Patienls who have had previous gynecologic surgery, Patients especially myomectomy, myomectomy, may have had one uterine ardetecrabl e tery ligated during the procedure. This is not detectable before arteriography and is usua ll y not mentioned mentio ned in the usually operative records. In these cases, UAE is unlikely to be clinically successful. Arteriographically, Arteriographi cally, oone ne may see reconstitution of uterine artery branches br.mches in the body of the uterus via collaterals. coll aterals. Anatomic Variants Anatom ic variations va riations of the uterine arteries have not been bee n Anatomic re, as there was no cl inical well-described in the literatu literature, clinical relevance for such anomalies until recently. We have seen several patterns of vascu lar anatomic variation va riation that vascular UAE, impact UAE. Complete or partial absence of oone ne uterine artery occurs in about 1% of the population. popu lation. In these cases, the suppl artery. It is easier supplyy to the uterus us via the ovarian artery. tect complete replacement rep lacement of the uterine supply to to de detect the ova rian artery than it is to detect partial partial replacement. ovarian Collate ral supply suppl y to the uterus from an enlarged ovarova rCollateral artery can also be seen, ian altery seen, and appears to be more commo n than a true anatomic variation. va riatio n. common rian artery can ca n be Supply to the uterus from the ova ovarian bolization of the relevant ovarmanaged by proximal em embolization ian altery with Gelfoam (Pharmacia (Phannacia & Upjohn, Kalamagets after embolization of all demonstrable zoo, MI) pled pledgets uterine arteries. It is known from current curren t experience ith postpartum and postsurgical he morrhage that proxw with hemorrhage imal embolization of the uterine and ovarian arteries with Gelfoam oorr coils does not interfere with menstmal menstrual futu re fertility. In my experience, this maneufunction or future ver has apparently decreased the .filling filling pressure to the utems fibroids without uterus sufficiently to infarct the patients' .fibroids causing any symptoms of ova rian dysfu nction or failure. ovarian dysfunction In cases in which there is supply to the uterus by both ovarian arteries, alteries, I] have been unwilling unWilling to embolize both vessels because of possible ovarian damage. I do not feel (hat that bilateral ovarian artery embolization can be justified justi.fied or defended in the case of the elective treatment of fibroid disease. Bilateral absence of the uterine arteries has been demonstrated on two occasions in my experience. A further anatomic variant that has been seen in a few cases is the presence of multiple small branches of the anterior division of the internal internal iliac artery separately supplying the uterus, ute rus, w ithout the presence of a main without uterine artery trunk. In the occasions in which this has been seen, it was obviously not no t feasible fea sible to separately catheterize and embolize each of these vessels, so UAE plans were abandoned. Summary butt nO[ not always, be manVessel tortuosity can ca n usually, bu aged in performing UAE. Anatomic variants that ma mayy effect the outcome of UAE are rare rare,, but most can be managed if they are noted. One should always be se nsitive to the possibility of sensitive
the occurrence occu rrence of these variants so that they can be dealt with at the time of initial initial UAE. 4:20 pm
Embolization of Symptomatic Uterine EmboUzation Leiomyomata: Measuring Outcomes james B. S pies, MD James Spies, Georgetown iversity Medical Center Georgetoum Un University Washington, DC embo liza tion of uterine The initial published results of embolization le io myomara have been very positive and have generleiomyomata in terest in this alternative to surgical ated considerable interest surgical therapy. However, However, investigators need to better define [he the ooutcome utcome from this treatment and develop more comprehensive hen ive measures of therapeutic success. In this review, review, we will prov ide an overview of the ]n provide measures of outcome that have been used to date. date. We will discuss the potential pitfalls that may occur in depotential veloping outcome measures for assessing symptomatic improvement and suggest ways of improving our analtient outcome. We w ill also introduce alternapatient will ysis of pa bility in future studtive measures that may have applica applicability ies of this procedure. Current Published Experience There have been six case series that have been published in th 0 -6) and two more that thee EngHsh English literature Cl-6) will be published in October of 1999 (our (ou r own initial experience in 61 61 patients and a report by a group at iar with the measures used in our UCLA). As we are famil familiar own stu dy (7), we have included these in Table 1. study All these authors have