Uterine phlebography Correlation
of clinical
ROBERT
R.
DEAN
D.
Houston,
Texas
diagnoses
HUGHES,
CURTIS.
with
dye retention
M.D M.D.
‘l’he majority of the patients were from the gynecology clinic of the Jefferson l.>avis Hospital and were primarily indigent Negro women. Each patient was carefully interviewed by one of the authors and the results recorded on a standard form including the 20 most commonly elicited complaints listed by Taylor” in his study of 200 patients with the pelvic congestion syndrome. The pertinent physical findings from the abdominal and pelvic examinations were included. The diagnostic impressions were divided into three categories designated as normal, organic disease: and nonorganic conditions. The normals were without significant gynecologic complaints or findings on physical examination. Patients with complaints caused by obvious pelvic pathologic conditions were classified as having organic disease. Under the nonorganic category were listed patients with the clinical diagnosis of the pelvic congestion syndrome and those with atypical pelvic pain with minimal physical findings. Of the patients with the pelvic congestion syndrome, the most commonly encountered complaint was pelvic pain of long duration. Other frequent complaints were hackachc. dyspareunia, and dysmenorrhea. These symptoms were usually aggravated by premenstrual tension. There were multiple other associated symptoms. Pelvic examination re\,ealed bimanual tenderness, and an enlarged soft uterus, occasionally retroverted. with a somewhat soft and cyanotic cervix. In the study, two categories of the syndrome nw(
U T E R I N E phlebograms are roentgenograms of the venous outflow of the uterus, obtained by injecting radiopaque media into the myometrium of the fundus. The technique has been outlined by numerous authors.‘, z By taking immediate and delayed x-rays, the rapidity of the blood outflow: the number and size of the blood vessels, and the location of the venous drainage of the uterus and adnexa can be demonstrated. The purpose of this study is to correlate the venous drainage of the uterus and adnexa with a clinical diagnosis. One hundred and five patients were included in this study. 11 of whom were considered gynecologically normal with normal phlebograms. These were compared with 94 patients with various pelvic conditions including 22 with findings compatible with the pelvic congestion syndrome. The uterine phlebograms demonstrate definite physical changes in the venous drainage of the uterus and adnexa in patients with various clinical diagnoses. This paper outlines these findings. These one hundred and five patients were injected one or more times and 125 uterine phlebograms obtained. Ninety of these patients were injected during the first half of the menstrual cycle, to eliminate the physiologic vascular changes associated with premenstrual congestion. From the Department of Obstetrics and Gynecology, Baylor University College of Medicine, and Jefferson Dark Hospital.
156
Uterine,
Table I. Distribution
phlebography
157
by age and parity Parity
Category
NO.
Age
Normal
11
Organic
67
Nonorganic
27
Total
Table
17 30 16 30 26 30
group
0
to to to to to to
1 0 3 3 1 0
3 3 ‘I ‘I ;?I 8 8
8
63
2!9 4!) 29 48 29 44
105
II. Dye retention
6-12 0 4 5 15 3 7 34 ______~_
in the pelvic veins at 20 second delay films Moderate
Diagnosis Normal Organic Cervical erosions Pelvic inflammatory Pelvic relaxation Postpartum Other problems Nonorganic Pelvic congestion pain -__-Atypical
l-5
I
/
disease
syndrome
Total -__
No.
lxtramurai
or marked
of
dye
Uterine
1 Ozvzrian
11
0
1
0
1
9
14 21 19 6 7
4 3 0 4 2
8 3 2 3 2
7 7 2 5 3
11 7 3 6 4
79 33 33 100 57
22 - 5 ______._--
6 0
18 -.-_____~~- 1 38
15 1
22 1
100 20
40
55
-___-.52
105
\
account
19
noted. Both had the typical histories with only the more severe category having all the physical findings. Patients with atypical pain had intermittent dysmenorrhea, backache, and dyspareunia with variable pelvic findings.
