To many people, the term "microscopic" implies something so small it can be seen only with a microscope and that it is invisible to the unaided naked eye. "Microscopic endometriosis" is a term frequently thrown around by physicians for various reasons and with various interpretations.

For a pathologist who makes a living by looking at things under a microscope, all endometriosis would appear to be "microscopic" simply because that is the way pathologists view the disease. A pathologist has no reason to think about the gross or subtle visual appearance of endometriosis at surgery and how it compares to the microscopic view seen through the lens of the microscope, because a pathologist does not often come into the operating room. And it is also true that the diagnosis of endometriosis is most accurately made by observing tissue under a microscope, so the final diagnosis truly is one that depends on microscopy of some abnormality that has been biopsied by a surgeon who has seen evidence of that abnormality at surgery.

For the obstetrician-gynecologist who is busy delivering babies, and who does an occasional laparoscopy for pelvic pain, "microscopic endometriosis" can have a different meaning. The term would refer to endometriosis present on pelvic surfaces, but which cannot be seen because it is too small to be seen with the modest magnification (2X to 6X) afforded by a laparoscope. And certainly these surgeons would agree that microscopic endometriosis would be impossible to see at laparotomy, where the pelvic surfaces are viewed at arm's length. For such a surgeon, "Invisible Microscopic Endometriosis" (which will henceforth be abbreviated as IME) would be a ready explanation for persistent pain or justification for use of postoperative medical therapy to eradicate such disease. For surgeons with great experience in endometriosis, Invisible Microscopic Endometriosis simply does not exist.

For surgeons with great experience in endometriosis, IME simply does not exist. Virtually 100 percent of all the endometriosis in a woman's pelvis can be seen, but only if the surgeon knows what to look for. Use of the term "microscopic endometriosis" as the description of an apparently invisible entity is evidence of someone who is inexperienced in endometriosis and the history of its study. But how does one look for IME? Before we can answer that question, we have to know: How good is the human eye? The answer: quite good. Under ideal experimental conditions, the human eye can discern 0.5 seconds of arc [1]. At arm's length, this is approximately three microns. A human red blood cell is about seven microns in diameter. A human hair is approximately 100 microns in diameter. With the slight magnification of laparoscopy (let's say 4X), one could reasonably expect to see something as small as 25 microns across, or about one quarter the diameter of a human hair.

So how do researchers look for IME? They look for what appears to be normal peritoneum (the saran wrap-like lining of the pelvic cavity upon which endometriosis occurs), then they take a biopsy from this normal-appearing peritoneum and look at it under the microscope to see if the glands and stroma which are necessary for the diagnosis of endometriosis are present. If such apparently normal appearing peritoneum is found to have microscopic evidence of endometriosis, then this would mean that IME was present on that "normal" peritoneum.

Well, the ability to see something depends in part on its size. So, how big (or small) is a gland and stroma complex of endometriosis which "cannot" be seen? Are there deposits of endometriosis on the peritoneal surface smaller than 25 microns across (and therefore possibly "invisible")? The answer to this question comes from one of the studies of microscopic endometriosis from Belgium [2]. These authors found that undetected "microscopic" endometriosis found in apparently visually normal peritoneum ranged in size from 88 to 720 microns in greatest dimension, with the average diameter being 313 microns. This is well within the range of human visual acuity! To state these measurements in other terms, 300 microns is 0.3 of a millimeter, and this is three times the diameter of a human hair!

So, if researchers from Belgium confirmed that visually normal peritoneum may contain undetected small deposits of endometriosis ranging up to almost a millimeter in size on its surface, doesn't this prove that microscopic endometriosis exists? The answer is "No". The results of a study on the existence of IME are dependent upon the magnification used to examine the peritoneum and also the criteria by which normal peritoneum is judged. During laparoscopy, magnification is greater when the lens of the laparoscope is closer to the surface being examined.

Let's examine the studies on "microscopic endometriosis" with these two important factors in mind: magnification of "normal peritoneum" at surgery and the criteria used to select normal peritoneum. If higher magnification is used to select "normal peritoneum" and if stricter criteria of what constitutes "normal peritoneum" are used, will this affect the results of a study of IME? Many students of endometriosis might point to a study from Leuven, Belgium, published in 1984 [3] as the first article on IME, since the lesions were illustrated by electron photomicrographs. However, when one reads the article closely, it is clear the surgeons saw something on the peritoneal surface and then took biopsies, so this paper does not satisfy the basic requirement of a study of IME since visually normal peritoneum was not being biopsied.