used different measures of outcome, which are summarized in Table 1. Written sym ptom and satisfaction questionnaires were used in symptom four studies, phone interview in two (these two reports are from the same group, with the patients from the first mbination of combination study included in the second) and a co interview and questionnaire in oone. ne. Generally, the quessnldies were framed to assess imtions used in these studies provement rity of symplOms sym p[Qms pI' vement by comparison with the seve severity ca n be seen in Tab]e Table 1, a variety embolization. As can before embolization. of question fo rmats were used. This makes comparing compa ring formats patient ooutcome utCome difficult. In addition, beca use interview interv iew because and written questionnaires have been used, even more variability in response may have occurred. Therefore, altho ugh the reported repo rted results of UrE UFE are encouraging, e ncouraging, it although is difficult to determine if there is a difference in the resul ts of the various studies and how these results might results compare to alternative therapies such as myomectomy. wou ld be helpful if a standard standa rd format for questio ns It would questions iminate the could be adopted for use. This would el eliminate assess ment of symptom change and devariation in assessment potential bias that might be introduced by the crease the pmentiai wording or formatting of questions.
DeSigning Designing Symptom Scales to Measure Outcome During our experience at Georgetown University, we have learned lea rned the difficulties of questionnaire develop-
Table 1 Outcome measures used in initial publications of uterine fibroid embolization No. of
Source of followup data
Ravina cr aI, Lancet
Outcome measure scale
Success, failure, stable then D and C
Completely resolved, much or signiJicanl improved, slightly improved, unchanged, slightly worse, much worse
1995 Goodwin et ai, ]VlR
mean 5.8 months postprocedure Bradley et ai, Sf) Obstet Gynaecol,
Worthington-Kirsch et ai, Radiology, 1998
Interview 3 months posrprocedure
Burn el al, Bf J Radiol,
Symptom outcome by interview Satisfaction with outcome by questionnaire (strongly agreelagreel neutraVdisagree/su-ongly disagree) Stopped completely, greatly improved, moderately improved, slightly improved, unchanged, worse (for pain and pressure, "not applicable" also a choice) Not stated
Hutchins et ai, )
posrprocedure Telephone Interview at 3 months, 6 months and 12 months
Stopped completely, greatly improved, moderately improved, slightly improved, unchanged, worse (for pain and pressure, "not applicable" also a choice)
Spies et ai, ]V1R 1999
11 point scale from markedly worse (-5) to markedly improved (5), with o unchanged (see text for example)
Mean 8.7 momhs postprocedure
Table 2 The new scale used for evaluation of UFE procedures
Slightly Markedly Worse
ment through trial and error. Initially, we used a questionnaire to assess symptomatic improvement with a scale of 5 levels of choice: much worse, slightly worse, unchanged, slightly better, and much better. This scale was used for 3-month follow-up in 48 patients and for 6-month follow-up in 30 patients. As we progressed in our study, we noticed that this scale was not adequate for assessing patient improvement. Some patients indicated a degree of improvement between "slightly better" and "much better." Most of these chose "slightly better," but indicated in comments that their level of improvement was between the two choices. After some thought and discussion, we changed the scale we used to the one displayed as Table 2. For all the subsequent patients and for later assessments of the earlier patients, we have used this elevenpoint scale, ranging from markedly worse to markedJy
improved. This has allowed a broader range of choices, and it is our hope that we will better characterize patient response. However, this has left us with two sets of data, with some patients having been given two different questionnaires. For a cross sectional repon of patient outcome (such as the study we are publishing next month) the response from only one questionnaire is used and the two scales do not need to be combined. When the progress of patients over time is reported, all the interval questionnaires must be used and the two scales must be merged. There are no clear rules to guide this type of conversion. It is perhaps most reasonable to consider all previous responses of "slightly better" as comparable to "slightly better" on the new questionnaire and "much better" on the previous response to be equivalent to "markedly better" on the new questionnaire. .!\lthough this process does combine the data, it intro-
