- Email: [email protected]
Histopathology of five tubes after treatment with methotrexate for a tubal pregnancy
Vol. 57, No.2, February 1992
FERTILITY AND STERILITY
Printed on acid-free paper in U.S.A.
Copyright Cl 1992 The American Fertility Society
Histopathology of five tubes after treatment with methotrexate for a tubal pregnancy
Sjarlot Kooi, M.D.* Frans H. P. M. van Etten, M.D.t Hans C. L. V. Kock, M.D.*:j: Maria Hospital, and Elisabeth Hospital, Tilburg, The Netherlands
Objective: To investigate whether treatment with a local high dose of methotrexate (MTX) for a tubal pregnancy hampers tubal repair. Design: From a total of 67 cases treated with MTX for their tubal pregnancy, two tubes ruptured shortly after the MTX injection and were removed. One patient insisted on a removal of the affected tube in spite of the fact that serum human chorionic gonadotropin was negative. Two patients had her tube removed ±1 year after the MTX treatment during a salpingo-oophorectomy because of endometriosis and cystic enlarged ovary. Results: In the ruptured tubes there were no findings supporting the idea that tubal rupture occurred because of the injection of MTX into the tube. There were also no signs of a hampered tubal-tissue response to the insult of invading chorionic tissue. Macroscopic and microscopic evaluation of the unruptured tubes showed no residual tubal destruction. In the tubal wall, all layers were normal and continuous, all unruptured tubes were gracile without distention. Conclusion: Methotrexate applied in a high local dose does not seem to hamper tubal response against the insult of invading chorionic tissue or to interfere with the ultimate tubal repair. Fertil Steril1992;57:341-5 Key Words: Histopathology, methotrexate, tubal pregnancy, tubal repair
The classical therapy for tubal pregnancy is surgical removal of the tube and conceptus. Recently, more conservative surgical, microsurgical, and nonsurgical methods have been advocated, particularly for those women who wish to preserve their fertility (1). It sounds reasonable that preservation of the affected tube by any method will result in a higher fertility potential, but this is difficult to prove (2). After termination of an intrauterine pregnancy, it is normal for the endometrium to regenerate completely. For the endosalpinx even after the complete resorption of a
Received June 7, 1991; and revised and accepted September 20, 1991. * Department of Gynecology and Obstetrics, Maria Hospital. t Department of Pathology, Elisabeth Hospital. :j: Reprints requests: Hans C. L. V. Kock, M.D., Department of Gynecology and Obstetrics, Maria Hospital, Tilburg, The Netherlands. Vol. 57, No.2, February 1992
tubal pregnancy, as reflected by the serum human chorionic gonadotropin (heG) values, this is not so (3-7). At this moment it is impossible to prove that one conservative treatment is better than the other (8, 9). Histopathology of tubes removed after experimental treatment, observative management, and conservative surgical or medical treatment might give a clue to the answer (10, 11). Because ectopic pregnancies (EPs) are only seen in human, there is no animal model to evaluate treatment modality. Nelson et al. (12) studied the healing of linear ampullary salpingotomy, primarily or secondarily in the nonpregnant rabbit oviduct. Pregnancy rates, nidation indices, and percentage of adhesion-free tubes were greater after primary closure, but these differences were not statistically valid. Menard et al. (13) studied the influence of local methotrexate (MTX) injection on genital tract mucosa in the nonpregnant rat model and did not find any toxicity. Kooi et al.
Histopathology of tubal pregnancies after MTX
341
,
It is difficult to study the influence of medical treatment of the EP in the human because many years of radical surgery did not yield identical information on the histopathology of tubal pregnancy. Simple questions are unanswered or controversial, such as, how far goes the trophoblastic spread, what factors influence the depth of chorionic invasion, how is the decidual response of the tube, and how are all these factors related to clinical signs and symptoms, tubal rupture, and tubal repair? Data on the histopathology after conservative surgical treatment of the tubal pregnancy are even more scarce. Stock (14) reviewed 15 tubes, removed because of recurrent tubal pregnancies. All were treated for their first EP by salpingotomy or salpingostomy. All former linear incisions could be identified. The tubal muscularis was well healed regardless of whether a salpingotomy or salpingostomy was performed. The muscular layer of the overlying broad ligament showed marked fibrosis in all cases. Residual histologic remnants of the previous EP could be identified in 5 of the 16 cases. Arenthals et al. (15) studied one fallopian tube 6 weeks after salpingotomy for a tubal pregnancy and found an intact lumen bounded by an intact villous mucosal membrane in all the cross-sectional segments. The result of tubal repair after medical treatment by a selective embryocide of the ectopic fetus by potassium chloride 20% in one case of combined IUP and extrauterine pregnancy (EUP) was studied by Porecco et al. (16). In his case report, a caesarean was done at 40 weeks. During this session the affected tube was removed. Histologic evaluation showed a hematosalpinx with necrotic chorionic villi and marked signs of fibrosis. There is only one publication on tubal histology during treatment with MTX; there are no data after treatment with prostaglandin (17). We were able to study the affected tubes of five patients after treatment with MTX.
