INT. J. HYPERTHERMIA,
1991,
VOL.
7,
NO.
2, 213-220
WINNER OF THE 1990 LUND AWARD
Treatment of menorrhagia by radiofrequency heating M. V. PRIOR?, J. H. PHIPPSS, T. ROBERTS?, B. V. LEWIS$, J. W. HAND? and S . B. FIELD? tMRC Cyclotron Unit, Hammersmith Hospital, Ducane Road, London W12 OHS, UK $Department of Obstetrics and Gynaecology, Watford General Hospital, Vicarage Road, Watford WD1 8HB, Herts, UK
Int J Hyperthermia Downloaded from informahealthcare.com by University of Melbourne on 11/06/14 For personal use only.
(Received 12 February 1990; revision received 25 April 1990; accepted 5 July 1990)
A new technique is described for the treatment of menorrhagia by heating the whole of the endometrial cavity of the uterus. A capacitively coupled probe at 27 ‘ 12 MHz is inserted into the uterine cavity, which causes the basilis layer to be raised to approximately 50-55°C whilst the rest of the pelvic contents remain at approximately normal (body) temperature. A major advantage of the method is that no special hysteroscopic skills are required, unlike the two other techniques currently used for endometrial ablation: the Nd-Yag laser or the hysteroresectoscopicloop. Also no toxic flushing/distension fluids are necessary, as are required for all hysteroscopic surgery. The method, however, does require the application of a large amount of RF power to the probe and so care must be taken to position the probe correctly in order to prevent any serious complications. Of 32 patients given a single treatment at a power level of 550 W for 20 min, the ‘success rate’ was 84% with 31 % becoming amenhorrhoiec and 53 % showing significant reduction in menstrual bleeding. However, retreatment is possible and by this means, combined with improved treatment techniques, an even higher success rate could be achieved. In two of the earlier patients treated at 550 W a fistula was produced at the anterior vaginal wall which had to be surgically repaired. The probe was subsequently modified, since when this problem has not recurred. This new treatment approach offers an alternative to hysterectomy in the treatment of menorrhagia and may offer a number of significant advantages over methods currently used for endometrial ablation. Key words: RF heating, menorrhagia, endometrial ablation.
1. Introduction The potential for hyperthennia in the treatment of malignant disease is being investigated in centres throughout the world (Sugahara and Saito 1989). In addition it has also been reported that hyperthermia may be useful in the treatment of the non-malignant diseases benign prostatic hyperplasia (Yerushalmi et al. 1985, Astrahan ef at. 1989) and psoriasis (Orenberg et al. 1986). In the present paper the use of hyperthermia in the treatment of another non-malignant condition, namely menorrhagia, is presented. Menorrhagia is defined as excessive menstrual bleeding in the absence of organic pathology. It has no known aetiology and is presumed to be due to an inappropriate exposure of the endometrium to hormones, particularly oestrogen followed by progestogen. It should also be noted that it is a largely subjective condition. It is an exceedingly common problem, typically comprising approximately one in five outpatient referrals to gynaecological departments. Women suffering severe menorrhagia are at risk from chronic anaemia. They are often forced to develop a lifestyle which is dictated by the timing of their periods, making a normal social and sexual life impossible. First-line treatment is often by drug therapy with either progestogens such as norethisterone, anti-oestrogens such as danazol , o r anti-fibrinolytics such as cyclokapron. A major disadvantage is the need to administer drugs long-term, and in addition beneficial effects are likely to be temporary. Menorrhagia 0256-6736/91 $3.00 01991 Taylor & Francis Ltd
Int J Hyperthermia Downloaded from informahealthcare.com by University of Melbourne on 11/06/14 For personal use only.
2 14
M. I/. Prior et al.
is effectively dealt with by hysterectomy, but such major surgery carries the disadvantages o f a significant probability of associated mortality and morbidity. The mortality associated with hirnple hysterectomy is widely reported as between 0.05% and 0.2% (Sharp and Jordan i987). Serious morbidity may result from the complications of any major surgery, but in tlie case of hysterectomy there is a special risk to the urinary tract, particularly the bladder and ureters where damage may mean further major surgery to correct the injury. A number of physical and chemical methods have been tried as alternatives to h;ystercctomy, including the use of superheated steam, cryotherapy, urea injection, ‘superglue’, and radium packing (Goldrath et al. 1981). Of these cryotherapy met with limited success (Droegemueller et al. 1971) but was abandoned due to complications. Radiiiin packing was also found to be useful, but is no longer used due to the risk of carcinogenesis. The most commonly used methods as an alternative to hysterectomy are ablation of the endometrium either by using a neodymium-yttrium-aluminium-garnet (Nd-k’ag) laser (Davis 1987), or the diathermy hysteroresectoscope loop (Hamou 1988). These lechniques, in skilled hands (as they require to be performed under direct h:ysterorccopic vision), result in amenorrhoea in approximately 30-50% of patients with a substmtial proportion of the remainder hypomenorrhoiec (Lomano 1988) resulting in an ovrrall ‘success rate’ of approximately 85 % . Follow-up times for these patients vary from study to study but range from 6 months to 1 year. Longer-term follow-up data for these er,dometrial ablation studies are at present lacking. However, in order to maintain clear vision toxic fluid must be flushed through the cavity. This gives rise to a risk of cr:rebra!l and pulmonary oedema due to absorption of fluid into the circulation. The actual magnitu.de of this risk cannot at present be accurately established as, to date, inadequate numbers of patients have been studied. A further serious potential risk of these techniques is the possibility of perforation of the uterine wall. The present report describes a different approach to destroying the endometrium. There h,ave been many previous reports in the literature of using R F probes for heating body cavities (Schliephake 1935, Daels 1973, Hand et al. 1982, Oliguchi and Tsutsumi 1989). A modification of these techniques has been used to heat the superficial layers of the uterus. A capacitively coupled probe driven at 27- 12 MHz is inserted, under general anaesthesia, into the uterus, which allows the whole of the endometrium to be heated simultaneously. I t is Iesa. time-consuming and requires considerably less surgical skill than either the laser or diathermy loop methods. There is no requirement for the use of toxic distension/flushing media. Moreover, as the method relies on capacitive coupling between the probe and the tissue (rather than direct galvanic contact) the irregularity of the endometrial cavity does not rcstrict heating to only those points that are in intimate contact with the probe. Results of preliminary in vitro experiments, in vivo feasibility studies and a phase 1/11 trial o f the technique are reported.