duces uncertainty and may understate the benefit of the therapy. For example, with the old questionnaire, if a patient was improved, but not "much better," the only other choice was slightly better. If she had been given the new questionnaire, she might have the choice of a 2 or 3 or 4, all interval responses greater than "slightly better" and less than "markedly improved." Therefore, a patient with moderate improvement would be categorized as only "slightly better," leading to an underestimation of the benefit of therapy. A more fundamental potential difficulty is using symptom change as the primary measure of outcome. When asked to assess the change in symptoms since treatment, he patient must rely on memory of baseline symptoms to answer. As a prospective study progresses, the memory of the patient may fade and it may be difficult to accurately answer the question. In addition, with a comparative question, the patient is asked to judge whether the procedure was successful. There is a natural tendency to validate the choices one makes by confirming their successful outcome. We tend to "buyin" to the choices we make. By rating the therapy a marginal success or failure, a patient must admit that they made an error in choosing the procedure. There is a natural tendency to avoid this type of choice and hence comparative questions encourage a positive response bias. A more objective assessment might be obtained by asking a patient to rate the severity of their symptoms at baseline and at each follow-up, with use of the same questions. These ratings can be converted to a numerical score, which can be compared with use of the appropriate statistical tests. This may help to reduce the bias toward success and the reliance on memory. The optimum structure of a symptom questionnaire for this procedure has not yet been agreed upon and the previous discussion highlights some of the problems in the assessment of symptom outcome that might be encountered. We are hopeful that a fibroid-specific symptom questionnaire will be developed in the near future with the assistance of an experienced outcomes consultant. This could be brief, concentrating on the symptoms most closely associated with fibroids. If a simple scoring system was created, the questionnaire could be used by all practitioners performing this procedure, including those participating in the proposed national registry. It might also have applicability in assessing the outcome of other therapies for leiomyomata and it might serve as a basis of comparison among different therapies.
Measuring Health-related Quality of Life A broader assessment of health status and outcome from therapy is possible using a quality-of-life questionnaire. This type of questionnaire asks questions to assess overall health status, including broad categories such as general health perception, mental status, bodily pain, Vitality, emotional status, self image, and sexual functioning. At Georgetown, we have completed a pilot study using
a fibroid-specific quality of life questionnaire to assess outcome from .fibroid embolization (8). The questionnaire used was developed by Lewin-TAG, a health outcomes consulting group. A sample of the questions used are given in the Appendix. The goal of our study was to determine whether symptom changes occurring after fibroid embolization could measurably impact health-related quality of life. These changes were detected with statistically significant improvement in all the general health scales at month 3 and month 6, with marked improvement in general health perception, comparative health perception, physical functioning, energy, mental health, sexual functioning, and self-image. There were corresponding decreases in the three measures of restricted activity, difficulty with activity, pain, and health distress. All the symptom-specific measures Significantly improved at month 3 and all but one (backaches) improved at month 6. These findings suggest that health-related quality of life is substantially improved by FE. Because it was our intent to assess the changes that might occur in health status as a result of UFE, we needed an evaluative instrument that would allow us to assess outcome over time. There are two basic categories of health-related quality of life questionnaires, generic and disease-specific (9). A generic instrument is usually comprised of questions assessing broad categories of health status and does not contain disease-specific questions. An advantage of this approach is that these generic questionnaires have been extensively validated, have age- and sex-specific norms available, and often have extensive data available for individuals with different clinical conditions. A disadvantage of generic instruments is that they may not be sensitive enough to detect changes in the status of patients with a specific medical condition. This may be particularly tOle for a condition such as uterine leiomyomata that may only cause symptoms several days a month and impact only a few specific aspects of functioning and well-being. Generic instruments may be supplemented by the addition of symptom-specific questions and by the addition of