A paper published in 1986 actually represents the beginning of the concept of IME [4]. The study discussed previously (Vasquez et al., 1984) has been mischaracterized by others as being a study of visually normal peritoneum, where in fact it was a study of small peritoneal abnormalities that were biopsied. Anyway, Murphy et al (1986) investigated 20 patients with endometriosis who were undergoing treatment of their disease at laparotomy. Laparotomy is where the abdomen is opened with a scalpel and surgeons are operating "at arm's length" deep inside the incision. Getting light into the "hole" can be a problem. There is no mention of any magnification being used, and there is no mention of any criteria being used to select visually normal peritoneum other than the consensus of the operating surgeons. These authors found unsuspected "invisible microscopic endometriosis" in five of the 20 patients (25%). Most of their patients, therefore, did not have IME.

But clearly this study was not the final answer on IME. The patients were operated on in an era when all of the possible visual appearances of endometriosis were not widely known. Apparently no magnification was used, and no criteria of normal peritoneum were used. Therefore, this exploratory study does not achieve high validation as a study of IME as required above. Nonetheless, the commonly accepted finding of a 25% incidence of IME is interesting since this is the highest incidence of IME ever reported in the literature. Keep this in mind. I read this 1986 paper and realized that it didn't necessarily represent Ultimate Truth on IME. I was already keenly aware of the many subtle visual appearances of endometriosis and of the value of the magnification of the peritoneum gained by laparoscopy. I was also aware that the many subtle manifestations of endometriosis could be missed not only because of lack of magnification, but also because of the lack of awareness of their existence.

People often see what they expect to see. This applies to surgeons as well. So it seemed possible that surgeons or researchers could look at endometriosis but not see it! (It still happens all the time).

I determined that in a study of IME it was important not only to look at the peritoneal surface with some magnification gained by laparoscopy (since that was also becoming the modern way of diagnosing and treating the disease), but that it was also important to "know" what normal peritoneum looked like so it could be selected for biopsy (since biopsy of visually normal peritoneum is one of the requirements of a study of IME).

Magnification at laparoscopy is increased as the tip of the laparoscope approaches the peritoneal surface, and I realized that the maximum magnification could be attained by "near-contact laparoscopy," where the tip of the laparoscope was almost in touch with the peritoneal surface, certainly less than one cm away. But I also needed a set of criteria for what constituted visually normal peritoneum.

From my knowledge of the subtle appearances possible with endometriosis, I devised a simple set of criteria of what constituted visually normal peritoneum and which could possibly effectively exclude subtle endometriosis or possibly even exclude IME in biopsies of what I thought to be normal peritoneum. I then performed two studies where I biopsied peritoneum which was judged to be normal by these criteria and sent them to the pathologist for microscopic study [5, 6]. I found only one patient out of 69 (0.014%) had a glandular structure (with no surrounding stroma) which was not apparent in visually normal peritoneum. Since no stroma was present, this cannot be accepted as true endometriosis, so the incidence of IME in my studies was actually zero. Mine were the first studies to use laparoscopic magnification by near-contact laparoscopy and specific criteria for what constituted normal peritoneum in patients with endometriosis. I also studied visually normal peritoneum from patients without endometriosis and found no IME in any specimen.

While I was doing my first study, researchers from Australia incidentally mentioned an incidence of IME of zero in 10 patients studied by laparoscopy, although no specific criteria were used for selecting normal peritoneum [7]. Other studies on IME using laparoscopy followed and the importance of laparoscopic magnification to select visually normal peritoneum was confirmed. The paper by Nisolle et al (1990), which was already mentioned above, described a technique where the laparoscope was four to five cm away from the peritoneal surface, but no criteria for visually normal peritoneum were used. They found that between six percent and 13 percent of their patients had IME, although the size of the lesions was actually rather large, as previously indicated.

Then why couldn't these lesions be seen? I think the answer lies with how far the laparoscope was from the peritoneal surface (four to five cm as compared to than less than one cm with near-contact laparoscopy) and because no criteria were used for what represented visually normal peritoneum. Nonetheless, their incidence of IME was much less than the original estimate of 25% by Murphy et al obtained at laparotomy.