2. Materials and methods 2 . 1 . Flmting system ‘The method chosen to heat the endometrium of the uterus was to deliver power at 2’7.12, MHz to a probe placed into the uterine cavity and capacitively coupled to a second electrcde consisting of a belt placed around the patient. A diagram of the probe is shown in Figure 1. The ‘active’ region, which consists of ;t 10 nini diameter stainless-steel cylinder 70 mm in length, is placed into the uterine cavity. A n elcc?rical connection is then made between the stainless-steel cylinder and a miniature BNC connector at the end of probe. This is done via a stainless-steel rod (1 mm in diameterj in an insulated support, machined from nylon. ’The belt. which forms the return electrode of the capacitive network, consists of a
215
.t
Int J Hyperthermia Downloaded from informahealthcare.com by University of Melbourne on 11/06/14 For personal use only.
.~ .................................................
~
......
~
~~~
wire mesh embedded in insulating material (to avoid galvanic contact between the belt and the patient). It is of sufficient length to be wrapped around the patients’ hips, so that it can be in the same plane as the active part of the probe. The heating system consists of a 27.12 MHz signal source (Kenwood Transceiver model TS-520SE) feeding a linear amplifier (Kenwood model TL922). The output from the linear amplifier is then passed through a power/SWR meter (Hansen model FS-301) into an antenna tuning unit (Heathkit model SA-2040), the output from which is connected to the probe and belt. 2.2. Thermometry All thermometric studies were carried out using a Luxtron 3000 fibreoptic thermometry system (Luxtron Corporation, Mountain View, CA). This system allows up to eight channels of temperature data to be monitored in the presence of electromagnetic fields, so that temperature readings are possible whilst the power is on. The system is capable of reading temperatures either from probes consisting of four individual sensors or from a probe consisting of a linear array of sensors at various spacings. In this study either probes with single sensors or a linear array with 1 cm sensor spacing were used.
3. Results 3.1 In vitro studies The initial investigation into the feasibility of heating uterine tissue with 27.12 MHz capacitively coupled RF, was carried out on five freshly excised human uteri, with thermometry probes placed at various depths radially out from the probe. The excised uterus was first dilated and the probe placed into the cavity with an insulated metallic sheet wrapped around the uterus to act as the return electrode. In all the in vitro studies it was found that the antenna tuning unit was capable of matching the load into the 50 fl output of the linear amplifier and it was possible to obtain an SWR of 1. Figure 2 illustrates the temperature rise as a function of time at various distances in the uterus radially out from the probe surface using an incident power of 70 W. Similar results were obtained in each of the five studies. The figure shows clearly that a significant rise in temperature occurs at the probe surface and falls off rapidly with increasing distance. This effect is due principally to the geometry of the system, the surface area of the probe being significantly less than that of the return electrode. The electric field intensity is therefore greatest at the probe surface and decreases with distance from the probe, heating being proportional to the square of the electric field strength. The differential between temperatures at various depths will decrease with time due to thermal conduction. In these
M . V. Prior et al.
2 I6
0 mrn 5 mm 10 mm 20 mrn
Int J Hyperthermia Downloaded from informahealthcare.com by University of Melbourne on 11/06/14 For personal use only.
1
0
100
200
Time (seconds) Figure 2. Temperature rise at various distances from the probe during heating in vitro.
itz vitro studies, there is of course no blood flow, which would modify the temperature distribution in vivo. The differences are expected to be particularly apparent at depths greater than about 10 mm where uterine blood flow becomes significant and would result in a further reduction in temperature. The results of the in vitro study indicated that the method was feasible. However, the effects of convection in vivo due to perfusion, coupled with the fact that the return electrode would be at a greater distance from the probe, lead to the requirement for more power, but also result in a more rapid fall off in temperature with distance from the probe.
3.2. I n vivo studies In order to assess the feasibility of applying the method safely in vivo, preliminary ir1vestig;itions were carried out on ten patients scheduled to have an abdominal hysterectomy. Luxtron temperature sensors were placed both into the uterus and at various other locations, e.g. bladder, liver and bowel, so that it was possible to monitor any significant temperature changes occurring within and outside the uterus. After placing the temperature sensors the belt was positioned around the patient and the probe energized. The first patient was treated at 275 W for 20 min. The power applied was then systematically increased on subsequent patients up to 550 W. The results from one such study on a patient who received 550 W for 20 min are shown in Figure 3. As can be seen in the figure, a significant temperature rise occurred at the endometrial surface (giving i~ steady-state temperature of 66°C) and at the basalis layer (steady-state temperature of 52 "C) but not at the uterine serosa or in bowel, bladder or abdomen. In all the patients in which this investigation was performed no significant heating was found to occur outside the uteriirs. Blood flow through the uterus is substantial and will certainly provide an important cooling effect. In order to assess whether the treatment caused any changes in the blood itself, samples were taken via the uterine vein from these patients during and after heating, and examined microscopically. No red cell deformity was observed. Since this endpoint has been shown to be a sensitive indicator of thermal damage to blood (Coakley et al. 1979), il: was concluded that no significant heating of the blood occurred.
RF treatment of menorrhagia
217
serosa
Int J Hyperthermia Downloaded from informahealthcare.com by University of Melbourne on 11/06/14 For personal use only.
layer (52 d e g C )
All other t i s s u e s at
normal temperature.
Figure 3. Diagram of the anatomy of the uterus showing the locations of the steady-state temperature measurements recorded in vivo during a treatment at 550 W for 20 min.