domains that reflect areas of importance for that condition. Our questionnaire contained several generic scales from the Medical Outcomes Study and thus overlaps the SF-36 to some degree. But very importantly, fibroid-specific questions and domains for restricted actiVity, sexual functioning, self image, and health distress were added to include areas important in this condition. Quality of life assessments have been used in studies of gynecologic conditions, with outcome of treated patients compared to the normal population. In one study of hysterectomy in Great Britain using the Nottingham Health Profile, the health status of subjects at 6 weeks after surgery became indistinguishable from women in the general population (10). A number of other studies of hysterectomy and other fibroid therapies have measured outcome with the use of health-related quality of life questionnaires. In a pro-
spective spective cohort study of 418 women undergoing hysterectomy, ectomy, Carlson Carlson used a variety of quality of life measures including including a global rating, a mental health index, a general eral health health index, and an activity index (l1). (11). Significant improvements improvements in mental health, health, general heailh, health, and activity tivity indices indices were reported at 6 months and sustained at 11 year. year. Because Because the symptoms which patients experience from from fibroids fibroids are variable and subjective and because these these symptoms symptoms impact overall quality of life, further study study of health-related quality qua lity of life will add significantly cantl y to to our our understanding of the success of therapy in this this group group of patients. Menstrual Menstrual calendars Menstrual calendars are pictorial records of the quantity of of menstrual menstrual flow flow recorded by a patient during a menstrual strual period. period. They Th ey are a more objective measure of the effect effect of therapies directed at controlling menorrhagia than than questionnaires. questionnaires. We have begun using a menstrual calendar calendar to compare severity of menstrual bleeding before aher embolization. We hope to report our initial fore and and after results results at the SCVIR SCVIR 2000 2000 Annual Meeting in San Sa n Diego. This This type type of record may help he lp validate questionnaires questio nnaires that arc menorrhagia. are developed developed for assessing control of menorrhagia. These These records records have been lIsed used in other leiomyomata therapies anorher basis for therapies and may thereby proVide provide another comparing patients. patienrs. comparing Complications Complications Just as as standardized standardi zed assessment of outcome would imJust prove the the evaluation of this therapy, therapy , so would uniform uni form prove report') reporting of complications. None of the current report'5 reponing (includ ing our our own) have used a standard classification (including for complications. complications. A standard scale has been developed for by the the SCV1R SCV1R and its use should shou ld be encouraged. This by scale was was developed to rate the severity of complications scale based on their outcome. o utcome. Complications are rated as folbased lows: lows:
we gather gather regarding regarding fibrOid fibroid It is important that the data we of physicians, physicians, papaembolization with'stand emboli zation with· stand the scrutiny of tient.'i, and payers with an interest in this thiS condition. condition. We We tient'), of the the measures measures use of must determine the outcome with use interpretable, and and that are clear, standardized, readily interpretable, available standard standard reliable. As our knowledge of the available to report report results results in in terms terms scales grows, we will be able to by skeptics skeptics and and more that are mo re likely to be accepted by advocates alike.
References D, Ciraru-Vigneron N, N, et et al. al. 1. Ravina J, Herbreteau D, treat uterine uterine myomata. myomata. LanLanArterial embolisation to treat cet 1995; 346,671-672. 2. Goodwin S, Vedantham S, McLucas McLucas B, B, Forno Forno A, A, Perrella R. Preliminary experience experience with with uterine uterine arartery embolization for uterine fibroids. JV1R JV1R 1997; 1997; 8,517526. 8517-526. 3. Bradley E, Reidy), Reidy J, Forman R, Jarosz Jarosz J,J, Braude Braude P. P. Transcatheter uterine artery embolisation embolisation to to treat treat large uterine fibroids Obst Gynaecol fibroids.. Br) Br JObst Gynaecol 1998; 1998; !O5, 105, 235-240. 4. Worthington-Kirsch R, Popky G, Hutchins Hutchins F. F. Uterine Uterine arterial embolization for the management of management of le iomyomas: quality-of-life assessment leiomyomas: assessment and and clinical clinical response response.. Radiology 1998; 2080625-629. 208062>-629.
Minor Complications Minor 1. No No therapy, no consequence 1.
