SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Surgical technique
  4. Discussion
  5. Conclusion
  6. References

Traditionally radical hysterectomy has formed the mainstay of treatment for early stage cervical carcinoma. More recently radical trachelectomy and laparoscopic lymphadenectomy have been introduced to allow preservation of fertility. We present a new approach to fertility-sparing surgery, namely abdominal radical trachelectomy. The technique is similar to a standard radical hysterectomy and lymphadenectomy. In our technique the ovarian vessels are not ligated and, following lymphadenectomy and skeletonisation of the uterine arteries, the cervix, parametrium and vaginal cuff are excised. The residuum of the cervix is then sutured to the vagina and the uterine ateries re-anastomosed.

Traditionally the treatment for invasive cervical carcinoma which has progressed beyond microinvasion has been radical hysterectomy. Long term experience of radical surgery for Stage Ib carcinoma has shown that it produces excellent results in terms of survival but that morbidity may be significant, and there is always loss of potential for future childbearing. Increasingly, large numbers of young women (24–35 years) are being diagnosed with cervical cancer1. Clearly, the loss of fertility in these women can be devastating.

In recent years there has been a move towards more conservative approaches for the treatment of cervical carcinoma. Conisation of the cervix has become acceptable practice for the management of FIGO classification Stage la(i) tumours. In an attempt to develop a more conservative operation for early invasive carcinoma of the cervix, Dargent et al.2 described a new technique suitable for exophytic tumours of Stages la to 2a which allowed preservation of the uterus but removed the cervix, parametrium and upper one third of the vagina. He called this procedure ‘radical trachelectomy’. His patients also underwent laparoscopic pelvic lymphadenectomy, with negative histology results. The description of the operation includes visualisation of the ureters and ligation of the descending cervical and vaginal branches of the uterine artery. This is a technically demanding operation requiring skill as both a laparoscopic and vaginal surgeon.

We wondered whether an abdominal approach to radical trachelectomy might be possible and initially performed cadaveric dissection to investigate this possibility.

We consulted the literature with respect to the spread of cervical carcinoma. Burghardt and Girardi3 state that “the spread of cervical cancer is predominantly lateral i.e. into the parametrial tissue”. Spread may be continuous whereby the tumour infiltrates in a confluent manner towards the pelvic sidewall, or discontinuous when there may be vessel and parametrial lymph node involvement. In Burghardt and Girardi's series of 395 women, 149 had Stage 1b carcinoma and of these 16.1% had parametrial involvement, while in women with Stage 2a tumours (n= 8), 25% had parametrial involvement.

Vertical spread of cervical cancer is less common than lateral spread. Little information is available on this subject as it is normally a late feature of the disease. In the series of Burghardt and Girardi there was no case of vertical spread in Stage 1b or 2a cancers. In 220 Stage 2b tumours there were 11 cases (20%) of spread to the uterine corpus. Other investigators quote incidences of 26%4 and 24%5. Baltzer6 found age to be an important variable in the vertical spread of disease in women with Stage 2b tumours, with 9.5% of women younger than the age of 50 years but 32% older than 50 years having uterine involvement.

The uterus is supplied by two uterine arteries, two ovarian arteries via collaterals and two vaginal arteries via collaterals. At the Society of Gynecologic Oncologists (SGO) Conference 1996 (New Orleans, Louisiana, USA) in an interactive session between the panel and members, where the membership were able to vote on aspects of clinical management, discussion took place as to uterine viability with ligation of various vessels. There was agreement among the membership that in the presence of vaginal collaterals both uterine arteries could be ligated but not both uterines and one ovarian artery (i.e. of the six arteries, the membership thought that at least four should be present in the absence of uterine arteries, although there is no controlled study to demonstrate this in practice). A trachelectomy requires the division of the vaginal arteries in order to remove the cervix and upper vagina. To perform a radical trachelectomy via the abdominal approach requires division of the uterine arteries in order to achieve adequate ureteric dissection as it is not technically possible simply to dissect the uterine arteries from their origins at the internal iliac arteries and hold them out of the operating field. An abdominal radical trachelectomy probably requires that the uterine arteries are re-anastomosed as the uterus would otherwise receive arterial supply from the ovarian arteries only.

We propose a new technique which gives the same pelvic lymph node and parametrial clearance as a radical hysterectomy while preserving the uterus and thus

the capacity to have children (Fig. 1). We write this article to invite comment and stimulate debate among gynaecologic oncologists. The discussion section incorporates some of the arguments for and against this technique as well as other methods of radical trachelectomy.

image

Figure 1. Cervix, upper third of the vagina and parametrium are all removed. The uterine arteries are divided.