3.3. Clinical studies The results of the in vitro and in vivo studies showed that it should be possible, using this technique, to heat the endometrium of the uterus to cytotoxic temperatures without heating non-target tissue. On this basis a phase 1/11 trial was begun. All patients entered into this study were treated under general anaesthesia as day surgery cases. Eligibility for entry was judged on the following criteria: 1. A clinical history of menorrhagia. 2. Organic pathology had been excluded within the past 6 months by diagnostic curettage with or without hysteroscopy. 3. Age 35-50. 4. Patient in the postmenstrual phase of the cycle. 5. Patients had completed their families. (This was a requirement because of the uncertainty of the effects of the treatment on fertility.) 6 . The uterus was of clinically normal size and shape. 7. There was no serious intercurrent illness. 8 . Informed consent had been obtained. Patients were examined 1 week after treatment and at intervals of 6 weeks thereafter. The total follow-up at the time of submission of this manuscript is 6-8 months for all patients reported in Table 1. In this study menstrual blood loss was not measured. Menorrhagia is exceedingly difficult to assess objectively, although several attempts have been made in the past. Most Table 1. Response of patients treated in both phase I and I1 trials Treatment time (min) Number treated Amenorrhoiec Significant improvement No improvement Success rate
10 10 0 3 (30%) 7 (70%) 30 %
15 10
0 6 (60%) 4 (40%) 60 %
20 32 10 (31%) 17 (53%) 5 (16%) 84 %
The power used was 550 W in all cases, and follow-up times were a minimum of 6 months
Int J Hyperthermia Downloaded from informahealthcare.com by University of Melbourne on 11/06/14 For personal use only.
LL8
M . V . Prior et al.
of these methods, which on the whole involve saving used sanitary protection and measuring eil.her blc bod content or radioactivity after previous intravenous radioactive isotope injection, are extremely difficult to carry out and notoriously inaccurate. All studies published SO far on a n y form of endometrial ablation are based on patient history. It is noteworthy that most gynaecologists feel that if a patient feels her periods are excessive, they are therefore by definition excessive. The question as to whether patients are sterile after endometrial ablation has yet to hc anxwc.:red. Therefore in this study all patients were either sterile or advised to continue iising contraception after treatment. The rtirst phase I trial was performed on 10 patients to test the safety and efficacy o f treating ]for 10 min at a power level of 550 W. This was found to be safe and produced n o complications. At this power level 30% of the patients were rendered hypomenorrhoiec, the rcnrainder showing no improvement in their menstrual blood loss. A further 10 patients were tticn treated for 15 min at 550 W. The clinical results were improved, with 60% beconiin!; hypomenorrhoiec and the remaining 40% showing no improvement. There were no. significant side-effects. The next group of patients were treated at 550 W with a treatment time of 20 min. 'I'hirly-.rwo patients have been entered into this group so far. In the majority of cases a Ihur-senwr Luxtron probe was placed adjacent to the probe surface and the temperatures recordcd were in agreement with those measured at the probe surface in the in vivo studies, 1.c. a probe surface steady-state temperature of about 66"C, so that the basalis temperature was 50--:55"C.Of these 32 patients. 31 % have become amenorrhoiec, 53% have become hypomerrorrhoiec and the remaining 16% have shown no improvement. The .results for all the patients treated so far, for the various treatment times used, arc sumtnarized in Table 1 . In the table the 'success rate' is determined by combining the re:jults oafthe arnenorrhoiec patients with those who experienced a significant improvement. It represcnts the number of patients who no longer wish to be treated surgically after initially requesting hysterectomy. In all the patients premenopausal gonadotrophin levels were present hefore and after treatment. Following treatment, some patients experienced pain and discomfort for a limited period, the intmsity being greater in patients treated for 20 min. In some cases lower abdominal cmnip was experienced for up to 48 h post-treatment. This was satisfactorily dealt with using conventional analgesics. A proportion of patients complained of lower abdominal divcomfort and a blood-stained serous discharge lasting normally up to a week, but occasionally up to 4-6 weeks following treatment for 20 min. Clearly this technique leaves open the possibility of retreatment. Four of the patients who w(:ic treated in the first phase I study (i.e. at 550 W for 10 min) and who did not resporitl. were subsequently (at their own request) retreated at 550 W for 20 min. This second treatment was successful in three of the four patients. In tht:. early clinical tests it was noted that contact between the probe and the vaginal ~t-allCVLIS possible, giving rise to excess local heating, which unfortunately resulted in a tisttila rn two patients. Both of these patients were excessively obese (in excess of 100 kg). This rciiiltcd in bulging of the anterior vaginal wall and bladder into the operative field. hringirig it into close proximity with the shaft of the probe. Both of these patients presented d t c r 3 -4weeks post-treatment with urinary incontinence. Cystoscopy revealed a small cleftxt jtj!jt proximal to the trigone. These were both repaired trans-abdominally . Thc probe was subsequently modified by the addition of a nylon tubular speculum to prevent such contact. Since this change was made patients have been treated without such prohlc~ni.~.
Int J Hyperthermia Downloaded from informahealthcare.com by University of Melbourne on 11/06/14 For personal use only.
RF treatment of menorrhagia
219
4. Discussion This study indicates the potential for using RF heating in the treatment of a very common and unpleasant benign condition, i.e. menorrhagia (Phipps et al. 1990). The results of the in vitro and in vivo studies, combined with those from the phase 1/11 trial. show that by using this RF technique it is possible to treat this condition in a way that is much simpler, less time-consuming and probably safer, than other currently used methods. The in vitro studies showed that it was possible to capacitively couple the RF into a uterus. It also showed that the chosen geometry of the heating method led to a temperature distribution which was characterized by high temperature at the probe surface and a rapid fall-off with radial distance. This was a requirement of the heating system since the aim was to destroy selectively the basalis layer (approximately 5 mm from the endometrial surface). In vivo studies were then carried out to obtain information on the temperature distribution produced using this technique in patients. The results were satisfactory, with no significant heating of tissues outside the uterus. Also the tests on red cell deformity of blood samples taken before and after heating implied that there was no effective heating of the blood. The next stage in the evaluation of this technique was a phase 1/11 study. Table 1 shows that the percentage of patients who experience amenorrhoea or a significant reduction in flow increased with the treatment time used. In the patients who experienced reduced menstrual blood loss, many have had periods which have become successively lighter during their follow-up period. This phenomenon has also been reported with other endometrial ablation methods, and is thought to result from shrinkage of the cavity as a result of scarring. It is therefore possible that on continued follow-up a higher proportion of patients may respond. No significant side-effects were seen with treatments at either 10 or 15 min treatment times. All patients who were treated for 20 min developed some pain, which was successfully managed in all cases by the use of conventional analgesics, and blood-stained discharge after treatment, which was variable in degree. In the majority of cases these symptoms settled within I week, but sometimes lasted for up to 6 weeks. The greatest risk of damage is probably to the urinary bladder by accidental heating of the anterior vaginal wall. This resulted in the development of a fistula in two of the patients treated in the phase I1 part of the study. The probe was subsequently modified, since when this problem has not recurred. Retreatment of patients who fail to respond is possible. So far only patients who did not respond to the 10 min treatment have been assessed, but three out of four of these responded to a second treatment for 20 min. It is plausible that a proportion of failures resulted from technical difficulties. If so, then by retreating failures it should be possible to achieve an even higher success rate. Alternatively the treatment techniques might be improved. The results show that with the present design of treatment and equipment it is necessary to treat for approximately 20 min at 550 W for a high success rate to be achieved. However, the system used in this study was a prototype. Improved apparatus or a different treatment technique might yield even more impressive results. For example, by redesigning the output stage to deliver the power more efficiently we believe that a reduction in the absolute power used could be achieved without prejudicing the high success rate. This work is currently in progress. The exact mechanism by which this technique results in cell death is not known. However, the effect of heating the endometrium to the temperatures we achieved is likely to result in direct cell death with resulting necrosis of the endometrium.