despite the good clinical result. In cases 4 and 5 a salpingo-oophorectomy was done because of endometriosis and cystic enlargement. The removed tubes were serially blocked after fixation. Serial cross sections were stained with hematoxylin and eosin (H and E). Additional sections were stained with van Gieson's and Perls' iron stain. All sections were studied by light microscopy. RESULTS
During laparoscopic inspection and local MTX injection, the serosa of all tubes was intact. Serum hCG concentrations declined after treatment (Fig. 1). In case 1, the tubectomy was performed because of hematoperitoneum 3 days after the initial MTX treatment. The EP was localized in the right ampulla; the tube was distended at its distal position. In contrast to the findings at the initiallaparoscopy, the serosa was now ruptured. At gross inspection much of the distention of the tube was because of an intraluminal bloodclot rather than a conceptual mass. The former site of puncture could not be identified. At histology, most trophoblastic cells were found in the lumen of the tube. The placental villi were covered by cytotrophoblastic cells with nuclear clumping and fibrinoid exudation. These villi were partly degenerated and partly viable. Cross sections showed a predominant intraluminal spread of the trophoblast. Embryonic tissue could
hCG mlU/mL ,100 mg MTX local 10000
I
I
I
1000
100
MATERIALS AND METHODS
All five patients were treated with injection of MTX, 100 mg dissolved in 4 mL isotonic natrium chloride, pH 8.2 to 8.4, using a 22-gauge needle, directly into the tubal swelling after having injected 10 mL adrenaline 1:80,000 under the serosa of the mesosalpinx under laparoscopic control as reported earlier (9). All five patients had their tubes removed at laparotomy. The interval between MTX treatment and surgical removal varied from 3 days to 1 year. In two cases, tubal rupture made surgical removal necessary (cases 1 and 2). In case 3, the tube was removed because the patient insisted on removal 342
Kooi et al.
Histopathology of tubal pregnancies after MTX
10
* 0123456789mnU~Mmrev.w
DAYS
Figure 1 Serum hCG disappearance curves in all five patients with a tubal pregnancy treated with local MTX injection of 100 mg (tl. A laparotomy because of tubal rupture was done in cases 1 and 2 on days 3 and 12, respectively (tl. Case 2 received additional 1M MTX therapy (tl. The other three cases had an uneventful clinical course followed by a tubectomy after a variable time interval ranging from 1 to 13 months. Serum hCG concentrations are expressed in mlU/mL First IRP. Case 1-. -; case 2 - + -; case 3 ---*---; case 4 --- D ---; case 5 - X-.
Fertility and Sterility
be identified intraluminally. This showed hardly any evidence of autolytic changes (Fig. 2). The tubal wall had a marked decidual response (Fig. 3). Inflammatory changes were identified in all three coats of the tube. The muscular layer was thin and partly destroyed by the hemorrhage. On the serosal side was a fibrinous inflammatory reaction. There was a complete rupture of the tubal wall in the vicinity of embyronic tissue. At this site, the number of viable placental villi was not different from these numbers in the other parts of the tube. The decidual reaction in the part of the tube adjacent to the tubal rupture seemed more marked. The tubal folds showed changes compatible with acute salpinigitis over the whole tubal length. At laparoscopy the tubal pregnancy in case 2 seemed to fill the whole left tube. The local injection of 100 mg MTX resulted in an initial fall of the serum hCG concentrations. Additional systemic MTX therapy was given because of steady hCG levels. On day 12, a laparotomy was performed because of intraperitoneal bleeding, and the ruptured tube was removed (Fig. 1). It was distended over the whole tubal length as it had been 12 days before at laparoscopy. Histologic examination showed a predominantly extraluminal spread of trophoblast. The rupture of the serosa did not communicate with the tubal lumen. The tubal lumen was of normal size, not distented, and bordered by an intact continuous mucosa. The folds were edematous with connective tissue cords. All sections showed signs of chronic salpingitis. There were swelling and edema of the myosalpinx, and the circumferential lymphatic vessels showed marked engorgement. Trophoblastic invasion into the myosalpinx and the concommitant
Figure 2 Case 1. Hand E staining. Magnification was 80X. Fetal tissue in the tubal lumen. It showed remarkable little evidence of autolytic changes. Despite a sharp decrease in serum hCG tubal rupture occurred. Vol. 57, No.2, February 1992
Figure 3 Case 1. Hand E staining. Magnification was 400X. The tubal wall had a marked decidual response. The decidual reaction in the part of the tubal rupture seemed more marked.
hemorrhage marked the complete destruction of the muscular layer. Here, sheets of cytotrophoblasts, syncytiotrophoblasts, tissue necrosis, and inflammatory infiltration were seen. Most of the placental villi were necrotic; some were still viable. The tubal wall showed a patchy decidual response. Embryonic parts were not seen. Cases 3, 4, and 5 had a laparotomy and tubal removal for reasons not related to the clinical course of the EP. In these patients, the serum hCG concentration was negative before surgery. In case 3, an intact ampullary tubal pregnancy was treated with a local MTX injection. The initial serum hCG concentration of 1,590 mIU/mL (First International Reference Preparation [IRP]) returned to normal within a few days. Three weeks later the affected tube was removed. On macroscopic inspection it was normal. All cross-sectional segments showed an intact lumen, bounded by an intact villous mucosal membrane and no constrictions. The folds were gracile, normally shaped, without adhesions or fibrosis. An implantation site could not be identified. There were no signs of pre-existent salpingitis. All layers in the tubal wall were normal and continuous. Locally, there was a light inflammatory reaction with granulocytes and some edema. In case 4, the fallopian tube was removed> 1 year after tubal pregnancy. This patient was treated twice for an EP. Her first tubal gestation was localized on the left side. This was treated with intramuscular (IM) MTX, 5 X 50 mg. The initial serum hCG concentration of 2,235 mIU /mL returned to 0 in 14 days. Several months later she had a tubal pregnancy on the right side that was treated by a MTX injection of 100 mg into the tubal swelling. The left tube was Kooi et al.
