The Small Blue Cell Dilemma Associated With Tamoxifen Therapy.
MATERIALS AND METHODS
Papanicolaou tests from patients who were on tamoxifen therapy, according to their specimen requisition forms, were selected from the files of Upstate Medical University, State University of New York (Syracuse, NY) and Community General Hospital of Greater Syracuse (Syracuse, NY) for a recent 10-year period. Also included were Pap tests positive for small blue cells found in routine screening but without requisition history of tamoxifen therapy; the physician offices of these patients were contacted to clarify the medical history. Additional relevant clinical data collected for all cases included the patient's age when tamoxifen therapy was initiated, history of hysterectomy or other gynecologic procedures, status of menopause, and the time interval between the initiation of tamoxifen administration and the identification of small blue cells. When a Pap test positive for small blue cells was identified, prior Pap test specimens were re-screened for the presence of small blue cells. The maturation index of each specimen (cervical or vaginal smears) was determined based on the method of Wied and Bibbo. When available, follow-up surgical specimens from these patients were also reviewed.
The Student t test was performed to determine the significance of difference on selected data obtained from patients positive and negative for small blue cells.
The small blue cells associated with tamoxifen therapy have a distinct and characteristic morphology, as illustrated in the Figure. These cells usually occur as small, tightly cohesive clusters with scant to absent cytoplasm. The nuclei of these cells are similar to those of intermediate squamous cells and have fine hyperchromatic chromatin and indistinct or minute nucleoli.
A total of 154 Pap tests from 60 patients on tamoxifen therapy were reviewed, among which 18 tests from 12 patients were without requisition history of tamoxifen therapy and the history was identified by subsequent clinician contact. These patients were separated into either positive or negative groups based on the respective presence or absence of small blue cells in their Pap tests. Twenty-four (40%) of these patients had positive Pap tests and 36 (60%) patients had negative Pap tests. The positive group was an average of 9 years older than the negative group at the time tamoxifen therapy was initiated (P [is less than] .01). Relevant clinical data for these patients are summarized in Table 1.
Table 1. Summary of Clinical History of Patients Receiving Tamoxifen Therapy Clinical History Positive Group (n = 24) Average age at menopause, y 46 (range 40-54) Postmenopause patients, % 85 Posthysterectomy patients, % 42 Age of patient at initiation of 66 (range 47-87)(*) tamoxifen therapy, y Latency between therapy initiation and identification of small blue cells, y 2-4 Clinical History Negative Group (n = 36) Average age at menopause, y 46 (range 41-57) Postmenopause patients, % 84 Posthysterectomy patients, % 30 Age of patient at initiation of 57 (range 37-78)(*) tamoxifen therapy, y Latency between therapy initiation and identification of small blue cells, y Not applicable (*) Four patients from the positive group and 3 patients from the negative group were omitted from the calculation owing to insufficient data.
Small blue cells were seen in the Pap tests of women who had undergone hysterectomy. In Pap tests from women with a uterus, no definitive endometrial cells were identified in any of the specimens with small blue cells. Additional findings in the study population included Trichomonas vaginalis infection (1 case) and atypical squamous cells with subsequent biopsy negative for intraepithelial lesion (1 case).
The maturation index of all the specimens, including cervical or vaginal smears, was determined, acknowledging that cervical smears are not the ideal specimens for hormonal assessment. The mean of the maturation index of the positive group was significantly higher than that of the negative group (P [is less than] .01), as demonstrated in Table 2.
Table 2. Mean Maturation Indices of Papanicolaou Tests in Positive and Negative Groups(*) Maturation Index, % Small Blue Cells (Parabasal: Intermediate: Superficial) Positive group 13 ([+ or -] 17):63 ([+ or -] 23):24 ([+ or -] 20) (range 0-53: 25-100: 0-60) Negative group 24 (][+ or -] 38):67 ([+ or -] 39):9 ([+ or -] 8) (range 0-100:0-100:0-25) (*) Student t test on superficial cells between the 2 groups, P < .01.