Int J Hyperthermia Downloaded from informahealthcare.com by University of Melbourne on 11/06/14 For personal use only.
220
M. V. Prior et a].
5. Conclusions A new RF technique has been developed for the treatment of menorrhagia. It offers substantial advantages over the present methods used to treat this condition. It avoids the need for major surgery. Alternatively, by comparison with other outpatient procedures (such as use of the Nd-Yag laser or the resectoscope), it is far simpler to perform, requires no special surgical skills, is less time-consuming and possibly safer. The results of an initial phase 1/11 trial of this technique show it is possible to achieve a success rate of approxirnately 85 % with no serious side-effects (as long as care is taken to avoid contact with the anterior vaginal wall). In addition, should the first treatment fail, retreatment is possible with a high percentage of success. Retreatment of failures, combined with improvements in treatment techniques, ultimately might enable a higher success rate not far from 100% to be achieved. References ,&iTRAtiAN, M. A , , SAPOLINK, M. D., COHEN,D., LUXTON,G., KAMPP, T. D., BOYD,S. and PETROVICH, %. , 1989, Microwave applicator for transurethal hyperthermia of benign prostatic perplasia. Inrcwutional Journal of Hyperthermia, 5, 283-296. , W. T , , BAKER, A. J., CRUM,L. A. and DEELEY,J. 0. T., 1979, Morphological changes, molysis and microcirculation of heated human erythrocytes. Journal of7herrnul Biology, -1, 85-93, DAIL.S..I. 1973, Microwave heating of the uterine wall during parturition. Obstetrics and (;vnaecology, 42, 76-78. DAVIS,J.: 1987, The principles and use of the Nd-Yag laser in gynaecological surgery. Clinical 0hic~tric.xcind Gyaec-ology, 1, 33 1-50, ~ l ~ O t ~ ( ; ~ . ~ l ~ ~ W., t ~ . l GREER, - ~ ~ E R . B . and MAKOWSKI, E., 1971, Cryosurgery in patients with cly:-fiinctional uterine bleeding. Obstetrics and Gynaecology, 38, 256-8. (.;o~.I)KA,I H. M . H.. FULLER, T. A. and SEGAL,S., 1981, Laser photovaporization of endometrium t o r menorrhagia. American Journal of Obstetrics and Gynecology, 140, 14-19. H.4MOI:, J . , 1988, Resection of the endometrial cavity. Proceedings of the First International ('(wfirenc~,on Endoscopic,Surgery in Gynaecology, Antwerp, Belgium, December. HAND.J . W.. BLAKI:,, P . R.. HoPEWF,IJ., J. W.. LAMBERT, H. E. and FIELD,S. B., 1982. A COaxial applicator lor intracavitary hyperthermia of carcinoma of the cervix. Clinicd :'l7~,rt,iohir)logy,edited by M . Gautherie and E. Albert (Alan R. Liss, New York), pp. 635--639. LOMA'VO. J., 1988, Dragging technique versus blanching technique for endometrial ablation with the Nd-Yag laser in the treatment of chronic menorrhagia. American Journal of Obstetrics i l t l d C ; y t ~ c d ~ 159, g ~ , 152-1 55. Oi..I(ik!(.I4I.Y. and TsurS~Mr,S., 1989, RF capacitive heating by electrode inserted in stomach and oul:er one. Hyperrhennic Oncology 1988, Vol. 1, edited by M . Sugahara and M. Saito (Taylor Francis, L,ondon), pp. 804-806. C;? E. K., NOODLEMAN, F. R., KOPERSKI, J. A,, POUNDS, D. and FARBER, E. M.. 1986, C'ctrnparison of heat delivery systems for hyperthermic treatment of psoriasis. Internutional Jorcrnal of Hypertherrniu, 2, 23 1-241. Pmv?:. J~ H . , LEWIS,B. V., ROBERTS,T., PRIOR,M. V . , HAND,J. W., ELDER,M. and FIELD, 3 . B . , 1990, A new treatment for functional menorrhagia: radiofrequency induced thermal endometrial ablation. Luncet, 335, 374-376. S C H L . I ~ . PH A E.,K E1935. , Short Wuve Therapy (Actinic Press, London), pp. 53-54. SHARP, I-;. and JORDAN, J. A . , 1987, Gynaecological Laser Surgery (Perinatal Press, New York), p. 266. s l . i ( ; A t l k R A ~ M. and SAIIO, M. (eds), 1989, Hyperthermic Oncology 1988, Vols. 1 and 2 (Taylor & Francis, London). YEKI!SI-iII.l..MI, A., FISHEL.OVITZ, Y., SINGER,D., REINER,I., ARIELLY,J . , ABRAMOVICJ. Y.. ( ' ~ L r S E N ~ l . s o NR.. , LEVY,E. and SHANI,A,, 1985, Localized deep microwave hyperthermia i n the treatment of poor operative risk patients with benign prostatic hyperplasia. Journal of Ilrolo,qy, 133, 873-876.