Histopathology of tubal pregnancies after MTX
343
normal on gross inspection during this laparoscopy. The clinical source was uneventful again; initial serum hCG levels of 1,710 mIU/mL returned to 0 in 13 days. Follow-up hysterosalpingography showed tubal patency on both sides. The right tube was removed during an ovariectomy for endometriosis 13 months after the second tubal pregnancy. On gross inspection, the tube was gracile and without distention. Cross sections showed an intact lumen with a continuous mucosal layer, without local constrictions. The endosalpinx showed signs of chronic salpingitis with clumsy folds. These folds showed fibrosis and a little edema. This was seen all along the tube, and the muscular layer was intact. An old hematoma in the submucosa layer with iron pigment in macrophages and little calcification marked the former implantation site. An intact mucosal layer covered this site. Case 5 had a left salpingo-oophorectomy because of a cystic enlarged ovary 8 months after an ipsilateral tubal pregnancy treated with a local MTX injection. The intact tubal pregnancy was localized in the left ampulla. The initial serum hCG concentration of 3,095 mIU /mL returned to normal within 2.5 weeks. Thereafter she experienced a regular ovulatory cycle. Eight months later she had a salpingo-oophorectomy. At gross inspection, only a few adhesions were seen from the tube to the adjacent ovary. The tube was mobile and gracile. Cross-sectional segments showed an intact tubal lumen bounded by an intact villous mucosal membrane. The folds were gracile and normally shaped. An implantation site could not be identified. In the tubal wall all layers were normal and continuous. DISCUSSION
In an overview of literature on MTX treatment of tubal pregnancies, the overall risk on tubal rupture is approximately 5% (1,18). Although high initial serum hCG concentrations, short hCG doubling times, and positive fetal heart rate on ultrasound all bear an increased risk of tubal rupture, there is no reliable parameter for the prediction of tubal rupture or bleeding for the individual patient (1, 16). In our patients it was even in retrospect impossible to distinguish between the two patients whose tubes (cases 1 and 2) eventually ruptured from the 65 patients whose tubes did not. Both tubal diameter and pregnancy duration were comparable; the initial serum hCG concentrations were 11,800 and 8,900 mIU/mL, but several other patients without tubal rupture under MTX treatment had such a high initial serum hCG concentrations (4). Furthermore, a steep de344
Kooi et al.
Histopathology of tubal pregnancies after MTX
cline in serum hCG was seen after MTX injection in both patients, pointing to an inhibition of trophoblast proliferation. Phillipe and Satge (19) concluded in their study that the tubal rupture is independent of the tubal distention, but most likely related to the development of the trophoblast and its implantation site. In contrast, Stock (20) concluded that the pathophysiological mechanism for spontaneous tubal rupture and the site of rupture might be related to local distention of the tube because of hemorrhage and not to invasion of the wall by chorionic villi. The trophoblastic spread was predominantly intraluminal in case 1 and extraluminal in case 2. Embryonic tissue was seen in case 1. This was situated near the tubal rupture. In the vicinity of the tubal rupture, the number of viable villi was comparable with the other sites and there was a marked decidual response. This suggests that rupture did not occur because of an enhanced local trophoblastic activity or a lack of tubal reaction to the insult of trophoblast invasion. It is remarkable that the embryonic tissue in case 1 showed little signs of autolysis (Fig. 2), although the decrease in serum hCG concentrations (Fig. 1) pointed to a decrease in the number of viable, hCGsecreting, chorionic cells. In 1980, Budowick et al. (21) studied the histopathology of the developing tubal EP and reported a predominantly extraluminal spread in most tubal pregnancies. The study of Randall et al. (22) on tubal placentation confirmed these data. Later studies of Pauerstein et al. (10) in 1986 and Senterman and others (23) in 1988 did not support this point of view. Pauerstein et al. (10) studied the trophoblastic spread in 25 tubal pregnancies. Nineteen cases had predominantly an intraluminal trophoblastic spread. The tube was unruptured in 86% of these cases. From the 6 cases with a predominantly extraluminal spread, 5 ruptured. Senterman et al. (23) reported on 84 ampullary tubal pregnancies: 47 were intraluminal, 6 were extraluminal, and 31 showed intraluminal as well as extraluminal spread. Nine of these tubes were ruptured; all but one had a combined intraluminal and extraluminal spread. These data indicate that an intraluminal spread leads to an earlier diagnosis and treatment or that extraluminal spread is more likely to cause rupture. It might well be that all tubal pregnancies have an initially intraluminallocalization and develop into an extraluminar stage (21). During that stage, tubal rupture might occur (23). The histologic phenomena of implantation in tubal pregnancy are very alike those of a normal intrauterine gestation, with allowance for the strikFertility and Sterility
ing failure of the tube to mount a proper decidual response. There is a general agreement that here the decidual cells do not form a continuous and intact layer but merely a patchy pattern (20, 24, 25). To differentiate between decidual transformation and decidual-like changes is difficult, mainly because of the discontinuity (20, 23). Both our cases with tubal rupture had an extended decidual response in tubal wall and endosalpingeal folds. This decidual response was notable at the site of the tubal rupture, which is compatible with the view that invasion of the tubal wall by placental villi is not merely because of its failure to mount a decidual reaction (22). Pauerstein et al. (10) also noticed this decidual reaction in 7 of 25 tubes from EUPs. Four of these were ruptured and 3 were intact. In cases 3, 4, and 5, the absence of serum heG reflects the absence of viable persistent trophoblastic tissue. The ultimate tubal repair after MTX treatment in these three tubes could be studied. The muscular and mucosal layer of these tubes showed no signs of former insult by the placental villi or the hematoma, nor was there evidence that MTX hampers the process of tubal repair. For instances, the muscular and mucosal layers were continuous, and the endosalpingeal folds were identical throughout the whole tube. The signs of chronic salpingitis were diffuse and throughout the whole tubal length and therefore interpreted as pre-existing. From our data, although they concern a limited number of observations, it can be concluded that there is no evidence that: (1) tubal rupture during MTX therapy is because of a locally enhanced chorionic activity; (2) tubal rupture is associated with a lack of decidual response in the tube; and (3) MTX applied in a high local dose hampers the tubal reaction to the insult of trophoblast invasion and the ultimate tubal repair. There was evidence that MTX in a high local dose gives necrosis of chorionic tissue.