Download figure to PowerPoint

Surgical technique

  1. Top of page
  2. Abstract
  3. Surgical technique
  4. Discussion
  5. Conclusion
  6. References

The abdomen is opened with a vertical subumbilical incision. The round ligaments are divided and the broad ligaments entered. The external and internal iliac and obturator fossae lymph nodes are removed and examined histologically by frozen section. The ureters are dissected as in a standard radical hysterectomy until the uterine arteries are revealed. The uterine artery is dissected from the body of the uterus to produce a skeletonised vessel, and is clipped and divided 4 cm from its origin at the internal iliac artery. Bulldog clips are applied, and heparinised saline instilled into the ends of the vessel in both directions. The ureteric tunnels are opened in standard fashion. The uterosacral ligaments and the parametrium are divided. The uterus remains connected by the ovarian pedicles. The vagina is opened and the uterus, cervix, vagina and parametrium are swung superiorly and the cervix with attached parametrium and vagina is amputated (Figs 2 and 3). The superior surface of the amputated cervix is examined histologically by frozen section to ensure that there is a tumour-free margin. If a reasonable margin exists, a cervical cerclage suture is then inserted into the remaining cervical stump to maintain cervical competence in any future pregnancy.

The two ends of the ligated uterine artery are now re-anastomosed. First the ends are trimmed. They are then joined using sutures. The vault of the vagina is sutured using an interlocking circumferential suture to give adequate haemostasis and the body of the uterus sutured to the vault with a modified Sturmdorf suture. The abdomen is then closed.

So far, this procedure has not been performed in its entirety. Initially, the technique was performed as far as the stage of re-anastomosing the uterine vessels in a woman who was undergoing a radical hysterectomy. The re-anastomosis was then performed ex vivo on the basis that prolongation of anaesthetic time could not be justified in an experimental situation. The anastomosis in this case was performed using the 3M Precise Microvascular Anastomotic Device—a system designed for anastomosis of small vessels using two interlocking plastic rings with pins placed circumferentially over which the vessel ends are placed (Figs 4 and 5). We demonstrated, by removing the clip temporarily, that the uterine artery remained patent at the point where it was divided from the internal iliac artery. We also demonstrated by angiography that the anastomosis was patent and did not leak (Fig. 6).

image

Figure 6. Angiography of the reanastomosed uterine artery to demonstrate both patency and lack of leakage.

Download figure to PowerPoint

The technique was then performed on a woman with a Stage Ib tumour who gave her fully informed consent and who wished to preserve her fertility. In this case, the re-anastomosis was performed using the 3M Precise Microvascular Anastomotic Device but unfortunately broke down soon after. This was due to misalignment of the vessel ends. Anaesthetic complications prevented a further attempt at re-anastomosis. Interestingly, this patient's uterus remained viable thereafter supplied only by the ovarian arteries. This was confirmed by Doppler flow studies post-operatively.

Discussion

  1. Top of page
  2. Abstract
  3. Surgical technique
  4. Discussion
  5. Conclusion
  6. References

Vascular considerations

The 3M Precise Microvascular Anastomotic Device has been tested extensively in animals and humans7–14. We ourselves originally tested it by re-anastomosing the inferior cava and aorta of a rat. Oguchi and Van der Lei13 performed a study re-anastomosing rat femoral veins and showed a 92% patency rate immediately and of these, all remained patent for the six week duration of the study. In an animal experiment 4/25 arterial re-anastomoses resulted in aneurysm formation, and it was noted that arterial anastomosis was technically more difficult than venous anastomosis14. The main concern regarding anastomosed uterine arteries is the risk of separation of the vessel at the site of the anastomosis during pregnancy. A further concern is that the anastomosis will constrict the vessel at that point, and we do not know what effect this will have on blood flow during pregnancy. This is a particular concern where plastic nonabsorbable rings have been used as they will cause a fixed constriction in the uterine artery during pregnancy.

Our initial enthusiasm for the nonabsorbable rings was the apparent ease with which they could be applied. However, our concerns expressed above now lead us to believe that the use of sutures is preferable. In uterine arteries which have been anastomosed with sutures, the scar tissue at the anastomatic site is likely to stretch, at least to some extent, when the vessel starts to expand in pregnancy.

We intend to test the ability of the anastomosed uterine arteries to withstand the greatly increased flow during pregnancy in an animal model. We will use the pig as a model as it is anatomically the species closest to humans with regard to pelvic anatomy. We intend to investigate whether anastomosis is required at all and, if so, which suture is preferable. Six sows will undergo laparotomy and have their uterine vessels dissected. Some uterine arteries will remain untouched and act as the controls. The others will be divided and then anastomosed using either 6/0 absorbable PDS suture or 6/0 nonabsorbable Prolene. The sows will then be inseminated and undergo fortnightly Doppler flow studies of their uterine arteries during pregnancy. All sows will be delivered by caesarean section and the uterine arteries will be observed during the procedure.

We have demonstrated the possibility of anastomosis of the uterine arteries. At the end of the procedure the uterus will be supplied by two uterine and two ovarian arteries and drained by two ovarian veins. Arterial anastomosis may also be suitable for repair of damaged vessels during other gynaecological operations which would otherwise necessitate removal of an organ, for example the ovary. We suggest a microvascular surgeon carries out the arterial anastomoses.