1991,
VOL.
7,
NO.
2, 213-220
WINNER OF THE 1990 LUND AWARD
Treatment of menorrhagia by radiofrequency heating M. V. PRIOR?, J. H. PHIPPSS, T. ROBERTS?, B. V. LEWIS$, J. W. HAND? and S . B. FIELD? tMRC Cyclotron Unit, Hammersmith Hospital, Ducane Road, London W12 OHS, UK $Department of Obstetrics and Gynaecology, Watford General Hospital, Vicarage Road, Watford WD1 8HB, Herts, UK
Int J Hyperthermia Downloaded from informahealthcare.com by University of Melbourne on 11/06/14 For personal use only.
(Received 12 February 1990; revision received 25 April 1990; accepted 5 July 1990)
A new technique is described for the treatment of menorrhagia by heating the whole of the endometrial cavity of the uterus. A capacitively coupled probe at 27 ‘ 12 MHz is inserted into the uterine cavity, which causes the basilis layer to be raised to approximately 50-55°C whilst the rest of the pelvic contents remain at approximately normal (body) temperature. A major advantage of the method is that no special hysteroscopic skills are required, unlike the two other techniques currently used for endometrial ablation: the Nd-Yag laser or the hysteroresectoscopicloop. Also no toxic flushing/distension fluids are necessary, as are required for all hysteroscopic surgery. The method, however, does require the application of a large amount of RF power to the probe and so care must be taken to position the probe correctly in order to prevent any serious complications. Of 32 patients given a single treatment at a power level of 550 W for 20 min, the ‘success rate’ was 84% with 31 % becoming amenhorrhoiec and 53 % showing significant reduction in menstrual bleeding. However, retreatment is possible and by this means, combined with improved treatment techniques, an even higher success rate could be achieved. In two of the earlier patients treated at 550 W a fistula was produced at the anterior vaginal wall which had to be surgically repaired. The probe was subsequently modified, since when this problem has not recurred. This new treatment approach offers an alternative to hysterectomy in the treatment of menorrhagia and may offer a number of significant advantages over methods currently used for endometrial ablation. Key words: RF heating, menorrhagia, endometrial ablation.
1. Introduction The potential for hyperthennia in the treatment of malignant disease is being investigated in centres throughout the world (Sugahara and Saito 1989). In addition it has also been reported that hyperthermia may be useful in the treatment of the non-malignant diseases benign prostatic hyperplasia (Yerushalmi et al. 1985, Astrahan ef at. 1989) and psoriasis (Orenberg et al. 1986). In the present paper the use of hyperthermia in the treatment of another non-malignant condition, namely menorrhagia, is presented. Menorrhagia is defined as excessive menstrual bleeding in the absence of organic pathology. It has no known aetiology and is presumed to be due to an inappropriate exposure of the endometrium to hormones, particularly oestrogen followed by progestogen. It should also be noted that it is a largely subjective condition. It is an exceedingly common problem, typically comprising approximately one in five outpatient referrals to gynaecological departments. Women suffering severe menorrhagia are at risk from chronic anaemia. They are often forced to develop a lifestyle which is dictated by the timing of their periods, making a normal social and sexual life impossible. First-line treatment is often by drug therapy with either progestogens such as norethisterone, anti-oestrogens such as danazol , o r anti-fibrinolytics such as cyclokapron. A major disadvantage is the need to administer drugs long-term, and in addition beneficial effects are likely to be temporary. Menorrhagia 0256-6736/91 $3.00 01991 Taylor & Francis Ltd
Int J Hyperthermia Downloaded from informahealthcare.com by University of Melbourne on 11/06/14 For personal use only.
2 14
M. I/. Prior et al.
is effectively dealt with by hysterectomy, but such major surgery carries the disadvantages o f a significant probability of associated mortality and morbidity. The mortality associated with hirnple hysterectomy is widely reported as between 0.05% and 0.2% (Sharp and Jordan i987). Serious morbidity may result from the complications of any major surgery, but in tlie case of hysterectomy there is a special risk to the urinary tract, particularly the bladder and ureters where damage may mean further major surgery to correct the injury. A number of physical and chemical methods have been tried as alternatives to h;ystercctomy, including the use of superheated steam, cryotherapy, urea injection, ‘superglue’, and radium packing (Goldrath et al. 1981). Of these cryotherapy met with limited success (Droegemueller et al. 1971) but was abandoned due to complications. Radiiiin packing was also found to be useful, but is no longer used due to the risk of carcinogenesis. The most commonly used methods as an alternative to hysterectomy are ablation of the endometrium either by using a neodymium-yttrium-aluminium-garnet (Nd-k’ag) laser (Davis 1987), or the diathermy hysteroresectoscope loop (Hamou 1988). These lechniques, in skilled hands (as they require to be performed under direct h:ysterorccopic vision), result in amenorrhoea in approximately 30-50% of patients with a substmtial proportion of the remainder hypomenorrhoiec (Lomano 1988) resulting in an ovrrall ‘success rate’ of approximately 85 % . Follow-up times for these patients vary from study to study but range from 6 months to 1 year. Longer-term follow-up data for these er,dometrial ablation studies are at present lacking. However, in order to maintain clear vision toxic fluid must be flushed through the cavity. This gives rise to a risk of cr:rebra!l and pulmonary oedema due to absorption of fluid into the circulation. The actual magnitu.de of this risk cannot at present be accurately established as, to date, inadequate numbers of patients have been studied. A further serious potential risk of these techniques is the possibility of perforation of the uterine wall. The present report describes a different approach to destroying the endometrium. There h,ave been many previous reports in the literature of using R F probes for heating body cavities (Schliephake 1935, Daels 1973, Hand et al. 1982, Oliguchi and Tsutsumi 1989). A modification of these techniques has been used to heat the superficial layers of the uterus. A capacitively coupled probe driven at 27- 12 MHz is inserted, under general anaesthesia, into the uterus, which allows the whole of the endometrium to be heated simultaneously. I t is Iesa. time-consuming and requires considerably less surgical skill than either the laser or diathermy loop methods. There is no requirement for the use of toxic distension/flushing media. Moreover, as the method relies on capacitive coupling between the probe and the tissue (rather than direct galvanic contact) the irregularity of the endometrial cavity does not rcstrict heating to only those points that are in intimate contact with the probe. Results of preliminary in vitro experiments, in vivo feasibility studies and a phase 1/11 trial o f the technique are reported.