Twenty-two surgical specimens were available for 19 patients in the positive group, and 3 specimens were available from 3 patients in the negative group. No malignant diagnoses were made (Table 3). The majority (73%) of patients in the positive group had a diagnosis of atrophic or inactive endometrium (36.5%) or benign cervical/vaginal biopsies (36.5%). An additional 23% of these patients had endometrial biopsies showing proliferative-phase endometrium.
Table 3. Follow-up Surgical Biopsy of Patients Receiving Tamoxifen Therapy Positive Negative Group Group Diagnoses (n = 24) (n = 36) Atrophic/inactive endlometrium 8 0 Proliferative endometrium 5 2 Endometrial breakdown 0 1 Endometriosis 1 0 Benign vaginal/cervical biopsy 8 0
Four patients were dropped from this study. Among them were 2 patients whose Pap tests were positive for small blue cells, but whose histories were negative for tamoxifen therapy or other hormonal therapy (one was a 77-year-old woman with breast carcinoma 23 years earlier, and the other was a 58-year-old woman with no history of malignancy). The other 2 patients had Pap tests that were negative for small blue cells, yet these women were taking tamoxifen for diseases other than breast carcinoma (one for endometrial carcinoma and the other for the Breast Carcinoma Prevention Trial).
We observed a high incidence of distinctive small blue cells in Pap tests of patients treated with tamoxifen for breast carcinoma. These cells were initially thought to be of endometrial origin; however, the occurrence of these cells in Pap tests from hysterectomized patients indicates a different origin. Reviews of the limited number of surgical specimens failed to define the origin of these cells, but based on the morphology and the sampling sites of these specimens, we speculate that these cells represent hyperplastic reserve cells of cervix and/or vaginal epithelium. Although immunocytochemistry was not performed on our cases owing to the need to keep the slides for archival records, our hypothesis is further supported by a recent study. As reported in an abstract, small cells were found in 15 (21.4%) of 70 cases involving patients treated with tamoxifen. Similarly, a nonendometrial origin was also suggested by the authors, based on the cytologic morphology and immunocytochemical profile, including positivity for cytokeratin and estrogen receptor, but negativity for vimentin. Complex endometrial hyperplasia was found in only 1 of 14 endometrial biopsies. Thus, the benign nature of all the follow-up surgical biopsies from patients of both studies strongly suggests that these small blue cells are nonneoplastic.
Although our data showed a 40% positive rate for small blue cells in the Pap tests of patients treated with tamoxifen, the actual positive rate is probably much lower since negative tests from patients on tamoxifen could only be recognized by the history on the requisition. As this history is often incomplete, our measured population of all patients on tamoxifen is grossly underestimated. Nonetheless, the incidence of these cells was much higher in the tamoxifen-treated population than in the remaining population. Only rare cases in our current study revealed infection (Trichomonas) or reactively atypical squamous cells coexistent with small blue cells; thus, infection and inflammation were not likely responsible for the incidence of the small blue cells.
In our files, only 2 patients with a confirmed negative history of tamoxifen therapy had Pap tests that were positive for these cells. The etiology of the small blue cells in these 2 cases is not clear Hormonal replacement therapy does not appear to increase the incidence of small blue cells in Pap tests, as we would expect more positive cases unrelated to tamoxifen to have been identified in routine screening. Thus, we hypothesize that the high incidence of these small blue cells is most likely related to the estrogen-agonist effect of tamoxifen, especially in the postmenopausal population; furthermore, such an effect is different from that caused by estrogen, most likely acting through a different mechanism of action, as has been suggested by others.[8,9]
The various side effects of tamoxifen therapy appear to be age dependent. Thus, tamoxifen exerts an antiestrogenic effect in vaginal and uterine tissue of premenopausal patients,[5,15-17] but has an estrogenic effect in the same tissue of postmenopausal patients.[4,5,17,18] It is theorized that the age-dependent effect of tamoxifen is related to the hormonal milieu into which the drug is induced. Thus, tamoxifen may act as an agonist when endogenous estrogens are low, and conversely may act as an antagonist when circulating estrogens are high.