REFERENCES 1. Leach RE, Ory SJ. Modern management of ectopic pregnancy. J Reprod Med 1989;34:324-37. 2. Oelsner G, Morad J, Carp H, Mashiach S, Serr DM. Reproductive performance following conservative management of tubal pregnancy. Br J Obstet Gynaecol 1988;94:1078-83. 3. Olson CM, Holt JA, Alengant E, Greco S, Lumpkin JR, Geanon GD. Limitations of qualitative serum beta-hCG assays in the diagnosis of ectopic pregnancy. J Reprod Med 1983;28:838-42.
Vol. 57, No.2, February 1992
4. Lonky NM, Sauer MV. Ectopic pregnancy with shock and undetectable human chorionic gonadotropin. A case report. J Reprod Med 1987;32:559-60. 5. Taylor RN, Padula C, Goldsmith PC. Pitfall in the diagnosis of ectopic pregnancy: immunocytochemical evaluation in a patient with false-negative serum hCG levels. Obstet Gynecol 1988;71:1035-8. 6. Gomel V, Filmar S. Arrested tubal pregnancy. Fertil Steril 1987;48:1043-7. 7. Haney AF. Bilateral tubal occlusion secondary to asymptomatic ectopic pregnancies. Obstet Gynecol 1986;67:52S4S. 8. Thornburn J, Philipson M, Lindblom B. Fertility after ectopic pregnancy in relation to background factors and surgical treatment. Fertil Steril 1988;49:595-601. 9. Kooi S, Kock HCLV. Treatment of tubal pregnancy by local injection of methotrexate after adrenaline injection into the mesosalpinx: a report of 25 patients. Fertil Steril 1990;54: 580-4. 10. Pauerstein CJ, Croxatto HB, Eddy CA, Ramzy I, Walters MD. Anatomy and pathology of tubal pregnancy. Obstet GynecoI1986;67:301-7. 11. Ory SJ. Chemotherapy for ectopic pregnancy. Obstet Gynecol Clin North Am 1991;18:123-33. 12. Nelson LM, Margara RA, Winston RML. Primary and secondary closure of ampullary salpingotomy compared in the rabbit. Fertil Steril 1986;45:292-5. 13. Menard A, Crequat J, Mandelbrot L, Hauuy J -P, Madelenat P. Treatment of un ruptured tubal pregnancy by local injection of methotrexate under transvaginal sonographic control. Fertil Steril 1990;54:47-50. 14. Stock RJ. Histopathology and fallopian tubes with recurrent tubal pregnancy. Obstet GynecoI1990;75:9-14. 15. Arenthals JV, Dekker N, Hanselaar T, Exalto N. Morphology of the fallopian tube 6 weeks after salpingotomy: a case report. Eur J Obstet Gynecol Reprod Bioi 1988;28:79-84. 16. Porreco RP, Burke MS, Parker DW. Selective embryocide in the nonsurgical management of combined intrauterine and extrauterine pregnancy. Obstet Gynecol 1990;75:498-501. 17. Tulandi T, Bret PM, Mostafa A, Senterman M. Treatment of ectopic pregnancy by transvaginal intratubal methotrexate administration. Obstet GynecoI1991;77:627-30. 18. Kooi S, Kock HCLV. A review of the literature on non-surgical treatment in tubal pregnancies. Obstet Gynecol Surv In press. 19. Philippe E, Satge D. Etude histologique de 100 trompes gravides. J Gynecol Obstet Bioi Reprod (Paris) 1988;17:467-76. 20. Stock RJ. Histopathologic changes in tubal pregnancy. J Reprod Med 1985;30:923-8. 21. Budowick M, Johnson TRB, Genardy R Jr, Parmley TH, Woodruff JD. The histopathology of the developing tubal ectopic pregnancy. Fertil Steril1980;34:169-71. 22. Randall S, Buckley CH, Fox H. Placentation in the fallopian tube. Int J Gynecol Pathol 1987;6:132-9. 23. Senterman M, Jibodh R, Tulandi T. Histopathologic study of ampullary and isthmic tubal ectopic pregnancy. Am J Obstet Gynecol 1988;159:939-41. 24. Novak ER, Woodruff JD, editors. Novak's gynecologic and obstetric pathology. 7th ed. Philadelphia: Saunders 1974:486. 25. Parmley TH. The histopathology of tubal pregnancy. Clin Obstet and GynecoI1987;30:1l9-28.
