Predictors for survival in fallopian tube carcinoma
Patrick F. Timmins, M.D.*1.; Anisa I. Kanbour, M.D.†; and Fredric V. Price, M.D.*
Division of Gynecologic Oncology, Department of Obstetrics and Gynecology and †Department of Pathology, Magee-Womens Hospital, 300 Halket St., Pittsburgh, PA 15213.
Correspondence to: Dr Timmins, New York Oncology Hematology, 317 South Manning Blvd., Suite 280, Albany, NY 12208
Background: Fallopian tube carcinoma is usually a surprise diagnosis in patients undergoing an exploratory laparotomy for another presumed malignancy. Although it is treated using ovarian cancer surgery and chemotherapy, it may have a different biology and prognosis than its more well known counterpart. This study was designed to investigate prognostic factors of overall survival in this uncommon and poorly understood malignancy.
Methods: The tumor registry and practitioner databases at Magee-Womens Hospital were searched for patients with fallopian tube carcinoma. After pathological confirmation, a chart review was undertaken of the 47 patients identified to collect data for survival predictors. Statistical analysis was performed using standard models with significance at p .05.
Results: Five separate predictors of overall survival were identified in the univariate analysis. Negative peritoneal cytology (p<.01), no residual disease at primary cytoreductive surgery (p<.01), pelvic disease (p=.02), abnormal vaginal bleeding as a presenting symptom (p=.02), and negative second look laparotomy (p=.03) were all predictive of prolonged overall survival.
Conclusions: The results from our large series provide significant clinical information about fallopian tube carcinoma. Negative peritoneal cytology was the strongest predictor of survival, and the only one to remain significant in multivariate analysis. Abnormal vaginal bleeding is a predictor of survival and is not affected by stage.
Fallopian tube carcinoma is the most uncommon carcinoma of the mullerian system, accounting for 3.6 cases per million women per year. This rarity has prevented prospective trials that could answer questions of optimal treatment and prognostic factors. Only large retrospective series have been able to find significant prognostic indicators for this disease. Since our institution has one of the largest gynecologic tumor registries in the United States, we undertook a study to attempt to identify new clinically relevant prognostic indicators and/or support others previously reported in the literature.
Materials and Methods:
Medical records of all patients with possible primary fallopian tube carcinoma seen at The Magee-Womens Hospital from April 1963 to September 1994 were reviewed. This study had full Institutional Review Board approval. Patients were identified through searches of the institutional tumor registry and practitioner data bases. Pathology slides were reviewed by one senior pathologist (AIK), and satisfied the diagnostic criteria for primary fallopian tube cancer proposed by Hu, Taymor, and Hertig and modified by Sedlis. All tumors were classified by the 1991 International Federation of Gynecologists and Obstetricians' (FIGO) staging for fallopian tube cancer. Staging was completed using surgical and pathological findings extracted from the patients' records. Kaplan-Meier life-table analysis was used to generate survival curves. Survival based on categorical variables was compared using the Peto-Peto Wilcoxon test. Multivariate analysis was undertaken on significant univariate predictors using the Cox proportional hazards regression model.
Forty-seven patients met diagnostic criteria. Of the 47 patients, 37 were diagnosed since 1980. The median age at diagnosis was 62 years (range = 40-81 years). Forty-three (91%) of the patients were white, three (6%) were black, and one (2%) was asian. Five percent of the patients were nulliparous and 26% were para one or less. The menstrual status of 46 patients was known: 15 (33%) were pre-menopausal and 31 (67%) were postmenopausal. Five (11%) had previous hysterectomies. Nine (20%) had a family history of breast, uterine, colon, and/or ovarian cancer (BUCO CA) in a first degree relative.
The most common presenting symptoms were vaginal bleeding (28%), abdominal pain (21%), and abdominal distention (15%). Thirty-six percent were asymptomatic. The duration of symptoms before diagnosis was determined for 33 patients: Symptoms had been present from 0 to 19 months with a median of 2 months and an interquartile range of 6 months. Only one patient was correctly diagnosed pre-operatively.
Of the 42 patients in whom the amount of residual disease could be quantified, 36 (77%) were considered to have no residual disease, and 6 (13%) had largest disease less than 2 cm. Lymph node sampling was performed in 17 patients; 5 were positive.
All 47 patients had surgical removal of their tumor as primary therapy. In addition 30 patients received chemotherapy. Many different regimens were employed. 24 patients received platinum-based chemotherapy, 5 patients received 5-FU, and 3 patients received thiotepa. Eight patients underwent radiation therapy either alone or in combination with chemotherapy, 6 were stage I/II.
Fifteen patients had no recurrence in the follow-up period, 16 patients recurred, and 16 patients had persistent disease. Fifteen patients underwent second look laparotomy. Seven had a negative second look and one of those recurred two years later.