The incidence of small blue cells in our study has also demonstrated this age-related phenomenon. At the initiation of tamoxifen therapy, the average age of the positive group was 9 years older than that of the negative group. This significant age difference between the 2 groups of patients is also reflected in the length of time that the patients had been postmenopausal prior to tamoxifen therapy. Theoretically, the estrogenic agonist effect of tamoxifen becomes increasingly more conspicuous as the endogenous estrogen levels decrease gradually after menopause. Thus, compared to patients in the positive group, most patients in the negative group had been postmenopausal for a shorter period of time, should have a higher endogenous estrogen level, and therefore experienced a weaker agonist effect from tamoxifen.
Papanicolaou tests from the positive group also displayed an increased ratio of superficial cells. Our study showed that the ratio of superficial cells to parabasal and intermediate cells was significantly higher in the positive group than that in the negative group (P [is less than] .01), which is similar to the findings of earlier studies showing that tamoxifen caused an increase in the maturation index in vaginal smears from postmenopausal women.[4,5,17] Although the cervical smear is not the ideal specimen for hormonal evaluation, it can indirectly indicate a patient's hormonal status, provided that the results are not affected by an inflammatory condition.
The small blue cells identified in our study are not endometrial in origin, an important diagnostic distinction in that the presence of endometrial cells in the Pap tests of patients receiving tamoxifen indicates a higher risk of endometrial adenocarcinoma. Previous studies have demonstrated an increased incidence of endometrial diseases associated with tamoxifen therapy, including endometrial carcinoma, endometrial hyperplasia, and endometrial polyps.[2,8-14] In our study, neither dysplasia nor inflammatory atypia was found to be associated with tamoxifen therapy, although reactively atypical glandular cells and squamous cells have been reported.[19-21]
It is important to recognize that these distinctive but benign small blue cells are associated with tamoxifen therapy, since the differential diagnosis includes endometrial cells, endometrial carcinoma, and metastatic breast carcinoma. Out-of-phase endometrial cells in premenopausal patients may indicate abnormalities such as endometrial hyperplasia or endometrial carcinoma. Unlike the tamoxifen-related small blue cells, benign endometrial cells usually have irregular nuclear contours, degenerative appearance, coarser chromatin, and may present as exodus. Malignant endometrial cells are larger, and their nuclei are hyperchromatic with coarser chromatin and conspicuous nucleoli. Uncommon gynecologic carcinomas that are composed of predominantly small cells include carcinoid tumor, small cell carcinoma, adenoid cystic carcinoma, and adenoid basal carcinoma. They can be distinguished from tamoxifen-related small blue cells by the morphologic features summarized ill Table 4. Lobular breast carcinoma also is composed of relatively small cells and can be a diagnostic challenge, especially when the patient has a history of breast lobular carcinoma; however, lobular carcinoma cells are larger and usually have a plasmacytoid appearance, irregular nuclear membrane, coarser chromatin, and conspicuous nucleoli. Our data from follow-up surgical biopsies strongly suggested that these small blue cells are nonneoplastic.