Kooi et al.
Histopathology of tubal pregnancies after MTX
345
FERTILITY AND STERILITY
Printed on acid-free paper in U.S.A.
Copyright Cl 1992 The American Fertility Society
Histopathology of five tubes after treatment with methotrexate for a tubal pregnancy
Sjarlot Kooi, M.D.* Frans H. P. M. van Etten, M.D.t Hans C. L. V. Kock, M.D.*:j: Maria Hospital, and Elisabeth Hospital, Tilburg, The Netherlands
Objective: To investigate whether treatment with a local high dose of methotrexate (MTX) for a tubal pregnancy hampers tubal repair. Design: From a total of 67 cases treated with MTX for their tubal pregnancy, two tubes ruptured shortly after the MTX injection and were removed. One patient insisted on a removal of the affected tube in spite of the fact that serum human chorionic gonadotropin was negative. Two patients had her tube removed ±1 year after the MTX treatment during a salpingo-oophorectomy because of endometriosis and cystic enlarged ovary. Results: In the ruptured tubes there were no findings supporting the idea that tubal rupture occurred because of the injection of MTX into the tube. There were also no signs of a hampered tubal-tissue response to the insult of invading chorionic tissue. Macroscopic and microscopic evaluation of the unruptured tubes showed no residual tubal destruction. In the tubal wall, all layers were normal and continuous, all unruptured tubes were gracile without distention. Conclusion: Methotrexate applied in a high local dose does not seem to hamper tubal response against the insult of invading chorionic tissue or to interfere with the ultimate tubal repair. Fertil Steril1992;57:341-5 Key Words: Histopathology, methotrexate, tubal pregnancy, tubal repair
The classical therapy for tubal pregnancy is surgical removal of the tube and conceptus. Recently, more conservative surgical, microsurgical, and nonsurgical methods have been advocated, particularly for those women who wish to preserve their fertility (1). It sounds reasonable that preservation of the affected tube by any method will result in a higher fertility potential, but this is difficult to prove (2). After termination of an intrauterine pregnancy, it is normal for the endometrium to regenerate completely. For the endosalpinx even after the complete resorption of a
Received June 7, 1991; and revised and accepted September 20, 1991. * Department of Gynecology and Obstetrics, Maria Hospital. t Department of Pathology, Elisabeth Hospital. :j: Reprints requests: Hans C. L. V. Kock, M.D., Department of Gynecology and Obstetrics, Maria Hospital, Tilburg, The Netherlands. Vol. 57, No.2, February 1992
tubal pregnancy, as reflected by the serum human chorionic gonadotropin (heG) values, this is not so (3-7). At this moment it is impossible to prove that one conservative treatment is better than the other (8, 9). Histopathology of tubes removed after experimental treatment, observative management, and conservative surgical or medical treatment might give a clue to the answer (10, 11). Because ectopic pregnancies (EPs) are only seen in human, there is no animal model to evaluate treatment modality. Nelson et al. (12) studied the healing of linear ampullary salpingotomy, primarily or secondarily in the nonpregnant rabbit oviduct. Pregnancy rates, nidation indices, and percentage of adhesion-free tubes were greater after primary closure, but these differences were not statistically valid. Menard et al. (13) studied the influence of local methotrexate (MTX) injection on genital tract mucosa in the nonpregnant rat model and did not find any toxicity. Kooi et al.
Histopathology of tubal pregnancies after MTX
341
,
It is difficult to study the influence of medical treatment of the EP in the human because many years of radical surgery did not yield identical information on the histopathology of tubal pregnancy. Simple questions are unanswered or controversial, such as, how far goes the trophoblastic spread, what factors influence the depth of chorionic invasion, how is the decidual response of the tube, and how are all these factors related to clinical signs and symptoms, tubal rupture, and tubal repair? Data on the histopathology after conservative surgical treatment of the tubal pregnancy are even more scarce. Stock (14) reviewed 15 tubes, removed because of recurrent tubal pregnancies. All were treated for their first EP by salpingotomy or salpingostomy. All former linear incisions could be identified. The tubal muscularis was well healed regardless of whether a salpingotomy or salpingostomy was performed. The muscular layer of the overlying broad ligament showed marked fibrosis in all cases. Residual histologic remnants of the previous EP could be identified in 5 of the 16 cases. Arenthals et al. (15) studied one fallopian tube 6 weeks after salpingotomy for a tubal pregnancy and found an intact lumen bounded by an intact villous mucosal membrane in all the cross-sectional segments. The result of tubal repair after medical treatment by a selective embryocide of the ectopic fetus by potassium chloride 20% in one case of combined IUP and extrauterine pregnancy (EUP) was studied by Porecco et al. (16). In his case report, a caesarean was done at 40 weeks. During this session the affected tube was removed. Histologic evaluation showed a hematosalpinx with necrotic chorionic villi and marked signs of fibrosis. There is only one publication on tubal histology during treatment with MTX; there are no data after treatment with prostaglandin (17). We were able to study the affected tubes of five patients after treatment with MTX.