/ TyxL
injected following an intravenous sensitivity test. The dye is injected in a period of 30 to 40 seconds. A film is taken immediately at the end of the injection and a second film 20 seconds later. If an intravenous pyelogram is desired, the usual number and intervals for those films are obtained.
Technique The technique of injection is similar to that outlined by Topolanski-Sierra.? Preparation similar to that done for a hysterosalpingogram is performed with the patient in the lithotomy position. No premeditation is given. A special cannula with a needle is inserted through the endocervical canal to the fundus. The needle is then cleared of ail and projected through the cannula into tht myometrium. One hundred and fifty units of hyaluronidase + in 3 ml. of normal saline is injected into the myometrium. Twenty milliliters of a radiopaque media? is then *Wydase,
Wyeth
Laboratories,
tRenografin, 60 per Ywk, New York.
cent,
Philadelphia, E.
R.
Squibb
Pennsylvania. & Sons,
Nerv
Method
of interpretation
The interpretation of results u-as made as objective as possible. Measurements of the size of individual vessels and groups of vessels were made directly from the films. Even though the measurements were exaggerated by the technical aspects of x-rays, it was felt that comparisons were justified since the distortion should be fairly constant. The location of the vessels and the amount of dye in each of them were recorded. The concentration of the radiopaque media in the vessels was judged as being none, minimal, moderate, or marked and graded as 0, 1, 2, or 3, respectively, for each of the 3 sets of veins found, i.e., the intramural, ovarian, and
58
Hughes
and
Curtis
A
B Fig. 1. Normal uterine phlebogram. A, The injection film reveals veins and the right uterine vein. B, The 20 second delay film with vessels.
opacification of the ovarian little opacification of pelvic
A
B Fig. 2. Uterine phlebogram illustrating severe pelvic congestion syndrome. A, The injection shows the ovarian veins bilaterally. B, The 20 second delay film shows the ovarian veins faintly with arrows pointing at the marked opacification of the uterine veins.
film more
Uterine
uterine veins. The 20 second delay film was compared to the injection film and changes in the amount of dye was noted. Results
The distribution, number, and parity of the subjects are noted in Table I. Sixty-seven patients had organic disease, 11 were normal, and 27 had nonorganic problems. In each category, the patients are divided into those above and below 30 years of age. The ages and parity in the three categories were comparable. No patient listed under the nonorganic category was younger than 26 years of age. In Table II, the findings of the 20 second delay films are outlined. Only a moderate or marked amount of dye seen in the intramural, uterine, and ovarian veins on the delay films is recorded. Those vessels with a minimal amount of dye are not included. The three categories are broken down for a specific diagnosis. The last column shows the total number and the percentages of each category demonstrating a delay in the excretion of the dye. Only one of the 11 normal patients had retention of dye in 20 seconds. This dye was seen in the uterine veins. This particular patient had no gynecologic complaints and no abnormal findings. Fig. 1 illustrates the typical normal. On the 20 second delay film, only a small amount of dye is seen remaining in the pelvis. All of the patients diagnosed as having the pelvic congestion syndrome had retention of dye in the 20 second delay film. Eighteen of the 22 patients had pooling of the dye in the uterine vessels. As noted in Fig. 2 on the injection film, there is no visualization of the uterine vessels. Twenty seconds later, a:s shown at the arrow point, the uterine veins are filled with a considerable amount of dye. In most of the cases, the uterine veins were visualized on the injection film and did not clear by the 20 second delay film. One patient with pelvic pain not typical of the pelvic congestion syndrome showed congestion on the phlebogram. There were findings of the pelvic congestion syndrome, but