5. Burn P, McCall J, Chinn R, Healy). Healy). Embolization Embolization of of uterine fibroids. Br) Br J Radiol Radial 1999; 72,159-161. 72,159-161.
2. Nominal Noll1.inal therapy, no consequence; includes over2. night admission for obseIVation observation only. o nly. night
6. Hutchins F, Worthington-Kirsch R, Berkowitz Berkowitz R. R. SeSelective uterine artery embolization as primary embolization as primary treattreatment for symptomatic leiomyomata leiomyomata uteri. uteri. JJ Am Am AsAssoc Gyn Laparosc 1999; 6,279-284. 6,279-284.
Major Complications Major 3. Require Require therapy, minor hospitalization «48 hours) 3. 4. ReqUire Require major major therapy, therapy , unplanned increase in level 4. of ca re , prolonged hospitalization (>48 hours) of care, 5. Permanent Permanent adverse adverse sequelae 5.
complication. endometrial infection infecrion would be a level 33 complica tion. and required reqUired aa What if that patient did not respond and hysterecromy the same admission? Is this this aa level level 44 comcomhysterectomy levelS? unanticipated hysterectomy hyslerecromy plication or is it leve l 5? Is an unamicipated clarifica~ sequelae? further clarificapermanent adverse seq uelae? Without further comparison of of complication, ambiguity may cloud a comparison differem studies. SCVlR SCVlR has has appointed appointed tions reported in different UFE, committee to to develop reporting slandards smndards for for UFE, a cOllunittee central part part of of with the resolution of issues such as this aa central charge. the committee charge.
6. Death Death 6. If all all authors reported complications grouped by seIf verity number level, it would allow more uniform assessverity ment of of the safety of the procedure. Some additional ment consensus clarification will likely be needed to ensure e nsu re consensus uniform application applica tio n of these measures in UFE. UFE. For exuniform ample, most most practitioners practitioners would would agree agree that that readmission readmission ample, for intravenous intravenous amibiorics antibiotics for for 22 days days for for suspected suspected mild mild for
7. Spies J, Scialli A, Jha ]ha R, et al. Initial Initial results results from from uterine fibroid embolization embolization for for symptomatic symptomatic leiomyomata. JVlR 1999; 10,11 49-11 57. leiomyomata.)VIR 10,1149-1157. 8. Spies J, Warren E, Mathias S, Walsh Spies], Walsh S, S, Roth Roth A, A, PenPentecost M. Uterine fibroid embolization: embolization: measurement measurement of health-related quality quality of life li.fe before before and and after after thertherapy. JVlR 1999; 10,1 293-1303. apy.)VIR 10,1293-1303. 9. 9. Patrick D, Devo R. R. Generic and and disease-specific disease-specific measures in assessing health status status and and quality quality of of life. Med Care 1989; 27,S2 17-S232. life. Med Care 27,S217-S232. 10. 10. Clarke Clarke A, A, Black Black N, N, Rowe Rowe P, P, Mott Mott S, 5, Howle Howle K. K. IndiIndi-
cations for and outcome of toral total abdominal hysterectomy for benign disease: disease: a prospective prospeccive cohort 620. study. Brit] Obstet Gynaecol 1995; 102,611102:611-620. K, MUler Miller B, Fowler F. The Th e Maine Women 's 11. Carlson K, Women's 1. Outcomes of hysterectomy. hysterectomy. Obstet Health Study; L Gynecol 1994; 83556-565. 83:556-565. Appendix A: Sample Items from a Fibroid HealthQuesti01Ulaire related Quality-of-Life Questionnaire MONTH, how much difficulty did you have In the PAST MONTIf, doing your daily work, both inside and outside the house, because of your health? 1. None at all 2. A little bit 3. Some 4. Quite a bit 5. Could not do daily work at all MONTIi, how much of the time rime During the PAST MONTIi, were you frustrated about your health? 1. All All of the time tim e 2. Most of the time 3. A good bit of the time 4. Some of the time 5. A little of the time
6. None of the time I am more active than I was before receiving this trea tment for my fibroids treatment 1. I. Strongly agree 2. Somewhat agree 3. Neither agree nor disagree 4. Somewhat disagree 5. Strongly disagree MONTIf: During the PAST MONTI-I, (Circle one number on each line) 1= = Not at all; 2 = A little; 3 = Quite a bit; 4
