- Email: [email protected]
Breast Cancer During Pregnancy and Lactation
Breast Cancer: Strategies for the 1990s 11
0039-6109/90 $0.00
+ .20
Breast Cancer During Pregnancy and lactation
Herbert C. Hoover, Jr, MD*
Breast cancer, the most common cancer, and in many geographic areas, the most frequent cause of cancer deaths in women, is also the cancer most likely to be seen in pregnancy or lactation. 11 The occurrence of breast cancer during pregnancy and lactation presents a challenging clinical problem, which is compounded by its infrequent occurrence and by the emotional issues involved. One of life's most joyous experiences can become a nightmare, often unnecessarily. Because most physicians deal with this clinical situation so infrequently, anecdotal experiences can easily propagate misinformation. We will discuss the principal questions related to the diagnosis, therapy, and prognosis of cancer of the breast in pregnancy and lactation.
HISTORICAL ASPECTS Until Halsted's pioneering work, breast cancer diagnosed during pregnancy was considered incurable. In 1896, Halsted performed a radical mastectomy on a lactating woman who was reported to be alive and well more than 30 years later. This long-term follow-up was provided by Bloodgood as a part of a report by Kilgore 26 describing a 70% 5-year survival rate in lactating women with node-negative breast cancer and a 11.5% 5year survival rate in patients with nodal involvement. These early writings rejected the idea that milk fistula would uniformly result from a breast biopsy during pregnancy or lactation. Bloodgood also mentioned that general anesthesia apparently had no harmful effect on the fetus and intimated that breast cancer during pregnancy should, in fact, be treated much as in the nonpregnant woman. However, an attitude of gloom was propagated by the 1943 publication by Haagensen and Stoutl6 stating that breast carcinoma developing during pregnancy and lactation was categori-
*Associate
Professor of Surgery, flarvard Medical School; and Chief, Surgical Oncology Research, Massachusetts General Hospital, Boston, Massachusetts
Surgical Clinics of North America-Va!' 70, No.5, October 1990
1151
1152
HERBERT
C.
HOOVER, JR
cally inoperable. This view was based on their experience with only 20 patients (19 of whom had positive nodes), all of whom had died of their breast cancer. Haagensen reversed his view later14 after more extensive experience, but the pessimistic attitude was widely prevalent and remains so today. However, studies in the past 30 years have demonstrated convincingly that the poor results are attributable to delay in diagnosis and reluctance to treat patients aggressively during pregnancy rather than to a detrimental effect of pregnancy or lactation itself.
EPIDEMIOLOGY Breast cancer diagnosed during pregnancy or lactation is relatively rare. Indeed, because the incidence of breast cancer peaks between the ages of 50 and 55, one would not expect to see the frequent coexistence of pregnancy and breast cancer. White and coworkers44 reviewed 1296 such cases, including 60 of their own, among 45,881 breast carcinoma cases, an incidence of approximately 2.8%. Of 300,860 pregnancies reviewed, there were only 93 breast cancers documented, or 3 per 10,000 pregnancies. Wallack and associates,43 in a large literature review, put the incidence even lower, with breast cancer associated with pregnancy or lactation in only 1%-2% of cancer cases. In most series, the average age of patients has been 34 to 35 years. With the current trend toward postponing childbearing to the mid to late 30s, it is expected that the incidence of breast cancer diagnosed during pregnancy and lactation will increase.
ETIOLOGY Even though pregnancy and cancer are the only two biologic conditions in which antigenic tissue is known to be tolerated by a seemingly intact immune system,13 there is no evidence to implicate pregnancy or lactation in either the etiology or the progression of breast cancer. There is evidence for depressed cell-mediated immunity, such as an abnormal lymphocyte response to phytohemagglutinin in vitro and impairment of in vivo responses to common recall antigens used in testing for delayed cutaneous hypersensitivity.36 Additionally, some of the pregnancy-induced hormonal changes, such as the elevation of circulating corticosteroids, could be immunosuppressive. 28 Theoretically, these changes could promote more rapid tumor growth but are unlikely to play any significant etiologic role. According to our knowledge of the biology of breast cancer, a cancer manifesting itself during pregnancy begins months or even years before conception. Of course, the etiology of breast cancer remains unknown, but there appears to be nothing unique about breast cancer in pregnancy.