despite the good clinical result. In cases 4 and 5 a salpingo-oophorectomy was done because of endometriosis and cystic enlargement. The removed tubes were serially blocked after fixation. Serial cross sections were stained with hematoxylin and eosin (H and E). Additional sections were stained with van Gieson's and Perls' iron stain. All sections were studied by light microscopy. RESULTS
During laparoscopic inspection and local MTX injection, the serosa of all tubes was intact. Serum hCG concentrations declined after treatment (Fig. 1). In case 1, the tubectomy was performed because of hematoperitoneum 3 days after the initial MTX treatment. The EP was localized in the right ampulla; the tube was distended at its distal position. In contrast to the findings at the initiallaparoscopy, the serosa was now ruptured. At gross inspection much of the distention of the tube was because of an intraluminal bloodclot rather than a conceptual mass. The former site of puncture could not be identified. At histology, most trophoblastic cells were found in the lumen of the tube. The placental villi were covered by cytotrophoblastic cells with nuclear clumping and fibrinoid exudation. These villi were partly degenerated and partly viable. Cross sections showed a predominant intraluminal spread of the trophoblast. Embryonic tissue could
hCG mlU/mL ,100 mg MTX local 10000
I
I
I
1000
100
MATERIALS AND METHODS
All five patients were treated with injection of MTX, 100 mg dissolved in 4 mL isotonic natrium chloride, pH 8.2 to 8.4, using a 22-gauge needle, directly into the tubal swelling after having injected 10 mL adrenaline 1:80,000 under the serosa of the mesosalpinx under laparoscopic control as reported earlier (9). All five patients had their tubes removed at laparotomy. The interval between MTX treatment and surgical removal varied from 3 days to 1 year. In two cases, tubal rupture made surgical removal necessary (cases 1 and 2). In case 3, the tube was removed because the patient insisted on removal 342
Kooi et al.
Histopathology of tubal pregnancies after MTX
10
* 0123456789mnU~Mmrev.w
DAYS
Figure 1 Serum hCG disappearance curves in all five patients with a tubal pregnancy treated with local MTX injection of 100 mg (tl. A laparotomy because of tubal rupture was done in cases 1 and 2 on days 3 and 12, respectively (tl. Case 2 received additional 1M MTX therapy (tl. The other three cases had an uneventful clinical course followed by a tubectomy after a variable time interval ranging from 1 to 13 months. Serum hCG concentrations are expressed in mlU/mL First IRP. Case 1-. -; case 2 - + -; case 3 ---*---; case 4 --- D ---; case 5 - X-.
Fertility and Sterility
be identified intraluminally. This showed hardly any evidence of autolytic changes (Fig. 2). The tubal wall had a marked decidual response (Fig. 3). Inflammatory changes were identified in all three coats of the tube. The muscular layer was thin and partly destroyed by the hemorrhage. On the serosal side was a fibrinous inflammatory reaction. There was a complete rupture of the tubal wall in the vicinity of embyronic tissue. At this site, the number of viable placental villi was not different from these numbers in the other parts of the tube. The decidual reaction in the part of the tube adjacent to the tubal rupture seemed more marked. The tubal folds showed changes compatible with acute salpinigitis over the whole tubal length. At laparoscopy the tubal pregnancy in case 2 seemed to fill the whole left tube. The local injection of 100 mg MTX resulted in an initial fall of the serum hCG concentrations. Additional systemic MTX therapy was given because of steady hCG levels. On day 12, a laparotomy was performed because of intraperitoneal bleeding, and the ruptured tube was removed (Fig. 1). It was distended over the whole tubal length as it had been 12 days before at laparoscopy. Histologic examination showed a predominantly extraluminal spread of trophoblast. The rupture of the serosa did not communicate with the tubal lumen. The tubal lumen was of normal size, not distented, and bordered by an intact continuous mucosa. The folds were edematous with connective tissue cords. All sections showed signs of chronic salpingitis. There were swelling and edema of the myosalpinx, and the circumferential lymphatic vessels showed marked engorgement. Trophoblastic invasion into the myosalpinx and the concommitant
Figure 2 Case 1. Hand E staining. Magnification was 80X. Fetal tissue in the tubal lumen. It showed remarkable little evidence of autolytic changes. Despite a sharp decrease in serum hCG tubal rupture occurred. Vol. 57, No.2, February 1992
Figure 3 Case 1. Hand E staining. Magnification was 400X. The tubal wall had a marked decidual response. The decidual reaction in the part of the tubal rupture seemed more marked.
hemorrhage marked the complete destruction of the muscular layer. Here, sheets of cytotrophoblasts, syncytiotrophoblasts, tissue necrosis, and inflammatory infiltration were seen. Most of the placental villi were necrotic; some were still viable. The tubal wall showed a patchy decidual response. Embryonic parts were not seen. Cases 3, 4, and 5 had a laparotomy and tubal removal for reasons not related to the clinical course of the EP. In these patients, the serum hCG concentration was negative before surgery. In case 3, an intact ampullary tubal pregnancy was treated with a local MTX injection. The initial serum hCG concentration of 1,590 mIU/mL (First International Reference Preparation [IRP]) returned to normal within a few days. Three weeks later the affected tube was removed. On macroscopic inspection it was normal. All cross-sectional segments showed an intact lumen, bounded by an intact villous mucosal membrane and no constrictions. The folds were gracile, normally shaped, without adhesions or fibrosis. An implantation site could not be identified. There were no signs of pre-existent salpingitis. All layers in the tubal wall were normal and continuous. Locally, there was a light inflammatory reaction with granulocytes and some edema. In case 4, the fallopian tube was removed> 1 year after tubal pregnancy. This patient was treated twice for an EP. Her first tubal gestation was localized on the left side. This was treated with intramuscular (IM) MTX, 5 X 50 mg. The initial serum hCG concentration of 2,235 mIU /mL returned to 0 in 14 days. Several months later she had a tubal pregnancy on the right side that was treated by a MTX injection of 100 mg into the tubal swelling. The left tube was Kooi et al.