2. Materials and methods 2 . 1 . Flmting system ‘The method chosen to heat the endometrium of the uterus was to deliver power at 2’7.12, MHz to a probe placed into the uterine cavity and capacitively coupled to a second electrcde consisting of a belt placed around the patient. A diagram of the probe is shown in Figure 1. The ‘active’ region, which consists of ;t 10 nini diameter stainless-steel cylinder 70 mm in length, is placed into the uterine cavity. A n elcc?rical connection is then made between the stainless-steel cylinder and a miniature BNC connector at the end of probe. This is done via a stainless-steel rod (1 mm in diameterj in an insulated support, machined from nylon. ’The belt. which forms the return electrode of the capacitive network, consists of a
215
.t
Int J Hyperthermia Downloaded from informahealthcare.com by University of Melbourne on 11/06/14 For personal use only.
.~ .................................................
~
......
~
~~~
wire mesh embedded in insulating material (to avoid galvanic contact between the belt and the patient). It is of sufficient length to be wrapped around the patients’ hips, so that it can be in the same plane as the active part of the probe. The heating system consists of a 27.12 MHz signal source (Kenwood Transceiver model TS-520SE) feeding a linear amplifier (Kenwood model TL922). The output from the linear amplifier is then passed through a power/SWR meter (Hansen model FS-301) into an antenna tuning unit (Heathkit model SA-2040), the output from which is connected to the probe and belt. 2.2. Thermometry All thermometric studies were carried out using a Luxtron 3000 fibreoptic thermometry system (Luxtron Corporation, Mountain View, CA). This system allows up to eight channels of temperature data to be monitored in the presence of electromagnetic fields, so that temperature readings are possible whilst the power is on. The system is capable of reading temperatures either from probes consisting of four individual sensors or from a probe consisting of a linear array of sensors at various spacings. In this study either probes with single sensors or a linear array with 1 cm sensor spacing were used.
3. Results 3.1 In vitro studies The initial investigation into the feasibility of heating uterine tissue with 27.12 MHz capacitively coupled RF, was carried out on five freshly excised human uteri, with thermometry probes placed at various depths radially out from the probe. The excised uterus was first dilated and the probe placed into the cavity with an insulated metallic sheet wrapped around the uterus to act as the return electrode. In all the in vitro studies it was found that the antenna tuning unit was capable of matching the load into the 50 fl output of the linear amplifier and it was possible to obtain an SWR of 1. Figure 2 illustrates the temperature rise as a function of time at various distances in the uterus radially out from the probe surface using an incident power of 70 W. Similar results were obtained in each of the five studies. The figure shows clearly that a significant rise in temperature occurs at the probe surface and falls off rapidly with increasing distance. This effect is due principally to the geometry of the system, the surface area of the probe being significantly less than that of the return electrode. The electric field intensity is therefore greatest at the probe surface and decreases with distance from the probe, heating being proportional to the square of the electric field strength. The differential between temperatures at various depths will decrease with time due to thermal conduction. In these
M . V. Prior et al.
2 I6
0 mrn 5 mm 10 mm 20 mrn
Int J Hyperthermia Downloaded from informahealthcare.com by University of Melbourne on 11/06/14 For personal use only.
1
0
100
200
Time (seconds) Figure 2. Temperature rise at various distances from the probe during heating in vitro.
itz vitro studies, there is of course no blood flow, which would modify the temperature distribution in vivo. The differences are expected to be particularly apparent at depths greater than about 10 mm where uterine blood flow becomes significant and would result in a further reduction in temperature. The results of the in vitro study indicated that the method was feasible. However, the effects of convection in vivo due to perfusion, coupled with the fact that the return electrode would be at a greater distance from the probe, lead to the requirement for more power, but also result in a more rapid fall off in temperature with distance from the probe.
3.2. I n vivo studies In order to assess the feasibility of applying the method safely in vivo, preliminary ir1vestig;itions were carried out on ten patients scheduled to have an abdominal hysterectomy. Luxtron temperature sensors were placed both into the uterus and at various other locations, e.g. bladder, liver and bowel, so that it was possible to monitor any significant temperature changes occurring within and outside the uterus. After placing the temperature sensors the belt was positioned around the patient and the probe energized. The first patient was treated at 275 W for 20 min. The power applied was then systematically increased on subsequent patients up to 550 W. The results from one such study on a patient who received 550 W for 20 min are shown in Figure 3. As can be seen in the figure, a significant temperature rise occurred at the endometrial surface (giving i~ steady-state temperature of 66°C) and at the basalis layer (steady-state temperature of 52 "C) but not at the uterine serosa or in bowel, bladder or abdomen. In all the patients in which this investigation was performed no significant heating was found to occur outside the uteriirs. Blood flow through the uterus is substantial and will certainly provide an important cooling effect. In order to assess whether the treatment caused any changes in the blood itself, samples were taken via the uterine vein from these patients during and after heating, and examined microscopically. No red cell deformity was observed. Since this endpoint has been shown to be a sensitive indicator of thermal damage to blood (Coakley et al. 1979), il: was concluded that no significant heating of the blood occurred.
RF treatment of menorrhagia
217
serosa
Int J Hyperthermia Downloaded from informahealthcare.com by University of Melbourne on 11/06/14 For personal use only.
layer (52 d e g C )
All other t i s s u e s at
normal temperature.
Figure 3. Diagram of the anatomy of the uterus showing the locations of the steady-state temperature measurements recorded in vivo during a treatment at 550 W for 20 min.