phlebography
159
the history was too atypical to be included under that diagnosis. The other 4 patients in this group had pain inconsistent with the pelvic congestion syndrome and no positive physical findings other than pelvic tenderness. Thirteen patients with all degrees of cervicitis and erosion had phlebograms with 10 showing some congestion. One of this group had carcinoma in situ. An interesting case in this category was a 25-year-old Negro woman, para 4-O-O-3, who had a chief complaint of menorrhagia since a term delivery one year previously. She had steady lower abdominal pain with secondary dysmenorrhea and low back pain, premenstrual tension, vaginal pain, and a yellow vaginal discharge. On physical examination, she had lower abdominal tenderness and pain on manipulation of the cervix. Inspection of the cervix revealed marked cervical erosion. Pelvic examination otherwise was normal. Fig. 3 shows the phlebogram at the time of her initial visit. The large vessels showed marked congestion which is obvious on the delay film. The patient had complete remission of the symptoms and physical findings after 21 days of vaginal chemotherapy. Fig. 4 illustrates the marked change in the phlebogram showing a return to normal. The vessels are smaller and at 20 seconds little dye is seen. There were 21 patients with pelvic inflammatory disease. Seven of these showed congestion of the phlebograms. Of these 7, 5 were within one month of the acute phase of the disease while 2 had bilateral adnexal masses. Of these 7 patients, only 3 had uterine vein congestion. Of 19 patients with complaints related to pelvic relaxations, 3 had phlebographic evidence of congestion. Seven other patients had some degree of relaxation, but each was felt to have the pelvic congestion syndrome and were included under that category. Included in this study were 6 patients who were 2 to 6 months’ post partum. Each patient was selected because pelvic findings were similar to those in patients with severe pelvic congestion syndrome except for the
160
Hughes
and
Curtis
Fig. 3. Uterine phlebo~ram revealing at injrction. large dilated uterine veins. a large left ovarian vem. B, At 20 srconds. a ma&cd amount of dye runains. cerI%itis
lack of marked tenderness. Nontl of these patients had complaints of pelvic pain. Tht phlebograms rcvcalcd congestion in all 6 patients with the ovarian veins being in\-olvcd five times. This confirms the findings of a pre\~ious study that the main outflow from the fundus in thr, postpartum state is via the ovarian veins.” In contrast to patients with severe pelvic congestion syndrome, only 3 of the 6 patients had utcrinc kn pooling. ‘I’hree postpartum patients who were reinjcctcd 2 months later showed a dccrease in size? congestion. and venous flow through the ok arian \.cins. Of the patients listed under “o&r problems.” no 2 had the same disease: suhsequently, no significance can be attached to the findin,ys in those individual cases. However, over half of them did have some rctcntion of dye on thr 20 second film\. Of the 55 patients that demonstrated some congestion on phlebography, 4 gave a positive history for hemorrhoids. whereas 6 of
A, Bilateraliy with Diagnosis is ncut~
the 50 patients without congestion had hemorrhoids. None of the patients in this sc4es had leg L aricosities. Major operations wcw performed on 28 patients. Fifteen of these were patients with pelvic congestion syndrome. 13 of whom had hysterectomks. Two patients had a Doyle” type of neurrctomy combined with a vaginal xlterinc suspfkon. Vaginal hysterectomy with anterior d posterior colporrhaphy an was performed on 8 patients with syrnptotnatic pelvic relaxation. Each of the 2 patients with chronic pelvic inflammatory disease had an abdominal hysterectomy with hilateral salpingo-oophorectomy. The clinical impressions were confirmed i)y pathologic study of the remolled organs. C:hronic ccr\.icitis was the only pathologic tinding in those with a diagnosis of the pTl\;ic. congestion syndrome or pelvic relaxation. C%ronic cervicitis was found in 95 per cent of uteri removcxd at Jrfferson Davis Hospital.
