Endometriosis usually occurs in the pelvis, but rare cases have been reported in other areas around the body. It has even been reported to occur in males. Even if endometriosis exists far away from the pelvis, it still shares similarities with pelvic disease. When symptomatic endometriosis occurs elsewhere in the body, it usually draws attention to itself with pain. Like pelvic pain with endometriosis, the pain may be present all month long and may be aggravated during the menstrual flow or by movement of nearby structures. Paying attention to these generic similarities of endometriosis can help clinicians think of the diagnosis in different locations around the body, including the diaphragm.

The diaphragm is a large, fairly thin muscle which looks like an overturned cereal bowl. It is only about one-quarter to one-eighth inch thick. It separates the chest cavity from the abdominal cavity and is attached around the bottom of the rib cage. There is a right half and a left half. The rear part of the right diaphragm is hidden behind the liver. The diaphragm is supplied by the phrenic nerve, and it is the most important muscle used in breathing. The diaphragm is of interest primarily to general surgeons, who are called upon to repair ruptures of the diaphragm resulting from trauma, or to repair hiatal hernias, among other things. Gynecologists are aware of the diaphragm, but since it can't get pregnant, they generally pay it no mind.

However, endometriosis can occasionally affect the diaphragm, so gynecologists can't entirely ignore its existence. We have seen 5 patients with endometriosis of the diaphragm (Table 1) , which is apparently the largest series in the world.

Each patient had a very similar history. All had undergone previous treatments for known pelvic endometriosis, but only one had received a diagnosis of diaphragmatic endometriosis at laparoscopy. There are several reasons for this. Most gynecologists have not been trained to look at the upper abdomen during a pelvic laparoscopy. After all, gynecologists are not trained to perform surgery in the upper abdomen or on the diaphragm, so it is somewhat understandable that many may not look where they can't treat. Additionally, in each patient, pelvic pain began earlier in life than did the pain related to the diaphragmatic disease, so the surgeon may have had no reason to suspect anything in the upper abdomen.

Each patient reported the gradual onset (over the course of several months to several years) of right shoulder pain associated with the menstrual flow. The pain was described as coming from deep within the shoulder, and in some would seem to radiate from the right chest. The pain would spread down the upper right arm and up the right neck in some, the result of muscular tension reacting to pain. One patient described the pain as similar to the deep chest or shoulder pain one gets with running too hard. Several patients described difficulty sleeping in certain positions, being occasionally awakened by the pain. One patient couldn't sleep lying down because of severe pain and found partial relief only by sleeping upright in a chair with ice packs on her shoulder. Most of the patients eventually began to experience a low level of right shoulder pain throughout the month.

One patient had undergone several rounds of medical therapy directed at her diaphragmatic disease, including danazol and a GnRH agonist. She had also had an attempt at laparoscopic fulguration of a lesion on the diaphragm, but still her pain continued. None of the others had received diagnosis or treatment of diaphragmatic endometriosis.

When Dr. Sharpe and I performed surgery on these patients for their diaphragmatic disease, we began laparoscopy in the usual way with a laparoscope inserted through the umbilicus (belly button). In all patients, we could see a small lesion of possible endometriosis on the visible portion of the right diaphragm in front of the liver. However, we were suspicious that such small lesions were causing all these symptoms, so we put a second laparoscope into the abdomen higher up near the bottom of the right rib cage. This allowed us to see behind the liver. Lo and behold! Each patient had an ugly area of invasive disease nestled far behind the liver where it couldn't be seen.

In our first patient, we tried to resect the diseased area through the laparoscope, but this didn't work at all. The liver was in the way, and we couldn't get to it. Even if we had been able to resect this area, we would have run into a second problem: in each patient the endometriosis extended all the way through the diaphragm, so we would have had to remove a full-thickness portion of the diaphragm and repair it laparoscopically. This would have taken forever, and would have risked injury to the exposed lung.

We had to open the first patient to complete her surgery, and we have had to open the other four also. At open surgery, the liver can be retracted out of the way, a portion of the diaphragm removed, and the diaphragm repaired securely with suture.

All five patients seemed to have relief of right shoulder pain following surgery, and none has mentioned any difficulty breathing. Post-operative chest X-rays and fluoroscopy on our first two patients were normal, and we haven't felt compelled to check the others for this reason. Our first patient returned with endometriosis of the left diaphragm several months later (we were so focused on the right-sided nature of her chest pain that we didn't get a good look at the left diaphragm!).

We have learned several things about diaphragmatic endometriosis from all of this.

1. Suspect diaphragmatic endometriosis on the basis of right or left shoulder pain which may be associated with the menstrual flow.
2. Use the laparoscope to look at both sides of the diaphragm as well as possible. Even if the surgeon sees a lesion of endometriosis he doesn't feel capable of removing at that time, at least the diagnosis will have been made.
3. If a small diaphragmatic lesion is seen, it may just be a "sentinel" lesion which is only the edge of the iceberg. The main bulk of the disease may be hidden behind the liver and the correct diagnosis may require a second laparoscope inserted higher up in the abdomen.
4. Laser vaporization or electroco-agulation of a small visible lesion may be treating only a fraction of the disease. Since disease which is symptomatic seems always to involve the full thickness of the diaphragm, it would be necessary to burn all the way through the diaphragm with laser or electrocoagulation in order to treat disease completely. This would risk damage to the lungs or heart, and the surgeon would still be left with a hole in the diaphragm which would need to be closed.
5. The laparoscope is inadequate for treating diaphragmatic endometriosis. A laparotomy is necessary.

Dr. Sharpe presented this material at the 22nd Annual Meeting of the American Association of Gynecologic Laparoscopists in San Francisco in November 1993 and is preparing a scientific publication on this topic.

SUGGESTED READING:

• Redwine, DB. Laparoscopic en bloc resection for treatment of the obliterated cul de sac in endometriosis. Journal of Reproductive Medicine, 1992; 37: 695-8. Rationale for and technique of treatment of obliteration of the cul de sac.
• Redwine, DB. Laparoscopic excision of endometriosis with 3 mm scissors: Comparison of operating times between sharp excision and electro-excision. The Journal of the American Association of Gynecologic Laparoscopists, 1993; 1:24-30.

  Excision of endometriosis with electrosurgery is significantly faster than excision with mechanical scissors.

• Redwine, DB. Endometriosis persisting after castration: Clinical characteristics and results of surgical management. Obstetrics and Gynecology 1994; 83:(March). In press.

  A prospective study of 75 patients with endometriosis persisting after removal of the ovaries, with results of post-op pain questionnaires. Patients with persistent endometriosis after castration are more likely to have intestinal disease, and are more likely to have obliteration of the cul de sac.

• Vercellini, P, et al. A gonadotropin-releasing hormone agonist versus a low-dose oral contraceptive for pelvic pain associated with endometriosis. Fertility and Sterility, 1993; 60:75-9.

Although patients achieved pain relief with both types of treatment, pain returned shortly after treatment ceased. With this publication, medical therapy seems to have come full circle (or is it a downward spiral?): abandonment of surgical excision in the 1960's for birth control pill therapy; abandonment of birth control pill therapy in the 1970's for danazol therapy; abandonment of danazol therapy in the 1980's and 90's for GnRH agonist therapy; possible abandonment of GnRH agonist therapy in the 1990's for birth control pill therapy. What's wrong with this picture?