Nezhat et al (1991) found one example of IME among 76 patients with endometriosis, an incidence of 1.3%. The viewing laparoscope was one to two cm from the peritoneal surface, and apparently criteria for selecting visually normal peritoneum were used [8]. Balasch et al (1996) found three examples of IME among 53 patients with visually normal peritoneum, an incidence of about six percent. The distance of the laparoscope from the pelvic peritoneum was four to five cm. Criteria for selecting visually normal peritoneum were used [9]. Walter et al (2001) studied peritoneum, thought to be normal, by laparoscopic examination in seven patients without endometriosis and found that none of the specimens harbored IME [10]. While this is somewhat similar to other studies of IME, it differs in that normal peritoneum was obtained from patients without endometriosis elsewhere in the pelvis. It could be argued that since these seven patients did not have endometriosis, one would expect normal biopsy results from visually normal peritoneum.

The evidence from studies of IME is clear: the occurrence of IME is directly and more strongly related to the power of magnification used to select normal peritoneum for biopsy. The occurrence of IME is less strongly related to the application of criteria of what represents normal peritoneum.

When the pelvis is examined at laparotomy at arm's length and with no criteria, the incidence of IME is 25 percent. When the pelvis is examined at medium range by laparoscopy, the incidence is about 10 percent. When the pelvis is examined at very close range and strict criteria are used, the incidence of IME approaches 0 percent.

Regardless of the rate of IME, no one has ever shown that any tiny area of endometriosis which might be missed because of its small size will ever get larger or be important with respect to pain or infertility. It is reassuring that this low rate of occurrence of IME has been confirmed by several different authors in many dozens of patients, since this makes the results more believable. This should be reassuring to patients with endometriosis.

This evidence can be used by patients in at least two ways. First, patients can be optimistic that surgery with an experienced endometriosis surgeon will essentially eliminate the question of IME. Second, if a surgeon suggests to a patient that IME is a big postoperative problem and will be a cause of recurrent disease, or is a reason for postoperative medical therapy, this will indicate that such a surgeon is not up-to-date on endometriosis or surgical treatment of the disease and may not be the best choice for that patient.

References

1. Newman, M. (1970). In L. A. Levin et al. (Eds.), Adler's Physiology of the Eye (pp 561-583). Mosby: St. Louis.
2. Nisolle, M., Paindaveine, B., Bourdon, A., Berliere, M., Casanas-Roux, F., & Donnez, J. (1990). Histologic study of peritoneal endometriosis in infertile women. Ferility and Stertility, 5, 984-988.
3. Vasquez, G., Cornillie, F., & Brosens, I. A. (1984). Peritoneal endometriosis: scanning electron microscopy and histology of minimal pelvic endometriotic lesions. Fertility and Sterility, 42, 696-703.
4. Murphy, A. A., Green, W. R., Bobbie, D., Cruz, Z. C., & Rock, J. A. (1986). Unsuspected endometriosis documented by scanning electron microscopy in visually normal peritoneum. Fertility and Sterility, 46, 522-524.
5. Redwine, D. B. (1988). Is "microscopic" peritoneal endometriosis invisible? Fertility and Sterility, 50, 665-666.
6. Redwine, D. B., & Ycom, L. (1990). A serial section study of visually normal peritoneum in patients with endometriosis. Fertility and Sterility, 54, 648-651.
7. Jansen, R. P. S., & Russell P. (1986). Nonpigmented endometriosis: Clinical, laparoscopic, and pathologic definition. American Journal Obstetrics Gynecology, 155, 1154-1159.
8. Nezhat, F., Allan, C. J., Nezhat, C., & Martin, D. C. (1991). Nonvisualized endometriosis at laparoscopy International Journal of Fertility, 36, 340-343.
9. Balasch, J., Creus, M., Fabregues, F., Carmona, F., Ordi, J., Martinez-Roman, S., & Vanrell, J. A. (1996). Visible and non-visible endometriosis at laparoscopy in fertile and infertile women and in patients with chronic pelvic pain: a prospective study. Human Reproduction, 11, 387-391.
10. Walter, A. J., Hentz, J. G., Magtibay, P. M., Cornella, J. L., & Magrina, J. F. (2001). Endometriosis: Correlation between histologic and visual findings at laparoscopy. American Journal of Obstetrics and Gynecology, 184,, 1407-1413.