Uterine
phlebography
161
B Fig. 4. Uterine phlebogram, cervicitis. A, The decrease is seen at 20 seconds.
a reinjection of the patient of Fig. 3 after treatment of the acute in the size of the vessels is marked. B, In contrast to Fig. 3. no dye
An attempt to study vessel size revealed that only the ovarian veins can be measured. The uterine veins were so often scattered that assessment of the size was difficult. The intramural vessel sizes varied over such a small range that it was impossible to categorize them. The ovarian veins varied greatly in size with separate vessels being easily measured. The measurements were taken directly from the injection film. Table III shows that a medium-sized group of vessels from 4 to 1’2 mm. was most common. Over half of the nulliparous patients had ovarian vessels less than 4 mm. in width. Conversely, the grand multiparas had almost three times as many large vessels (greater than 12 mm.) as small (less than 4 mm.). Thus, ovarian vessel size was found related to parity, women with higher parity having larger vessels. This is not accepted as significant because so many factors other than parity apparently influence vessel size. If enough normals were available, these other
factors fluence
could be eliminated and only the inof parity on vessel size studied.
Comment
Radiographic evidence of variability of genital blood flow is demonstrated in the 105 patients of this study. As had been outlined elsewhere and confirmed in this study, there are multiple factors causing pelvic congestiom4 A delay in the outflow of the dye injected into the uterine fundus was interpreted as demonstrating passive congestion. It was seen in only one of 11 patients considered Lgynecologically normal. Before elaborating on various categories associated with congestion, limitations and variables in this phlebographic technique should be discussed. The radiopaque dye used is important. The dispersibility of dye is secondary to its viscosity and rate of injection. In this series, these variables were controlled by the use of a single dye in all but 4 patients and by maintaining a steady in-
162
Hughes
and
Curtis
Table III. Size of ovarian Parity 0 1 to 5 6 to 12 Total
Under 4 mm. 5 9 5 19
--.
veins ______ 4 to 12 mm. 3 34 15
Over 12 mm. 0 18 14
52
3’,
However, since laxation occur upright position, conclusion since tempted in this Phlebographic
most symptoms in pelvic rtbwhile the patient is in the this might not be a valid no standing films were atseries. findings
in pelvic
infections
jection rate of 30 to 40 seconds. The needles used were of the same size (No, 19). The depth of penetration into the myometrium was varied from 3 to 5 mm. depending on where a steady adequate flow could be obtained. Retrodisplaced uteri had to be straightened before the cannula and needle could be inserted; however, after injection the uteri were released with a return to their previous position. Occasionally, the dye was injected subserosally, intraperitoneally, or into the endometrial cavity creating unsatisfactory films. Using these many criteria as limitations, phlebograms of many patients were unsatisfactory and were not included. With experience. less than 10 per cent of thr, films from recent attempts had to be reinjetted. The number of vessels visualized varied from case to case. Generally, reinjection of the same patient under the same circumstances revealed no change in the vessels visualized. The vessels most often visualized were on the right. No significance was attached to this as this did not seem to influence the retention or excretion of the dye. Although many variables do exist, we feel these were uniformly controlled so that I‘Csults could be compared. Normal phlebograms were demonstrated in patients whose ages ranged from 17 to 49 years of age and whose parity ranged from 0 to 12. Although there were only 1 I patients included in the normal category, we feel our data suggest that phlebographic evidence of congestion is not often seen in normal subjects of any age or parity. Since 16 out of 19 patients with symptomatic pelvic relaxation showed no evidence of congestion, we feel that the symptoms arc generally unrelated to pelvic congestion.
