensiveCytopathology 1104p 2008 read online Home Toggle text on/off