a. to what extent were you interested in sex? 12 1 2 3 4 b. to what extent were you sexually active? 1 2 3 4 c. to what extent has sex been enjoyable for you you?? 1 2 3 4 d. to what extent has sex been painful for you'123 you? 1 2 3 4 4:40 pm Uterine Fibroid Embolization in a Community Hospital Setting john C. Lipman, lvID John MD Piedmont Hospital Atlanta, Georgia Uterine fibroid embolization (UFE) is just one of a number of gynecologic procedures that we perform. This Th is HSG, sonohysterography, and a pliments HSG, procedure com compliments
series of other procedures thal that are part of a declicated gynecologic service. This seIVice service is part of a larger diviinteIVentional radiologists in sion of fellowship-trained interventional our practice. To better identify ourselves to our colleagues in the hospital and to the community at large we formed MICRO. MlCRO. sta nds for Minimally Invasive InVasive Care through MICRO stands Radiologically directed Outpatient therapy. Although the word y, it is impoltant fo r the interventional inteIVentional important for title is rather wordy, seIVice to be separate from the rest of oour ur practice. To service borrow from Barry Katzen, the key strategies for practice development were as follows: (0 (i) to educate about benefits of interventional procedures, (ii) to prove a commitment [Q to deliver dinical clinical care and close follow-up follow-up,, and (iiO to become an equal partner in the delivery of care. (tiD MICRO was the first nrst step in an evolution of a dedicated seIVice. clinical inteIVentional interventional service. procedure, but I would caution ca utio n UFE is an exciting procedure, those that want to begin performing perfo rming this procedure. This procedure, procedure, although technically not difficult for the skilled interventionalisl, interventionalist, does require a tremendous commitmenl on the part of the whole section. sectio n. There mitment ll1ere are two main issues with performing UFE: the new clinical skills necessary in regards (Q to the management of fibroids and parienr's and her gynecologist's gynecologis['s expectations with the patient's regard to he herr after care. My two main recommendations to those that want to ro practice are: incorporate UFE into their interventional interventional dialogue/relationsh.ip with their Ob1. Develop a dialogue/relationship Gyn colleagues early on: gyneco logists Talk to the chief of Ob-Gyn and other gynecologists UFE. Begin the exchnage of information info lmation about the about UFE. da ta in clinical managemenl management of uterine fibroids. Present data the medical literature oorr lecture at Ob-Gyn Rounds. Discuss a possible research protocol or developing guidelines for patients. We have developed a joint interdeUFE . We are also on the partmental statement on UFE. hysterectomy pathway and UFE is part of the hysterectomy consent. 2. Stan Start a regular clinic: O ur clinic is every Thursday afternoo Our afternoonn and we schedule patients every 30 minutes. minutes. Charts Chans are created nsultations. It includes office notes, H&P, H&P , on all co consultations. recent imaging, and biopsy information from the gynecologist. We add to this our o ur 3-page form, form , which whlch includes a page on chief complaint, history of present illness, past medical/surgical hiStory, history, and any pertinent gynecologic history, a procedural procedura l page filled out the da izatio n, and a clinical follow-up follow -up dayy of the embol embolization, page. appointment. the UFE procepage. During the clinic appointment, dure is discussed in detail ptio ns are considdetail,, other ooptions eered, red, and potential complications are thoroughly th o roughly explained. appOintment, a letter is generated genera ted to plained. After the appointment, necologist with our recommendation recommendatio n for treatthe gy gynecologist men£. pament. We have hired a full-time specials nurse (se (separate from those that are employed by the hospital). She helps with the afternoon clinic and is responsible teleph one follow-up, scheduli ng imaging followscheduling followfor telephone