Forty-six patients were able to be staged retrospectively. Of these 19 were stage 1, 8 were stage 2, 15 were stage 3, and 4 were stage 4. 34 tumors were graded for differentiation. 18 tumors were poorly differentiated, 11 were moderately differentiated, and 5 were well differentiated. 24 of the tumors were of serous histology while 23 were nonserous.
The median overall survival for these 47 patients was 36 months (m) with a median disease free interval of 56 months (16 patients excluded from this analysis because they had persistent disease). The overall 5 year survival (5YS) was 37%.
Multiple possible predictors of overall survival (OS) were subjected to Kaplan Meier life-table analysis using the Peto-Peto Wilcoxon test for significance at the p .05 level. Table 1 presents the significant predictors of survival. The median OS for stage I disease was 64 m with a 5YS of 55%, Stage II 38 months with a 5YS of 29% Stage III 18 months with a 5YS of 13%, and Stage IV 28 m with a 5YS of 0%. There was, however, no statistically significant difference in survival among the individual stages. When the stages were broken down into pelvic disease (FIGO I & II) vs. extrapelvic disease (FIGO III & IV), as in the report by Podratz, et al. and Rosen, et al.[5,6] there was a statistically significant survival advantage to pelvic disease (p=.02). Patients with pelvic disease had a 5YS of 52%, whereas those with extrapelvic disease had a 5YS of 7%.
Other significant predictors in the univariate analysis included peritoneal cytology. Of 28 patients with cytology results, the 13 patients with negative cytology, even if ascites was present, had a 10YS of 58%, and the 15 patients with positive cytology had a 10YS of 7% (p < .01). Fifteen patients had a second look laparotomy . Those with a negative second look had a 5YS of 69% vs those with a positive second look who had a 5YS of 17% (p= .03).
When OS was compared by the patient's presenting symptom or lack of symptoms, a significant survival advantage was shown for patients who presented with abnormal vaginal bleeding over those who were either asymptomatic or presented with pain or distention (p=.03). Patients who presented with abnormal vaginal bleeding had a 5YS of 67%, those with abdominal pain as a presenting symptom had a 5YS of 38%, those with complaints of abdominal distention had a 5YS of 0%, while patients who were asymptomatic at presentation had a 5YS of 9%.
Patients were categorized by residual disease after first laparotomy into those who had no residual disease versus those who had documented residual disease or the status of residual disease was unknown. This revealed a significant OS advantage in the univariate analysis (p < .01).
Patients treated with platinum chemotherapy had a longer OS in all stages, but it did not reach statistical significance. Patients with pelvic disease did have a significantly longer DFI when treated with platinum than those who were not (p=0.05 (Logrank)). Of the 10 patients with pelvic disease who were treated with platinum chemotherapy the median DFI was > 123 m, while the 4 treated with nonplatinum chemotherapy had a DFI of 30 months.
After determining which variables were statistically significant in the univariate analysis, those variables found to be significant, except SLL, were subjected to the Cox proportional hazards model to find multivariate predictors of survival. Only the presence of malignant cytology in the peritoneal cavity was significant in this model with p<.01.
The rarity of fallopian tube carcinoma has inhibited the definition of its natural history and the delineation of optimal treatment. The demographic characteristics of our study population are generally consistent with those of other reports [1,5,7-39], except that our incidence of nulliparity is quite low when compared to other series.[5,7-9,11-14,18,21-25,29,31-33]
In other series fallopian tube carcinoma has been reported to be diagnosed at early stages because of a propensity to cause symptoms.[5,23] In our series, 62% of the patients had disease confined to the pelvis, but 36% were asymptomatic. We did find, as have others, that stage is an important predictor of OS.[5,18,19,23,25,33,37] Additionally, all of our patients were initially treated with surgical debulking and we found that the complete removal of gross disease was a significant predictor of OS in the univariate analysis. This agrees with four other large series.[5,16,18,25] Malignant peritoneal cytology was the strongest predictor of OS and the only one to be significant in the multivariate analysis. The report by Podratz, et al was the only other large series to report this.
When clinicians are faced with a malignancy that is rare and about which little is understood it is important to understand as much about prognostic factors as possible. In our population we found that bleeding as a presenting symptom was associated with a significantly longer OS regardless of stage. The rationale that bleeding is associated with prolonged survival does not relate to early detection. Perhaps these tumors were more angiogenically active and were more sensitive to chemotherapy in the advanced stage disease. Tumors that have a propensity to bleed may be associated with a more brisk inflammatory reaction as noted by Rosen, et al.