Histopathology of tubal pregnancies after MTX
343
normal on gross inspection during this laparoscopy. The clinical source was uneventful again; initial serum hCG levels of 1,710 mIU/mL returned to 0 in 13 days. Follow-up hysterosalpingography showed tubal patency on both sides. The right tube was removed during an ovariectomy for endometriosis 13 months after the second tubal pregnancy. On gross inspection, the tube was gracile and without distention. Cross sections showed an intact lumen with a continuous mucosal layer, without local constrictions. The endosalpinx showed signs of chronic salpingitis with clumsy folds. These folds showed fibrosis and a little edema. This was seen all along the tube, and the muscular layer was intact. An old hematoma in the submucosa layer with iron pigment in macrophages and little calcification marked the former implantation site. An intact mucosal layer covered this site. Case 5 had a left salpingo-oophorectomy because of a cystic enlarged ovary 8 months after an ipsilateral tubal pregnancy treated with a local MTX injection. The intact tubal pregnancy was localized in the left ampulla. The initial serum hCG concentration of 3,095 mIU /mL returned to normal within 2.5 weeks. Thereafter she experienced a regular ovulatory cycle. Eight months later she had a salpingo-oophorectomy. At gross inspection, only a few adhesions were seen from the tube to the adjacent ovary. The tube was mobile and gracile. Cross-sectional segments showed an intact tubal lumen bounded by an intact villous mucosal membrane. The folds were gracile and normally shaped. An implantation site could not be identified. In the tubal wall all layers were normal and continuous. DISCUSSION
In an overview of literature on MTX treatment of tubal pregnancies, the overall risk on tubal rupture is approximately 5% (1,18). Although high initial serum hCG concentrations, short hCG doubling times, and positive fetal heart rate on ultrasound all bear an increased risk of tubal rupture, there is no reliable parameter for the prediction of tubal rupture or bleeding for the individual patient (1, 16). In our patients it was even in retrospect impossible to distinguish between the two patients whose tubes (cases 1 and 2) eventually ruptured from the 65 patients whose tubes did not. Both tubal diameter and pregnancy duration were comparable; the initial serum hCG concentrations were 11,800 and 8,900 mIU/mL, but several other patients without tubal rupture under MTX treatment had such a high initial serum hCG concentrations (4). Furthermore, a steep de344
Kooi et al.
Histopathology of tubal pregnancies after MTX
cline in serum hCG was seen after MTX injection in both patients, pointing to an inhibition of trophoblast proliferation. Phillipe and Satge (19) concluded in their study that the tubal rupture is independent of the tubal distention, but most likely related to the development of the trophoblast and its implantation site. In contrast, Stock (20) concluded that the pathophysiological mechanism for spontaneous tubal rupture and the site of rupture might be related to local distention of the tube because of hemorrhage and not to invasion of the wall by chorionic villi. The trophoblastic spread was predominantly intraluminal in case 1 and extraluminal in case 2. Embryonic tissue was seen in case 1. This was situated near the tubal rupture. In the vicinity of the tubal rupture, the number of viable villi was comparable with the other sites and there was a marked decidual response. This suggests that rupture did not occur because of an enhanced local trophoblastic activity or a lack of tubal reaction to the insult of trophoblast invasion. It is remarkable that the embryonic tissue in case 1 showed little signs of autolysis (Fig. 2), although the decrease in serum hCG concentrations (Fig. 1) pointed to a decrease in the number of viable, hCGsecreting, chorionic cells. In 1980, Budowick et al. (21) studied the histopathology of the developing tubal EP and reported a predominantly extraluminal spread in most tubal pregnancies. The study of Randall et al. (22) on tubal placentation confirmed these data. Later studies of Pauerstein et al. (10) in 1986 and Senterman and others (23) in 1988 did not support this point of view. Pauerstein et al. (10) studied the trophoblastic spread in 25 tubal pregnancies. Nineteen cases had predominantly an intraluminal trophoblastic spread. The tube was unruptured in 86% of these cases. From the 6 cases with a predominantly extraluminal spread, 5 ruptured. Senterman et al. (23) reported on 84 ampullary tubal pregnancies: 47 were intraluminal, 6 were extraluminal, and 31 showed intraluminal as well as extraluminal spread. Nine of these tubes were ruptured; all but one had a combined intraluminal and extraluminal spread. These data indicate that an intraluminal spread leads to an earlier diagnosis and treatment or that extraluminal spread is more likely to cause rupture. It might well be that all tubal pregnancies have an initially intraluminallocalization and develop into an extraluminar stage (21). During that stage, tubal rupture might occur (23). The histologic phenomena of implantation in tubal pregnancy are very alike those of a normal intrauterine gestation, with allowance for the strikFertility and Sterility
ing failure of the tube to mount a proper decidual response. There is a general agreement that here the decidual cells do not form a continuous and intact layer but merely a patchy pattern (20, 24, 25). To differentiate between decidual transformation and decidual-like changes is difficult, mainly because of the discontinuity (20, 23). Both our cases with tubal rupture had an extended decidual response in tubal wall and endosalpingeal folds. This decidual response was notable at the site of the tubal rupture, which is compatible with the view that invasion of the tubal wall by placental villi is not merely because of its failure to mount a decidual reaction (22). Pauerstein et al. (10) also noticed this decidual reaction in 7 of 25 tubes from EUPs. Four of these were ruptured and 3 were intact. In cases 3, 4, and 5, the absence of serum heG reflects the absence of viable persistent trophoblastic tissue. The ultimate tubal repair after MTX treatment in these three tubes could be studied. The muscular and mucosal layer of these tubes showed no signs of former insult by the placental villi or the hematoma, nor was there evidence that MTX hampers the process of tubal repair. For instances, the muscular and mucosal layers were continuous, and the endosalpingeal folds were identical throughout the whole tube. The signs of chronic salpingitis were diffuse and throughout the whole tubal length and therefore interpreted as pre-existing. From our data, although they concern a limited number of observations, it can be concluded that there is no evidence that: (1) tubal rupture during MTX therapy is because of a locally enhanced chorionic activity; (2) tubal rupture is associated with a lack of decidual response in the tube; and (3) MTX applied in a high local dose hampers the tubal reaction to the insult of trophoblast invasion and the ultimate tubal repair. There was evidence that MTX in a high local dose gives necrosis of chorionic tissue.