CLINICAL PRESENTATION Delay in diagnosis appears to be the primary, and perhaps the only, reason for the generally worse prognosis overall for all patients with breast
BREAST CANCER DURING PREGNANCY AND LACTATION
1153
cancer diagnosed during pregnancy or lactation. 1. 5. 6. 15. 22. 27. 31. 34. 37. 38 The duration of symptoms before diagnosis averaged 5 to 15 months in some series, which is considerably greater than in series of nonpregnant patients. This delay can be attributed to physician neglect as readily as to the patient's. During pregnancy, the breasts become tense and often have a multinodular consistency. A discrete mass can be difficult to feel or can be obscured as the breasts hypertrophy. Because of the pregnancy, masses are often ignored even when noticed by patients or physicians, the plan being to evaluate them further if they persist after delivery. Mammography often is not recommended during pregnancy because of the radiation exposure to the fetus and the poor imaging obtained in the tense, hyperplastic breast tissue. The latter is a significant problem, but the former need be of no serious concern with modern techniques and proper abdominal and pelvic shielding. As a consequence of these delays, virtually all articles on patients with breast cancer during pregnancy report a more advanced stage than in nonpregnancy series. Ribeiro and Palmer37 reported on 88 patients with breast cancer concurrent with pregnancy, 19 of whom had inoperable, faradvanced tumors. Of the 69 patients with operable disease, 89% were found to have positive nodes in the ipsilateral axilla. Holleb and Farrow2 reported 72% with positive nodes among 117 patients. These figures are far in excess of the 40% to 50% node-positive rate reported in most series of breast cancer in nonpregnant women. 19 DIAGNOSIS Primary care physicians, especially obstetrician-gynecologists, play a pivotal role in the diagnosis of breast cancer during pregnancy and lactation. It is especially important that a careful breast examination be performed at the initial obstetric visit before the breasts become engorged and difficult to examine carefully. As mentioned earlier, studies suggest that both patients and their physicians often fail to recognize the potentially serious nature of breast masses during pregnancy. There should be nothing special about the diagnostic evaluation of pregnant or lactating patients with breast masses, with the exception that xerography tends not to be as helpful because of the parenchymal changes associated with gestation. Increased water density of the breasts reduces the discriminatory capacity of mammography.21 Radiation exposure to the fetus should be negligible with proper shielding of the abdomen. In 368 pregnant women undergoing mammography, one series reported no apparent damage to the fetus. 28 Nevertheless, in a patient with a discrete palpable mass, a mammogram would rarely influence therapy, so it need not be obtained. Any dominant mass should be evaluated promptly in pregnant or lactating patients. This evaluation normally should begin with a fine-needle aspiration to distinguish cystic from solid lesions. Any new solid mass requires serious consideration of a biopsy, which can almost always be performed under local anesthesia.
1154
HERBERT
C.
HOOVER, JR
The breast masses biopsied during pregnancy show the same spectrum of histopathology found in nonpregnant patients. Additionally, the likelihood of finding a carcinoma appears to be similar to that in breast masses biopsied in the nonpregnant population. 4 Byrd and associates 4 found that 22% of breast biopsies in pregnancy showed malignancy compared with 19% in all patients overall. Although the earlier literature 44 implied that inflammatory carcinoma was seen more frequently during pregnancy, modern series do not confirm this. For biopsies done during lactation, some surgeons prefer to suppress lactation preoperatively using bromocriptine, although there is no body of evidence to support this measure. The risk of milk fistulas is very low for peripheral lesions, although such fistulas occasionally cause a problem in deep, central lesions.
HORMONE RECEPTORS Steroid hormone receptors may be difficult to demonstrate in breast cancer tissue during pregnancy unless special methods are utilized. 12. 25, 39, 40 Pregnancy may depress levels of estrogen and progesterone receptors detectable in breast cancer cytosol fractions, resulting in false-negative studies because high levels of estrogens circulating in pregnant patients cause receptor translocation into the nucleus and occupy all the cytoplasmic receptors, leaving none available for assay. Therefore, unbound estrogen should first be removed from breast cancer specimens by treatment of the cytosol with dextran-charcoal. Exchange assays can then be performed to detect occupied estrogen receptors within the cytosol and nuclei. However, to my knowledge, there are no data available on estrogen receptors in breast cancer during pregnancy that show any meaningful prognostic significance.