Other authors have recommended the routine use of second look laparotomy (SLL) in fallopian tube carcinoma as a guide to further treatment.[16,40,41] In our series we found that a negative SLL was associated with a greatly increased median OS, and only one of the seven patients who had a negative SLL recurred. Treatment of rare malignancies is always problematic as there is usually no standard therapy based on randomized studies, and this holds true in fallopian tube carcinoma. Since fallopian tube carcinomas resemble ovarian carcinomas they have usually been treated in a similar fashion.[42,43] In our series we found some evidence to support this practice as our patients had a longer OS when treated with platinum than those who were not, but this trend was not significant.
Although our study was limited by sample size and the necessary retrospective nature, it confirms and adds further information to the relatively small body of literature available on fallopian tube carcinoma.
TABLE 1. Significant predictors for survival in univariate analysis
SECOND LOOK LAPAROTOMY
|SYMPTOMS AT DIAGNOSIS||DISTENTION||7||15||0.02|
Nonsignificant variables included Individual Stages, Grade, Age, Ascites,
Menopausal Status, Nodal Status, Vaginal cytology, Radiation therapy, Treatment,
and Synchronous/Metachronous primary.
1. Rosenblatt KA, Weiss NS, Schwartz SM. Incidence of malignant fallopian tube tumors. Gynecol Oncol 1989;35:236-9.
2. Hu CY, Taylor ML, Hertig AJ. Primary carcinoma of the fallopian tube. Am J Obstet Gynecol 1950;59:58-67.
3. Sedlis A. Primary carcinoma of the fallopian tube. Obstet and Gynecol Surv 1961;16:209-26.
4. Pettersson F. Staging rules for gestational trophoblastic tumors and fallopian tube cancer. Acta Obstet Gynecol Scan 1992;71:224-5.
5. Podratz KC, Podczaski ES, Gaffey TA, O'Brien PC, Schray MF, Malkasian GD. Primary carcinoma of the fallopian tube. Am J Obstet Gynecol 1986;154(6):1319-26.
6. Rosen A, Klein M, Lahousen M, Graf AH, Rainer A, Vavra N. Primary carcinoma of the fallopian tube--a retrospective analysis of 115 patients. Austrian Cooperative Study Group for Fallopian Tube Carcinoma. Br J Cancer 1993 Sep;68(3):605-9.
7. King A, Seraj I, Thrasher T, Slater J, Wagner RJ. Fallopian tube carcinoma: a clinicopathological study of 17 cases. Gynecol Oncol 1989;33:351-5.
8. Hirai Y, Kaku S, Teshima H, Shimizu Y, Chen J-T, Hamada T, et al. Clinical study of primary carcinoma of the fallopian tube: experience with 15 cases. Gynecol Oncol 1989;34:20-6.
9. Hée P, Pagel JD. Primary carcinoma of the fallopian tube. Eur J Obstet Gynecol Reprod Biol 1987;25:131-8.
10. Yeung HH, Bannatye P, Russell P. Adenocarcinoma of the fallopian tubes: a clinicopathological study of eight cases. Pathology 1983;15:279-86.
11. Chalmers JA, Marshall AT. Carcinoma of the fallopian tube. Br J Obstet Gynaecol 1976;83:580-3.
12. Harrison CR, Averette HE, Jarrell MA, Penalvar MA, Donato D, Sevin B-U. Carcinoma of the fallopian tube: clinical management. Gynecol Oncol 1989;32:357-9.
13. Semrad N, Watring W, Fu Y-S, Hallatt J, Ryoo M, Legasse L. Fallopian Tube Adenocarcinoma: Common extraperitoneal recurrence. Gynecol Oncol 1986;24:230-5.
14. Asmussen M, Kaern J, Kjoerstad K, Wright PB, Abeler V. Primary adenocarcinoma localized to the fallopian tubes: report on 33 cases. Gynecol Oncol 1988;30:183-6.
15. Kinzel GE. Primary carcinoma of the fallopian tube. Am J Obstet Gynecol 1976;125(6):816-20.
16. Barakat RR, Rubin SC, Saigo PE, Lewis JL, Jones WB, Curtin JP. Second-look laparotomy in carcinoma of the fallopian tube. Obstet Gynecol 1993;82(5):748-51.
17. Amendola BE, LaRouere J, Amendola MA, McClatchey KD, Han IH, Morley GW. Adenocarcinoma of the fallopian tube. Surg Gynecol Obstet 1983;157:223-7.
18. Peters WAI, Andersen WA, Hopkins MP, Kumar NB, Morley GW. Prognostic features of Carcinoma of the fallopian tube. Obstet Gynecol 1988;71(5):757-62.
19. Rosen AC, Reiner A, Klein M, Lahousen M, Graf AH, Vavra N, et al. Prognostic factors in primary fallopian tube carcinoma. Austrian Cooperative Study Group for Fallopian Tube Carcinoma. Gynecol Oncol 1994 Jun;53(3):307-13.
20. Hershey DW, Fennell RH, Major FJ. Primary carcinoma of the fallopian tube. Obstet Gynecol 1981;57(3):367-70.
21. Nordin AJ. Primary carcinoma of the fallopian tube: a 20 year literature review. Obstet Gynecol Surv 1994;49(5):349-61.