Of the patients whose phlebograms rcvealed congestion, it is easier to understand in those with infections. Eleven of 14 patients with cervicitis and erosion had congestion. Why 3 patients had no congestion seems more difficult to explain, except that perhaps these 3 had a very mild degree of infection. After treatment of the cervicitis, 2 patients were reinjected with the phlebograms returning to normal. Chronic pelvic inflammatory disease was not associated with congestion in 1-L of 21 with phlebograms patients. Five patients taken within one month of the acute phase of pelvic inflammatory disease had con,qcstion. Two patients with tuboovarian masses showed congestion. Consequently, the data suggest that patients with chronic inflammatory disease not associated with u~asses do not have congestion. Phlebographic congestion
findings
in
the
pelvic
syndrome
The most convincing evidence from this study confirming that the pelvic congestion syndrome is an entity is found in those cases designated as marked pelvic congestion syndrome in which the clinical diagnoses were clearly established and were correlated with positive phlebograms. Marked pooling and delay of disappearance of the radiopaque media was seen in the uterine veins on the phlebogram. ,4t hysterectomy, the large mottled uteri and the huge Lraricosities of the pedicles were seen. The pathologic studies revealed only mild chronic cervicitis. After more detailed study with special stains, no increase in fibrosis or elastic tissue was seen. The only significant difference from uteri of comparable parity \\-as an increase in ~cssel sixe and frequency. These findings will be the subject of another paper.“’ SubjcLctively,
Vdumc Number
83 2
Uterine
hysterectomy elicited a complete remission of symptoms in 11 of 13 patients followed for a period of 18 months. The patients with moderate pelvic congestion syndrome had symptoms similar to those with more marked pelvic congestion syndrome. The physical findings were less marked; tenderness and uterine retroversion were often the only findings. The uterine veins had moderate pooling in 14 of 18 patients. The other 4 patients had ovarian and/or intramural vein congestion. Findings at hysterectomy were similar to those of marked pelvic congestion syndrome and similarly the pathologic study was negative. In 2 patients with moderate pelvic congestion syndrome, a procedure combining a Doyle neurectomyl’ with a vaginal suspension elicited a favorable resp0nse.l’ Phlebograms showed congestion of the uterine and intramural veins before the operation, but no congestion afterward. The clinical results were satisfactory in one patient, with recurrence of symptoms after 6 months in the other. Diagnoses congestion
confused syndrome
with
pelvic
Although often the history is similar, the pelvic congestion syndrome can be differentiated from cervicitis by inspection of the cervix. The phlebographic findings are similar in both entities except that after adequate therapy of the cervicitis the phlebogram returns to normal. Pelvic relaxation can be differentiated from the pelvic congestion syndrome by history and by physical examination. We do not feel that uterine position or descensus contributes to congestion in the lithotomy position. The pelvic congestion syndrome can be differentiated from postpartum patients by history and phlebographic findings as discussed above. All of the patients with the pelvic congestion syndrome in this study were parous. Twelve out of 22 dated the onset of symptoms from a term delivery. These 12 might be classified by some under the diagnosis of subinvolution. However,
phlebography
163
since history and physical and phlebographic findings are the same as for the other 10, we do not feel that they should be segregated. For years it has been known that pelvic varicosities could be seen by various methods of injecting radiopaque media into the pelvic vessels.