ANESTHETIC CONSIDERATIONS There is no evidence that a breast biopsy poses any significant risk to either the mother or the fetus even if general anesthesia is required, assuming that proper precautions are used. Byrd and associates 4 report only one fetal loss in 134 breast biopsies on pregnant patients under general anesthesia, and this loss was in an older patient not known to be pregnant at the time of the operation, which therefore was done without the usual precautions. There are some unique management considerations because of the phy~iologic changes of pregnancy and the physical effects of an enlarged, gravid uterus. The special problems posed by general anesthesia in pregnant patients are reviewed in several recent publications, 2, 7, 23, 29, 32, 33 which have been extensively discussed by Wallack and coworkers 43 and will be summarized here. A number of cardiovascular and hematologic alterations during pregnancy necessitate anesthetic adjustments. Blood volume increases by ap-
BREAST CANCER DURING PREGNANCY AND LACTATION
1155
proximately 35%, but this is primarily in the plasma rather than red cell volume, giving a physiologic dilutional anemia. By term, the heart rate increases 10 to 15 beats per minute. Along with the increase in blood volume and heart rate is a 30% to 50% increase in cardiac output. There is a slight decrease in the systemic vascular resistance as well. The lower hemoglobin content leads to a decrease in oxygen-carrying capacity, and the task of maintaining oxygen saturation and cardiac output becomes critical. A high inspired oxygen concentration (at least 50%) and the avoidance of agents that depress cardiac output are advised. Pregnancy also alters coagulation. The platelet count increases nearly 50%, as does the fibrinogen level. The resulting hypercoagulable state can increase the risk of thromboembolism. Serum cholinesterase activity is decreased by approximately 20%, prolonging the metabolism of succinylcholine. With the patient supine, a gravid uterus may interfere with blood flow because of pressure on the distal aorta and inferior vena cava, making positioning during anesthesia of great importance. Tilting the operating table slightly to the left and putting a cushion under the patient's right hip can help to displace the uterus to the left and away from major vascular structures. A large gravid uterus raises the level of the diaphragm significantly, thus compromising lung volumes. Decreased functional residual capacity along with a pregnancy-associated increase in the breathing rate can lead to a more rapid build-up of anesthetic vapor within the aveoli and thus a more rapid induction of anesthesia. Forced hyperventilation and maternal respiratory alkalosis can affect the fetus adversely by causing progressive fetal acidosis. The anesthetist must exercise great care in maintaining a "normal for pregnancy" ventilatory balance. Because of the decreased functional residual capacity and diminished ability to store oxygen, coupled with the increased oxygen consumption in pregnancy, the pregnant patient is predisposed to hypoxemia during even brief apneic episodes. Preoxygenation for several minutes should always precede endotracheal intubation. Intubation itself should be carefully performed because of the hypervascularity in the respiratory tract mucosa in most pregnant women. Gastric emptying time is prolonged in pregnancy, being as much as 60% above normal as the patient approaches term. The pressure on the stomach makes reflux much more likely as well. Hypergastrinemia not uncommonly leads to more acidic gastric secretions. Therefore, pregnant patients require special intubation techniques to avoid aspiration of this very damaging acidic material. There are no published studies to confirm any deleterious effects of anesthetic agents on the developing fetus. There are many studies of the offspring of narcotic addicts, which have shown no long-term effect of morphine, heroin, or methadone other than low birth weight. 3 Fetal monitoring should be used whenever possible so that early distress can be recognized and appropriate adjustments made. Otherwise, it is felt that general anesthesia poses no real hazard as long as careful attention is given to the unique requirements of the situation.
1156
HERBERT
C. HOOVER, JR
STAGING EVALUATION Once the diagnosis of breast cancer is made, staging of the disease is critical before therapeutic decisions are made. Because all of the staging procedures, with the exception of blood studies, involve radiation exposure, there is considerable controversy concerning the extent of evaluation that should be performed in breast cancer patients who are pregnant. Liver function tests, calcium measurements, and carcinoembryonic antigen assays can give some useful information but do not permit a definitive diagnosis or localization of metastatic disease. A thorough understanding of the radiation risk to the fetus is helpful to the clinician deciding on the diagnostic tests to pursue. Hall 17 has published much useful information in this regard. Radiation delivered during the preimplantation phase, from fertilization to embryonic implantation in the uterine wall (days 0 to 10), can result in the death of the embryo. As little as 10 rad can be lethal in the mouse. He did find that those embryos that survived irradiation during this period appeared to develop normally. Irradiation during organogenesis (days 11 to 56) can lead to serious developmental malformations. In humans, the most frequently noted defect is microcephaly. Again, doses as low as 10 rad may be sufficient to produce a measurable increase in birth defects, and doses exceeding 100 rad are felt to produce congenital abnormalities in 100% of cases on the basis of the atomic bomb experience. 27 • 30 Mter day 57, during the stage when additional fetal growth is occurring, larger doses of radiation are required to cause significant effects. After weeks 25 to 30, the fetus is far more resistant to the damaging effects of radiation. 