3.3. Clinical studies The results of the in vitro and in vivo studies showed that it should be possible, using this technique, to heat the endometrium of the uterus to cytotoxic temperatures without heating non-target tissue. On this basis a phase 1/11 trial was begun. All patients entered into this study were treated under general anaesthesia as day surgery cases. Eligibility for entry was judged on the following criteria: 1. A clinical history of menorrhagia. 2. Organic pathology had been excluded within the past 6 months by diagnostic curettage with or without hysteroscopy. 3. Age 35-50. 4. Patient in the postmenstrual phase of the cycle. 5. Patients had completed their families. (This was a requirement because of the uncertainty of the effects of the treatment on fertility.) 6 . The uterus was of clinically normal size and shape. 7. There was no serious intercurrent illness. 8 . Informed consent had been obtained. Patients were examined 1 week after treatment and at intervals of 6 weeks thereafter. The total follow-up at the time of submission of this manuscript is 6-8 months for all patients reported in Table 1. In this study menstrual blood loss was not measured. Menorrhagia is exceedingly difficult to assess objectively, although several attempts have been made in the past. Most Table 1. Response of patients treated in both phase I and I1 trials Treatment time (min) Number treated Amenorrhoiec Significant improvement No improvement Success rate
10 10 0 3 (30%) 7 (70%) 30 %
15 10
0 6 (60%) 4 (40%) 60 %
20 32 10 (31%) 17 (53%) 5 (16%) 84 %
The power used was 550 W in all cases, and follow-up times were a minimum of 6 months
Int J Hyperthermia Downloaded from informahealthcare.com by University of Melbourne on 11/06/14 For personal use only.
LL8
M . V . Prior et al.
of these methods, which on the whole involve saving used sanitary protection and measuring eil.her blc bod content or radioactivity after previous intravenous radioactive isotope injection, are extremely difficult to carry out and notoriously inaccurate. All studies published SO far on a n y form of endometrial ablation are based on patient history. It is noteworthy that most gynaecologists feel that if a patient feels her periods are excessive, they are therefore by definition excessive. The question as to whether patients are sterile after endometrial ablation has yet to hc anxwc.:red. Therefore in this study all patients were either sterile or advised to continue iising contraception after treatment. The rtirst phase I trial was performed on 10 patients to test the safety and efficacy o f treating ]for 10 min at a power level of 550 W. This was found to be safe and produced n o complications. At this power level 30% of the patients were rendered hypomenorrhoiec, the rcnrainder showing no improvement in their menstrual blood loss. A further 10 patients were tticn treated for 15 min at 550 W. The clinical results were improved, with 60% beconiin!; hypomenorrhoiec and the remaining 40% showing no improvement. There were no. significant side-effects. The next group of patients were treated at 550 W with a treatment time of 20 min. 'I'hirly-.rwo patients have been entered into this group so far. In the majority of cases a Ihur-senwr Luxtron probe was placed adjacent to the probe surface and the temperatures recordcd were in agreement with those measured at the probe surface in the in vivo studies, 1.c. a probe surface steady-state temperature of about 66"C, so that the basalis temperature was 50--:55"C.Of these 32 patients. 31 % have become amenorrhoiec, 53% have become hypomerrorrhoiec and the remaining 16% have shown no improvement. The .results for all the patients treated so far, for the various treatment times used, arc sumtnarized in Table 1 . In the table the 'success rate' is determined by combining the re:jults oafthe arnenorrhoiec patients with those who experienced a significant improvement. It represcnts the number of patients who no longer wish to be treated surgically after initially requesting hysterectomy. In all the patients premenopausal gonadotrophin levels were present hefore and after treatment. Following treatment, some patients experienced pain and discomfort for a limited period, the intmsity being greater in patients treated for 20 min. In some cases lower abdominal cmnip was experienced for up to 48 h post-treatment. This was satisfactorily dealt with using conventional analgesics. A proportion of patients complained of lower abdominal divcomfort and a blood-stained serous discharge lasting normally up to a week, but occasionally up to 4-6 weeks following treatment for 20 min. Clearly this technique leaves open the possibility of retreatment. Four of the patients who w(:ic treated in the first phase I study (i.e. at 550 W for 10 min) and who did not resporitl. were subsequently (at their own request) retreated at 550 W for 20 min. This second treatment was successful in three of the four patients. In tht:. early clinical tests it was noted that contact between the probe and the vaginal ~t-allCVLIS possible, giving rise to excess local heating, which unfortunately resulted in a tisttila rn two patients. Both of these patients were excessively obese (in excess of 100 kg). This rciiiltcd in bulging of the anterior vaginal wall and bladder into the operative field. hringirig it into close proximity with the shaft of the probe. Both of these patients presented d t c r 3 -4weeks post-treatment with urinary incontinence. Cystoscopy revealed a small cleftxt jtj!jt proximal to the trigone. These were both repaired trans-abdominally . Thc probe was subsequently modified by the addition of a nylon tubular speculum to prevent such contact. Since this change was made patients have been treated without such prohlc~ni.~.
Int J Hyperthermia Downloaded from informahealthcare.com by University of Melbourne on 11/06/14 For personal use only.
RF treatment of menorrhagia
219
4. Discussion This study indicates the potential for using RF heating in the treatment of a very common and unpleasant benign condition, i.e. menorrhagia (Phipps et al. 1990). The results of the in vitro and in vivo studies, combined with those from the phase 1/11 trial. show that by using this RF technique it is possible to treat this condition in a way that is much simpler, less time-consuming and probably safer, than other currently used methods. The in vitro studies showed that it was possible to capacitively couple the RF into a uterus. It also showed that the chosen geometry of the heating method led to a temperature distribution which was characterized by high temperature at the probe surface and a rapid fall-off with radial distance. This was a requirement of the heating system since the aim was to destroy selectively the basalis layer (approximately 5 mm from the endometrial surface). In vivo studies were then carried out to obtain information on the temperature distribution produced using this technique in patients. The results were satisfactory, with no significant heating of tissues outside the uterus. Also the tests on red cell deformity of blood samples taken before and after heating implied that there was no effective heating of the blood. The next stage in the evaluation of this technique was a phase 1/11 study. Table 1 shows that the percentage of patients who experience amenorrhoea or a significant reduction in flow increased with the treatment time used. In the patients who experienced reduced menstrual blood loss, many have had periods which have become successively lighter during their follow-up period. This phenomenon has also been reported with other endometrial ablation methods, and is thought to result from shrinkage of the cavity as a result of scarring. It is therefore possible that on continued follow-up a higher proportion of patients may respond. No significant side-effects were seen with treatments at either 10 or 15 min treatment times. All patients who were treated for 20 min developed some pain, which was successfully managed in all cases by the use of conventional analgesics, and blood-stained discharge after treatment, which was variable in degree. In the majority of cases these symptoms settled within I week, but sometimes lasted for up to 6 weeks. The greatest risk of damage is probably to the urinary bladder by accidental heating of the anterior vaginal wall. This resulted in the development of a fistula in two of the patients treated in the phase I1 part of the study. The probe was subsequently modified, since when this problem has not recurred. Retreatment of patients who fail to respond is possible. So far only patients who did not respond to the 10 min treatment have been assessed, but three out of four of these responded to a second treatment for 20 min. It is plausible that a proportion of failures resulted from technical difficulties. If so, then by retreating failures it should be possible to achieve an even higher success rate. Alternatively the treatment techniques might be improved. The results show that with the present design of treatment and equipment it is necessary to treat for approximately 20 min at 550 W for a high success rate to be achieved. However, the system used in this study was a prototype. Improved apparatus or a different treatment technique might yield even more impressive results. For example, by redesigning the output stage to deliver the power more efficiently we believe that a reduction in the absolute power used could be achieved without prejudicing the high success rate. This work is currently in progress. The exact mechanism by which this technique results in cell death is not known. However, the effect of heating the endometrium to the temperatures we achieved is likely to result in direct cell death with resulting necrosis of the endometrium.
