The persisting confusion regarding the treatment and natural history of endometriosis suggests that some of our clinical impressions may need modern restudy. Some clinicians believe that endometriosis involves progressively more widespread areas of the pelvis as patients get older and that pregnancy confers relative protection against spread of the disease throughout the pelvis. To examine these beliefs, this study evaluates both the frequency distribution of disease by age groups as well as the number of pelvic areas involved compared with age groups and fertility status in patients with endometriosis.
Materials and methods
One hundred thirty-two consecutive patients, 16 to 52 years of age, undergoing either laparoscopy or laparotomy, constituted the study group. Chronic pelvic pain was experienced by 72%, dysmenorrhea by 45%, dyspareunia by 39%, and infertility by 29%. Thirteen patients were asymptomatic, their disease discovered at surgery done for other indications. Sixty-five percent had been pregnant, while 51% were parous. Fourteen percent had either terminated or spontaneously aborted their only pregnancy without a subsequent live birth.
Magnification of the peritoneal surface was frequently employed, consisting of routine near-contact laparoscopy by a double- or triple-puncture technique and intrauterine cannula, or use of the operating microscope at laparotomy in the early stages of the study. Any visual abnormality of the pelvic peritoneum was considered to be possible endometriosis. Most patients had several separate specimens submitted for pathologic study. Endometriosis was histologically confirmed in 129 patients by the presence of both grandular epithelium and stroma and diagnosed by the presence of black "powder-burn implants" in 3. Five patients had a prior diagnosis of endometriosis, and review of their operative reports indicated no change in position of their disease in the pelvis from their previous surgery.
A pelvic map dividing the pelvis into 16 separate areas was used prospectively in all patients to record distribution of biopsy-proven disease. The study design sought to identify noncontiguous disease spread of centimeters or more. Local "microspread," local spreading of fibrotic plaques, and presumed invasion of the peritoneum were not evaluated because they were not pertinent to the purpose of the study.
Patients were grouped into 5-year age intervals for study, and patients older than 36 years of age were combined.
A probability value (P) of less than 0.05 was chosen as a means of ruling out the null hypothesis that there was no significant difference between group means. Probability was determined by use of the standard error of the difference of the means.
The frequency of involvement by endometriosis of the study areas of the pelvis in the different age groups is shown in Table 1.
Table 1. Frequency of involvement of pelvic sites versus age
|Site||Frequency of involvement (%)|
|16-20 years||21-25 years||26-30 years||31-35 years||36-52 years||All ages|
|Right broad ligament||57.1||43.5||32.1||39.3||20.0||36.6|
|Right uterosacral ligament||57.1||39.1||43.4||25.0||12.0||37.9|
|Left uterosacral ligament||42.9||52.2||32.1||25.0||20.0||35.6|
|Left broad ligament||35.7||39.1||35.8||25.0||12.0||32.6|
|Left round ligament||0||0||5.7||3.6||0||3.0|
|Right abdominal wall||0||0||3.8||0||0||1.5|
|Right round ligament||0||0||0||0||0||0|
|Left abdominal wall||0||0||0||0||0||0|
|Total number of patients||14||22||52||25||19||132|
Patients 16 to 27 years of age (n = 55) had a mean of 3.25 ± 1.92 (standard deviation [SD]) pelvic areas involved, whereas patients 28 to 52 years of age (n = 77) had 2.60 ± 1.64 (SD) pelvic areas involved, a significant difference (P < 0.046).
Nulligravid women 16 to 27 years of age (n = 27) had a mean of 3.59 ± 2.12 (SD) pelvic areas involved, whereas nulligravid women 28 to 52 years of age (n = 20) had a mean of 3.00 ± 2.00 (SD) pelvic areas involved, not significantly different.
All nulligravid women (n = 46) had a mean of 3.39 ± 2.00 (SD) pelvic areas involved, while all ever-gravid women (n = 86) had a mean of 2.45 ± 1.64 (SD) pelvic areas involved, a significant difference (P<0.01).
Because ever-gravid patients were a mean 5 years older than nulligravid patients, and because older patients in this series for some reason have slightly fewer pelvic areas involved than do younger patients, age-controlled comparisons are shown in Table 2. With controls for age, in two age groups there is a significant difference in number of pelvic areas involved in patients who were delivered of no infants compared with patients having up to four infants. In three age groups a significant difference was not seen.
Table 2. Age-controlled number of pelvic areas involved versus parity
|Number of pelvic areas involved ± SD|
|16-20 years||21-25 years||26-30 years||31-35 years||36-52 years|
|0||2.80 ± 1.87 (n = 10)||4.20 ± 2.34 (n = 15)||3.27 ± 1.39 (n = 26)||2.56 ± 1.42 (n = 9)||2.67 ± 2.89 (n = 3)|
|1-4||3.50 ± 0.58 (n = 4)||2.00 ± 1.53 (n = 7)||2.27 ± 1.69 (n = 26)||2.50 ± 1.32 (n = 16)||2.31 ± 2.17 (n = 16)|
|P||0.1, NS*||<0.01||<0.046||>0.5, NS||>0.5, NS|
I believe there are only two ways to validate objectively the notion of progressive pelvic spread of endometriosis over time: (1) laparoscopically monitor a group of untreated patients serially over several years, prospectively mapping their pelvic involvement and (2) identify a larger proportion of patients in older age groups with more frequent involvement of individual pelvic areas or with more pelvic areas involved than in younger age groups. This study attempts the second approach. The results suggest that endometriosis is a positionally static disease, not spreading in the pelvis with advancing age.
The notion that pregnancy protects against progressive pelvic spread of disease can be tested either by making age- and fertility-stratified comparisons of the frequency distribution of pelvic involvement by disease or by finding an inverse relationship between number of pelvic areas involved and exposure to pregnancy. This study attempts the second approach. The finding of more widespread disease in nulligravid patients than in ever-gravid patients is mitigated by the finding on more detailed analysis that this inverse relationship was not found in all age groups.
A greater number of areas of involvement in nulligravid patients than in ever-gravid patients does not necessarily imply that greater amounts of disease cause relative infertility or are a result of postponing pregnancy. The presence of endometriosis may simply be a marker of some other acquired or developmental pathologic state that may be related to the infertility and that remains after the endometriosis is gone.
Because the most common symptom in this study was pain, did this select a group of patients with an aberrant distribution of disease? Concern over selection bias could apply to any study of endometriosis, because presumably the disease is discovered by virtue of surgery done for some clinical indication, apparently usually infertility. The decline in apparent disease with age noted here could mean that patients with slightly more disease experienced slightly more pain and therefore came to diagnose slightly earlier in life.
The suggestion that endometriosis may be a positionally static disease would give renewed emphasis to conservative surgery for treatment of patients desiring future fertility.
The pelvic mapping used in this study is a potentially powerful tool for investigating the disease, particularly when combined with peritoneal magnification and frequent, vigorous biopsy by clinicians attuned to the varied visual manifestations of the disease
In this series of patients, endometriosis does not involve more pelvic areas in older age groups. Exposure to pregnancy is associated with slightly less pelvic involvement, although this effect is inconstant when age groups are studied individually.