“-” Topolanski-Sierra and associates’ suggested that the pelvic congestion syndrome could be diagnosed by phlebography. However, in their study, they did not attempt to correlate the patients’ symptoms, physical findings, and treatment with the phlebograms. Duncan and Taylor,” in an elaborate study, demonstrated active hyperemia in the vaginas of patients with the pelvic congestion syndrome that were subjected to psychiatric interview. In neither of these studies were uteri removed to prove by gross and microscopic study that no other pathologic state existed. In this study, the criteria for diagnosing the pelvic congestion syndrome has been outlined. The pathologic findings in 13 patients subjected to hysterectomy are noted. By ruling out cervicitis, chronic inflammation, and pelvic relaxation, we feel we have narrowed the etiological factors of the pelvic congestion syndrome. Of those who recognize this entity, it is commonly felt that there is a psychosomatic element. Only a remission of phlebographic findings after adequate psychotherapy could support this thesis. This was not done in this study. There are 5 common causes of visceral pain as outlined by Wilson and Mussey.“: Two of these, anoxemia of functioning musculature and spasm of arterioles, could be factors in the cause of pain in the pelvic congestion syndrome. The blueness found in the cervix is assumed to result from an elevated blood CO, which is common in congestion. This alone does not cause pain as is evidenced by similar findings seen in the postpartum and the pregnant patient. Perhaps the congestion decreases the oxygen content with the resultant anoxemia of functioning musculature causing the pain. The psychosomatic element may enter into this by increasing the degree and frequency of uterine contractions. The resultant anoxemia also increases the chances for spasm of the
164
Hughes
and
Curtis
arterioles although the cause of this could he central in origin. This study has lxuught up sc\~al prohIcrns that deserve further investigation. No in the earl\ comparison of dye retention \.ersus the late stages of the menstrual cyck \vas made. Several patients wt:rt‘ injected txcause of interest but no conclusions art warranted. Another projected study included follow-up injection of postpartum patients in an effort to determine u-hen the in\x~llltiorl of the uterus is complete, as determined 1)~ this method of evaluation of the blood flow. Further investigation of the genital flolv with the use of radioisotopes is being- done to confirm t hc findings already presented in this study. Summary
The injection of radiopaque media into the fundus of the uterus demonstrated a \rariation in the sites of the venous otltflo\v. By our method, a 20 second delay film showed a \.ariation in the rate of outflow with a moderate to marked amount of dyes in the intramural, ovarian, or uterine vessels
REFERENCES
1.
2. 3. 4. 5. 6.
Guilhem, P., and Baux, R.: La phltbographie pelvienne par voies veineuse. osseuse et ut&ine: Application a l’ttude des phlthites et des cancers, Paris. 1954. Masson & Cie. Topolanski-Sierra. R.: AM. J. OBST. & Gsmc. 76: 44. 195X. Taylor. H. C.. Jr.: AM. J. OBYT. & GYNEC. 67: 1177, 1954. Taylor, H. G., Jr.: AM. J. OBST. 9r GI-NEC. 57: “11, 1949. Theobald, C. W.: J. Obst. & Gynaee. Brit. Emp. 58: 733, 1951. Greitz, ‘I’.: Acta radiol. 43: 429, 1958.
being accepted as congeslion. only one out of 1 1 cases blio~~d olltflow
of
Nor-~rtals
in
;I del;~), in
dvc,.
Ttvcnty-t\vo patients with thcx pelvic con,qestion syndrome had a delay in the outflow of dpc injecLed into thr: fundus of thcx utcrlls. bit11 18 of these brink mainly in the utcrinc l&s. FifLct:n of thcsc patitbnts catnt’ to operation with 13 ha\ing hysterectomy. No path&~+7 diagnosis othvr than mild chronic cc77kitis \z’x f011nd. Parity. ptAI\,ic relaxation. chronic cerlicitis, chronic ~~‘11 ic inflammatory disease, position of the uLrruk and patient age within the rcproductiu> I ang:e were found not to alter the circulation significantly. In I I of thr. 14 patients with cervicitis and uosion. ltlodcrate to lnarked congestion was seen. Afttlr adequate treatment of the ccr\icitis 111~ phlchograms returned to normal. Six postpartum patients showed delay in the outflow of dye injected into the fundus. ‘I’he majority of this dye was seen in the cxarian pcdiclcs. and xvith time this delay in outflo\v is decrt~ased.
T. 8. 9. 10. 11. 12. 13. l-1.
Lagcrgren, (1.: Acta radiol. 43: -t&l, 1958. Helaander, G. G., and Lindbom, A.: Acta radiol. 53: 97, 1960. Duncan, C. H., and Taylor, H. G., JI,.: AM. J. OBST. & c:SNEC. 64: 1, 1952. Kaufmati, R. H., and Hughes, R. R.: I In press. ) Doyle, J. 8.: AM. J. OBST. 6r GY\IK;. 70: 1, 1955. Wcxler, A. S.: Personal communication. Wilson. R. B., and Mussey. R. D.: J. A. M. A. 134: 857, 1947. Wegryn. S. B.. and Harron. R. ;2.: Obst. B Gynec. 15: 73. 196tl.