Oncological considerations

The operation is conceptually similar to radical abdominal hysterectomy. This contrasts with vaginal trach-electomy which requires advanced laparoscopic skills as well as ureteric dissection carried out through the vagina. Abdominal trachelectomy allows good clearance of the tumour as well as the lymph nodes and parametrium.

Criticisms of the technique include a possible increased risk of complications during and after surgery. The operation takes longer than a conventional radical abdominal hysterectomy. In addition if the anastomosis fails and there is no blood flow through it then hysterectomy might be required. This risk is possibly higher than with vaginal trachelectomy where the uterine arteries are untouched.

Ethical considerations

Approval from the Ethics Committee of Chelsea and Westminster Hospital was granted for the case of the woman who then went on to have a radical hysterectomy. Ethics Committee approval was not deemed to be required for the further patient who had the procedure performed, nor for the initial cadaveric work. The full informed consent of both patients was also obtained. A project license with the Home Office is currently pending approval for the sows to undergo laparotomy.

Conclusion

  1. Top of page
  2. Abstract
  3. Surgical technique
  4. Discussion
  5. Conclusion
  6. References

Abdominal radical trachelectomy with uterine artery anastomosis may be suitable for women with early invasive carcinoma of the cervix who strongly desire to have children. Our next step is to complete our animal experiments and to perform the operation in women with early invasive carcinoma of the cervix.

References

  1. Top of page
  2. Abstract
  3. Surgical technique
  4. Discussion
  5. Conclusion
  6. References
  • 1
    CRC. Cervical screening Factsheet 13.1. Cancer Research Campaign, 1994.
  • 2
    Dargent D, Brun JL, Roy M et al. La trachelectomie elargie (TE) une alternative à l'hysteréctomie radicale dans le traitement des cancers infiltrants developés sur la face externe du col uterin [in French]. Journal d'Obstetrique et Gynecologic 1994; 2: 285292.
  • 3
    Burghardt E, Girardi F. Cervical cancer; local spread. In: BurghardtE, editor. Gynecologic Surgical Oncology. Stuttgart : Georg Thieme Verlag, 1993: 203212.
  • 4
    Mitani Y, Jimi S, Iwasaki H. Carcinomatous infiltration into the uterine body in carcinoma of the uterine cervix. Am J Obstet Gynecol 1964; 89: 984.
  • 5
    Ferraris G, Lanza A, Re A et al. The significance of lymph node status at pelvic, common iliac and para-aortic levels. Baillieres Clin Obstet Gynaecol 1988; 2: 913920.
  • 6
    Baltzer J. Die operative Behandlung des Zervixkarzinoms. Klinische, histologische und tumormetrische. Untersuchungserbnisse einer kooperativen Studie an vier Universitatsfrauenkliniken bei 1092 Patientinnen mit Zervixkarzinom [dissertation]. Munich : University of Munich; 1978.
  • 7
    Berggren A, Ostrup LT, Lidman D. Mechanical anastomosis of small arteries and veins with the Unilink apparatus: A histologic and scanning electron microscopic study. Plast Reconstr Surg 1987; 80: 274283.
  • 8
    Ragnarsson R, Berggren A, Ostrup LT, Franzen L. microvascular anastomosis of interpositional vein grafts with the Unilink system: a comparative experimental study. Scan J Plast Reconstr Surg 1989; 23: 2328.
  • 9
    Gilbert RW, Ragnarsson R, Berggren A, Ostrup L. Strength of microvascular anastomoses: comparison between the Unilink anasto-motic system and sutures. Microsurgery 1989; 10: 4046.
  • 10
    Blair WF, Steyers CM, Brown TD, Gable RH. A microvascular anastomotic device: Part 1. A haemodynamic evaluation in rabbit femoral arteries and veins. Microsurgery 1989; 10: 2128.
  • 11
    Blair WF, Morecraft RJ, Steyers CM, Maynard JA. A microvascular anastomotic device: Part 2. A histologic study in arteries and veins. Microsurgery 1989; 10: 2939.
  • 12
    Falconer DP, Lewis TW, Lamprecht EG, Mendenhall HV. Evaluation of a microvascular anastomotic device in the dog. Presented at Royal Society of Medicine; Sep 1988.
  • 13
    Oguchi H, van der Lei B. The 3M precise microvascular anastomotic system for implanting PTFE microvenous prostheses into the rat femoral vein. Plast Reconstr Surg 1996; 97: 662665.
  • 14
    Borja-Morant A, Navarro-Vila C, Cuesta-Gil M, Martin-Sastre R. Experimental evaluation and clinical use in the head and neck of the 3 M Precise microvascular anastomotic device. J Craniomaxillofacial Surg 1995; 23: 305311.