Table 4. Comparison of Cytomorphology of Small Cells From Various Sources[22,23] Tamoxifen- Related Cytologic Small Blue Endometrial Lobular Feature Cells Cells Carcinoma Microscopy ar- Small tight Small tight Small clusters, chitecture clusters cluster single, Indian file Necrosis No No No Mitoses No No No Cells Small, bland Small, bland Small, bland Cytoplasm Minimal Minimal Vacuole Nuclei Small, dark, Small, bland, Eccentric, small, smooth irregular dark, irregular Nucleoli Indistinct Small Distinct Adenoid Adenoid Cytologic Basal Cystic Small Cell Feature Carcinoma Carcinoma Carcinoma Microscopy ar- Nests, cords, Cords, trabecu- Crowded sheets chitecture haphazard lae, hyaline material, cys- tic spaces Necrosis No Yes Yes Mitoses No Yes Yes Cells Small, bland Small, bland Small, oval Cytoplasm Minimal Minimal Scant Nuclei Small, dark, Small, dark, Oval, dark, irregular irregular irregular Nucleoli Indistinct Indistinct No Cytologic Carcinoid Feature Tumor Microscopy ar- Ribbons, nests, chitecture trabeculae Necrosis No Mitoses No Cells Small, round Cytoplasm Minimal Nuclei Round, stippled, smooth Nucleoli Small
In summary, our study demonstrated the presence of distinctive small blue cells in patients undergoing tamoxifen therapy for breast carcinoma. These cells are benign and nonendometrial in origin. Their morphology of tight clusters of undifferentiated small cells with minimal cytoplasm, round to oval nuclei, bland chromatin, and indistinct nucleoli is reminiscent of the reserve cells of the cervix and the vagina. Therefore, we hypothesize that these tamoxifen-related small blue cells are hyperplastic reserve cells of the vagina or cervix induced by the estrogen-agonist effect of tamoxifen, but the definitive origin of these cells requires further study. However, we believe these cells to be completely benign, and therefore their presence in the Pap tests of women on tamoxifen should cause no alarm and does not indicate a need for further workup to rule out more serious pathology.
[1.] Sunderland MC, Osborne CK. Tamoxifen in premenopausal patients with metastatic breast cancer: a review. J Clin cncol. 1991;9:1283-1297.
[2.] Barakat RR. The effect of tamoxifen on the endometrium. Oncology. 1995; 9:129-134.
[3.] Boccardo F, Bruzzi P, Rubagotti A, et al. Estrogen-like action of tamoxifen on vaginal epithelium in breast cancer patients. Oncology. 1981;38:281-285.
[4.] Lahti E, Vuopala S, Kaupilla A, Blanco G, Ruokonen A, Laatikainen T. Maturation of vaginal and endometrial epithelium in postmenopausal breast cancer patients receiving long-term tamoxifen. Gynecol Oncol. 1994;55:410-414.
[5.] Abadi MA, Barakat RR, Saigo PE. Effects of tamoxifen on cervicaovaginal smears from patients with breast cancer. Acta Cytol. 2000;44:141-146.
[6.] Wied GL, Bibbo M. Hormonal cytology. In: Bibbo M, ed. Comprehensive Cytopathology. Philadelphia, Pa: WB Saunders; 1991:85-114.
[7.] Nahal A, Waszcok W, Mandavilli S, Ludwig M. Cytopathologic features of small cells (SCT) seen in Pap smears after tamoxifen treatment. Acta Cytol. 1999; 44:867.
[8.] Magriples U, Naftolin F, Schwartz PE, Carcangiu MI. High-grade endometrial carcinoma in tamoxifen-treated patients. J Clin Oncol. 1993;11:485-490.
[9.] Silver EG, Tornos CS, Follen-Mitchell M. Malignant neoplasms of the uterine corpus in patients treated for breast carcinoma: the effects of tamoxifen. Int J Gynecol Pathol. 1994;13:248-258.
[10.] Fornander T, Rutqvist LE, Cedermark B, et al. Adjuvant tamoxifen in early breast cancer: occurrence of new primary cancers. Lancet. 1989;1:117-120.
[11.] Hardell L. Tamoxifen as risk factor for carcinoma of corpus uteri [letter]. Lancet. 1988;2:563.
[12.] Lahti E, Blanco G, Kaupilla A, Apaja-Sarkkinen M, Taskinen PJ, Laatikainen T. Endometrial changes in postmenopausal patients receiving tamoxifen. Obstet Gynecol. 1993;81:660-664.