Int J Hyperthermia Downloaded from informahealthcare.com by University of Melbourne on 11/06/14 For personal use only.
220
M. V. Prior et a].
5. Conclusions A new RF technique has been developed for the treatment of menorrhagia. It offers substantial advantages over the present methods used to treat this condition. It avoids the need for major surgery. Alternatively, by comparison with other outpatient procedures (such as use of the Nd-Yag laser or the resectoscope), it is far simpler to perform, requires no special surgical skills, is less time-consuming and possibly safer. The results of an initial phase 1/11 trial of this technique show it is possible to achieve a success rate of approxirnately 85 % with no serious side-effects (as long as care is taken to avoid contact with the anterior vaginal wall). In addition, should the first treatment fail, retreatment is possible with a high percentage of success. Retreatment of failures, combined with improvements in treatment techniques, ultimately might enable a higher success rate not far from 100% to be achieved. References ,&iTRAtiAN, M. A , , SAPOLINK, M. D., COHEN,D., LUXTON,G., KAMPP, T. D., BOYD,S. and PETROVICH, %. , 1989, Microwave applicator for transurethal hyperthermia of benign prostatic perplasia. Inrcwutional Journal of Hyperthermia, 5, 283-296. , W. T , , BAKER, A. J., CRUM,L. A. and DEELEY,J. 0. T., 1979, Morphological changes, molysis and microcirculation of heated human erythrocytes. Journal of7herrnul Biology, -1, 85-93, DAIL.S..I. 1973, Microwave heating of the uterine wall during parturition. Obstetrics and (;vnaecology, 42, 76-78. DAVIS,J.: 1987, The principles and use of the Nd-Yag laser in gynaecological surgery. Clinical 0hic~tric.xcind Gyaec-ology, 1, 33 1-50, ~ l ~ O t ~ ( ; ~ . ~ l ~ ~ W., t ~ . l GREER, - ~ ~ E R . B . and MAKOWSKI, E., 1971, Cryosurgery in patients with cly:-fiinctional uterine bleeding. Obstetrics and Gynaecology, 38, 256-8. (.;o~.I)KA,I H. M . H.. FULLER, T. A. and SEGAL,S., 1981, Laser photovaporization of endometrium t o r menorrhagia. American Journal of Obstetrics and Gynecology, 140, 14-19. H.4MOI:, J . , 1988, Resection of the endometrial cavity. Proceedings of the First International ('(wfirenc~,on Endoscopic,Surgery in Gynaecology, Antwerp, Belgium, December. HAND.J . W.. BLAKI:,, P . R.. HoPEWF,IJ., J. W.. LAMBERT, H. E. and FIELD,S. B., 1982. A COaxial applicator lor intracavitary hyperthermia of carcinoma of the cervix. Clinicd :'l7~,rt,iohir)logy,edited by M . Gautherie and E. Albert (Alan R. Liss, New York), pp. 635--639. LOMA'VO. J., 1988, Dragging technique versus blanching technique for endometrial ablation with the Nd-Yag laser in the treatment of chronic menorrhagia. American Journal of Obstetrics i l t l d C ; y t ~ c d ~ 159, g ~ , 152-1 55. Oi..I(ik!(.I4I.Y. and TsurS~Mr,S., 1989, RF capacitive heating by electrode inserted in stomach and oul:er one. Hyperrhennic Oncology 1988, Vol. 1, edited by M . Sugahara and M. Saito (Taylor Francis, L,ondon), pp. 804-806. C;? E. K., NOODLEMAN, F. R., KOPERSKI, J. A,, POUNDS, D. and FARBER, E. M.. 1986, C'ctrnparison of heat delivery systems for hyperthermic treatment of psoriasis. Internutional Jorcrnal of Hypertherrniu, 2, 23 1-241. Pmv?:. J~ H . , LEWIS,B. V., ROBERTS,T., PRIOR,M. V . , HAND,J. W., ELDER,M. and FIELD, 3 . B . , 1990, A new treatment for functional menorrhagia: radiofrequency induced thermal endometrial ablation. Luncet, 335, 374-376. S C H L . I ~ . PH A E.,K E1935. , Short Wuve Therapy (Actinic Press, London), pp. 53-54. SHARP, I-;. and JORDAN, J. A . , 1987, Gynaecological Laser Surgery (Perinatal Press, New York), p. 266. s l . i ( ; A t l k R A ~ M. and SAIIO, M. (eds), 1989, Hyperthermic Oncology 1988, Vols. 1 and 2 (Taylor & Francis, London). YEKI!SI-iII.l..MI, A., FISHEL.OVITZ, Y., SINGER,D., REINER,I., ARIELLY,J . , ABRAMOVICJ. Y.. ( ' ~ L r S E N ~ l . s o NR.. , LEVY,E. and SHANI,A,, 1985, Localized deep microwave hyperthermia i n the treatment of poor operative risk patients with benign prostatic hyperplasia. Journal of Ilrolo,qy, 133, 873-876.