22. Tamimi HK, Figge DC. Adenocarcinoma of the uterine tube: potential for lymph node metastases. Am J Obstet Gynecol 1981;132-7.
23. Roberts JA, Lifshitz S. Primary adenocarcinoma of the fallopian tube. Gynecol Oncol 1982;13:301-8.
24. Raju KS, Barker GH, Wiltshaw E. Primary carcinoma of the fallopian tube: Report of 22 cases. Br J Obstet Gynaecol 1981;88:1124-9.
25. Eddy GL, Copeland LJ, Gershenson DM, Atkinson EN, Wharton JT, Rutledge FN. Fallopian tube carcinoma. Obstet Gynecol 1984;64(4):546-52.
26. Benedet JL, White GW, Fairey RN, Boyes DA. Adenocarcinoma of the fallopian tube: experience with 41 patients. Obstet Gynecol 1977;50(6):654-7.
27. Boutselis JG, Thompson JN. Clinical aspects of primary carcinoma of the fallopian tube: a clinical study of 14 cases. Am J Obstet Gynecol 1971;111(1):98-101.
28. Gurney H, Murphy D, Crowther D. The management of primary fallopian tube carcinoma. Br J Obstet Gynaecol 1990;97:822-6.
29. Momtazee S, Kempson RL. Primary adenocarcinoma of the fallopian tube. Obstet Gynecol 1968;32(5):649-56.
30. Kneale BLG, Attwood HD. Primary carcinoma of the fallopian tube: report of 13 cases. Am J Obstet Gynecol 1966;94(6):840-8.
31. Hanton EM, Malkasian GD, Dahlin DC, Pratt JH. Primary carcinoma of the fallopian tube. Am J Obstet Gynecol 1966;94(6):832-9.
32. Dodson MG, Ford JH, Averette HE. Clinical aspects of fallopian tube carcinoma. Obstet Gynecol 1970;36(6):935-9.
33. McMurray Earlene H., Jacobs AJ, Perez CA, Camel HM, Kao M-S, Galakatos A. Carcinoma of the fallopian tube: Management and sites of failure. Cancer 1986;58(9):2070-5.
34. Rose PG, Piver MS, Tsukada Y. Fallopian tube cancer: the Roswell Park experience. Cancer 1990;66:2661-7.
35. Navani SS, Alvarado-Cabrero I, Young RH, Scully RE. Endometrioid carcinoma of the fallopian tube: a clinicopathologic analysis of 26 cases. Gynecol Oncol 1996;63:371-8.
36. Zheng W, Sung CJ, Cao P, Zhang Z-F, Cai R, Godwin TA, et al. Early occurrence and prognostic significance of p53 alteration in primary carcinoma of the fallopian tube. Gynecol Oncol 1997;64:38-48.
37. Wolfson AH, Tralins KS, Greven KM, Kim RY, Corn BW, Kuettel MR, et al. Adenocarcinoma of the fallopian tube: results of a multi-institutional retrospective analysis of 72 patients. Int J Radiat Oncol Biol Phys 1998 Jan;40(1):71-6.
38. Rosen AC, Sevelda P, Klein M, Graf AH, Lahousen M, Reiner A, et al. A comparative analysis of management and prognosis in stage I and II fallopian tube carcinoma and epithelial ovarian cancer. Br J Cancer 1994 Mar;69(3):577-9.
39. Henderson SR, Harper RC, Salazar OM, Rudolph JH. Primary carcinoma of the fallopian tube: Difficulties of diagnosis and treatment. Gynecol Oncol 1977;5:168-79.
40. Eddy GL, Copeland LJ, Gershenson DM. Second-look laparotomy in fallopian tube carcinoma. Gynecol Oncol 1984;19:182-6.
41. Cormio G, Gabriele A, Maneo A, Marzola M, Lissoni A, Mangioni C. Second-look laparotomy in the management of fallopian tube carcinoma. Acta Obstet Gynecol Scand 1997 Apr;76(4):369-72.
42. Barakat RR, Rubin SC, Saigo PE, Chapman D., Lewis JLJr, Jones WB, et al. Cisplatin-based combination chemotherapy in carcinoma of the fallopian tube. Gynecol Oncol 1991;42(2):156-60.
43. Cormio G, Maneo A, Gabriele A, Rota SM, Lissoni A, Zanetta G. Primary carcinoma of the fallopian tube. A retrospective analysis of 47 patients. Ann Oncol 1996 Mar;7(3):271-5.
Précis: Peritoneal cytology, residual disease, stage, and symptoms at presentation are found to be predictors for survival in fallopian tube carcinoma.
Special thanks to: Wayne A. Christopherson, M.D., Robert P. Edwards