[13.] Nuovo MA, Nuovo GJ, McCaffrey RM, Levene RU, Barron B, Winkler B. Endometrial polyps in menopausal patients receiving tamoxifen. Int J Gynecol Pathol. 1989;8:125-131.
[14.] Korley D, Rowe J, Curtis MT, Hogan WM, Noumoff JS, Livolsi VA. Postmenopausal bleeding from unusual endometrial polyps in women on chronic tamoxifen therapy. Obstet Gynecol. 1992;79:111-116.
[15.] Tajima C. Luteotropic effects of tamoxifen in infertile women. Fertil Steril. 1984;42:223-227.
[16.] Neumannova M, Kauppila A Kivinen S, et al. Short term effects of tamoxifen, medroxyprogesteron acetate, and their combination on receptor kinetics and 17[Beta]-hydroxysteroid dehydrogenase in human endometrium. Obstet Gynecol. 1985;66:695-700.
[17.] Ferrazzi E, Cartei G, Mattarazzo R, et al. Oestrogen-like effect of tamoxifen on vaginal epithelium [letter]. Br Med J. 1997;1:1351-1352.
[18.] Bonte J, Ide P, Billiet G, et al. Tamoxifen as a possible chemotherapeutic agent in endometrial adenocarcinoma. Gynecol Oncol. 1981;11:140-161.
[19.] Gill BL, Simpson JF, Somlo G, McGonigle KF, Wilczynski SP. Effects of tamoxifen on the cytology of the uterine cervix in breast cancer patients. Diagn Cytopathol. 1998;19:417-422.
[20.] Cenci M, Mancini R, Nofroni I, Vecchione A. Endocervical atypical glandular cells of undetermined significance, II: morphometric and cytologic analysis of nuclear features useful in characterizing differently correlated subgroups. Acta Cytol. 2000;44:327-331.
[21.] Liu K, Marshall J, Shaw HS, Dodge RK, Layfield LJ. Effects of chemotherapy and tamoxifen on cervical and vaginal smears in bone marrow transplant recipients. Acta Cytol. 1999;43:1027-1033.
[22.] DeMay RM. The Art & Science of Cytopathology. Chicago, III: American Society of Clinical Pathologists Press; 1996:71,882.
[23.] Powers CN, Stastny JF, Frable WJ. Adenoid basal carcinoma of the cervix: a potential pitfall in cervicovaginal cytology. Diagn Cytopathol. 1996;14:172-177.
Accepted for publication March 22, 2001.
From the Department of Pathology, Upstate University, State University of New York, Syracuse (Dr Yang and Ms Demoski); the Department of Pathology, Laboratory Alliance of Central New York at St Joseph's Hospital, Syracuse (Dr Trapkin and Ms Bellerdine); and the Department of Pathology, Medical College of Virginia, Virginia Commonwealth University, Richmond (Dr Powers).
Presented in part at the 45th Annual Scientific Meeting of the American Society of Cytopathology, Boston, Mass, November 4-8, 1997.
Reprints: Yi Jun Yang, MD, PhD, Department of Pathology, Upstate Medical University, State University of New York, 750 E Adams St, Syracuse, NY 13210 (e-mail: Yangy@upstate.edu).
|Printer friendly Cite/link Email Feedback|
|Author:||Yang, Yi Jun; Trapkin, Linda K.; Demoski, Roberta K.; Bellerdine, Jeannette; Powers, Celeste N.|
|Publication:||Archives of Pathology & Laboratory Medicine|
|Article Type:||Statistical Data Included|
|Date:||Aug 1, 2001|
|Previous Article:||The Histopathology of 103 Consecutive Colonoscopy Biopsies From 82 Symptomatic Patients With Acquired Immunodeficiency Syndrome: Original and...|
|Next Article:||Major Pancreatic Resections for Chronic Pancreatitis: Histologic Findings and Their Association With Persistent Pain.|