REFERENCES 1. Leach RE, Ory SJ. Modern management of ectopic pregnancy. J Reprod Med 1989;34:324-37. 2. Oelsner G, Morad J, Carp H, Mashiach S, Serr DM. Reproductive performance following conservative management of tubal pregnancy. Br J Obstet Gynaecol 1988;94:1078-83. 3. Olson CM, Holt JA, Alengant E, Greco S, Lumpkin JR, Geanon GD. Limitations of qualitative serum beta-hCG assays in the diagnosis of ectopic pregnancy. J Reprod Med 1983;28:838-42.
Vol. 57, No.2, February 1992
4. Lonky NM, Sauer MV. Ectopic pregnancy with shock and undetectable human chorionic gonadotropin. A case report. J Reprod Med 1987;32:559-60. 5. Taylor RN, Padula C, Goldsmith PC. Pitfall in the diagnosis of ectopic pregnancy: immunocytochemical evaluation in a patient with false-negative serum hCG levels. Obstet Gynecol 1988;71:1035-8. 6. Gomel V, Filmar S. Arrested tubal pregnancy. Fertil Steril 1987;48:1043-7. 7. Haney AF. Bilateral tubal occlusion secondary to asymptomatic ectopic pregnancies. Obstet Gynecol 1986;67:52S4S. 8. Thornburn J, Philipson M, Lindblom B. Fertility after ectopic pregnancy in relation to background factors and surgical treatment. Fertil Steril 1988;49:595-601. 9. Kooi S, Kock HCLV. Treatment of tubal pregnancy by local injection of methotrexate after adrenaline injection into the mesosalpinx: a report of 25 patients. Fertil Steril 1990;54: 580-4. 10. Pauerstein CJ, Croxatto HB, Eddy CA, Ramzy I, Walters MD. Anatomy and pathology of tubal pregnancy. Obstet GynecoI1986;67:301-7. 11. Ory SJ. Chemotherapy for ectopic pregnancy. Obstet Gynecol Clin North Am 1991;18:123-33. 12. Nelson LM, Margara RA, Winston RML. Primary and secondary closure of ampullary salpingotomy compared in the rabbit. Fertil Steril 1986;45:292-5. 13. Menard A, Crequat J, Mandelbrot L, Hauuy J -P, Madelenat P. Treatment of un ruptured tubal pregnancy by local injection of methotrexate under transvaginal sonographic control. Fertil Steril 1990;54:47-50. 14. Stock RJ. Histopathology and fallopian tubes with recurrent tubal pregnancy. Obstet GynecoI1990;75:9-14. 15. Arenthals JV, Dekker N, Hanselaar T, Exalto N. Morphology of the fallopian tube 6 weeks after salpingotomy: a case report. Eur J Obstet Gynecol Reprod Bioi 1988;28:79-84. 16. Porreco RP, Burke MS, Parker DW. Selective embryocide in the nonsurgical management of combined intrauterine and extrauterine pregnancy. Obstet Gynecol 1990;75:498-501. 17. Tulandi T, Bret PM, Mostafa A, Senterman M. Treatment of ectopic pregnancy by transvaginal intratubal methotrexate administration. Obstet GynecoI1991;77:627-30. 18. Kooi S, Kock HCLV. A review of the literature on non-surgical treatment in tubal pregnancies. Obstet Gynecol Surv In press. 19. Philippe E, Satge D. Etude histologique de 100 trompes gravides. J Gynecol Obstet Bioi Reprod (Paris) 1988;17:467-76. 20. Stock RJ. Histopathologic changes in tubal pregnancy. J Reprod Med 1985;30:923-8. 21. Budowick M, Johnson TRB, Genardy R Jr, Parmley TH, Woodruff JD. The histopathology of the developing tubal ectopic pregnancy. Fertil Steril1980;34:169-71. 22. Randall S, Buckley CH, Fox H. Placentation in the fallopian tube. Int J Gynecol Pathol 1987;6:132-9. 23. Senterman M, Jibodh R, Tulandi T. Histopathologic study of ampullary and isthmic tubal ectopic pregnancy. Am J Obstet Gynecol 1988;159:939-41. 24. Novak ER, Woodruff JD, editors. Novak's gynecologic and obstetric pathology. 7th ed. Philadelphia: Saunders 1974:486. 25. Parmley TH. The histopathology of tubal pregnancy. Clin Obstet and GynecoI1987;30:1l9-28.
Kooi et al.
Histopathology of tubal pregnancies after MTX
345