8 The majority of the reports on the effects of ionizing radiation on the developing organs are extrapolations from animals to man, with doses that are much higher than are pertinent to the radiologic evaluation for metastases in breast cancer, although they are more relevant in considering therapeutic breast irradiation. However, there are insufficient data in the literature to quantify the long-term risk to the fetus from the low radiation doses involved in most scanning procedures. For example, the radiation dose to the fetus from a bone scan is approximately 0.1 rad. Even though we conclude that any radiologic staging procedure is probably safe in a pregnant patient, we cannot say this with certainty. Therefore, bone scans generally are not recommended unless significant therapeutic decisions would be based on their outcome. This is similar to our policy in stage I and II nonpregnant patients with carcinoma of the breast, where the diagnostic yield in an asymptomatic patient is not felt to be high enough to justify the expense on a routine basis. Harbert18 reviewed a large number of studies showing a 3% true-positive yield from bone scans in 533 patients with stage I breast cancer, 7% in 696 patients with stage II disease, and 25% in 278 with stage III disease. When pregnant patients with stage I or II disease have skeletal pain, a bone scan is preferred to radiographic evaluation, as the radiation exposure from the scan would be considerably less than that of a skeletal survey. In clinical stage III disease, where a significantly higher positive scan rate would be expected, one would still prefer to defer the bone scan until the later stages of pregnancy
BREAST CANCER DURING PREGNANCY AND LACTATION
1157
unless one would decide to treat with systemic therapy at that stage of the pregnancy if bone metastases were demonstrated. This raises complex issues as to which patients should be considered for termination of pregnancy, a topic that will be discussed in the therapeutic section. Carcinogenesis from radiation exposure is another possible risk to the fetus. Although the association between fetal radiation exposure and cancer was not clearly shown in the atomic bomb experience, a recent study of twins showed that those who both developed cancer or leukemia were twice as likely to have been exposed to x-rays in utero as were control sets of twins.20 Rather than take chances, it is best to avoid any radiation exposure during pregnancy unless the data to be gained are critically important. This would not apply to routine chest radiographs, which can be done with essentially no risk to the fetus with proper abdominal and pelvic shielding. However, in stage I breast cancer, the chance of a positive chest study is sufficiently low that one could justify deferring the chest film until delivery of the fetus. There is no role for radionuclide brain or liver scans in these patients. Any patient with CNS symptoms would be better served by a CT scan, and the liver can be evaluated with ultrasound at no risk to the fetus in patients with liver function abnormalities. TREATMENT Although Haagensen and Stout concluded in 1943 that the prognosis was so poor that mastectomy was not justified in patients with breast cancer during pregnancy or lactation,16 Haagensen did not persist with that point of view for long. He soon realized that his initial experience with 20 patients, who all died, was not representative of the prognosis. 17 Virtually all reports since then have stressed that the surgical treatment should be essentially that of the nonpregnant patient in all stages of disease. More controversial has been whether termination of pregnancy should be advised. Obviously, the therapeutic decisions will be influenced by the stage of pregnancy and the stage of disease, so we must consider the treatment plan accordingly. Because the risk of spontaneous abortion during mastectomy is low, 4 mastectomy or other breast procedures should be performed promptly upon diagnosis at any stage of pregnancy. A mastectomy allows a normal pregnancy to continue with little risk to the mother or the fetus. This is especially true in that no modem series suggests that abortion favorably affects the course of patients with breast cancer. 9, 10 Therefore, any decision to terminate the pregnancy should be based on considerations other than abortion's influence on the progression of the disease. Termination of pregnancy has no role in the management of patients with stage I or II breast cancer. Likewise, there is no substantial evidence that oophorectomy influence's the course of breast cancer during pregnancy and lactation. With the trend toward breast conservation and radiation, some pregnant women with early stages of breast cancer will want to consider that option. It should be strongly discouraged except in cases where radiation therapy could reasonably be withheld until after delivery, because the
1158
HERBERT
C.
HOOVER, JR
standard technique of whole-breast irradiation with a boost dose to the tumor bed would produce a dose scattered to the fetus that would be unacceptably high. 28 There are two sources of radiation dosage outside the beam: scatter within the body, and leakage from the head of the machine. With proper lead blocks, the leakage rate can be reduced to a low range, less than 2 rad. However, internal scatter is dependent on field size, photon energy, and proximity to the field edge; and shielding cannot modify this scatter. Early in pregnancy, when the distance between the breast and uterus is great, the internal scatter dose may be only 10 to 20 rad for a total course of treatment. However, when the top of the uterus approaches the xiphoid, fetal doses of as much as 100 rad or even more may be given. 28 Obviously, special treatment arrangements can be made for women who insist on breast conservation. This could involve primary excision using a small boost of irradiation to the area with delivery of full breast irradiation after the patient gives birth. The patient would have to be aware of the potential risk to the fetus even with this low dose. Every effort should be made to discourage such an approach. A more reasonable alternative would be simple surgical removal of the primary tumor and axillary lymph nodes with observation until delivery, when further therapy could be given. Adjuvant chemotherapy has become standard for node-positive premenopausal women with breast cancer and is now considered in nodenegative patients. Because it is felt that therapeutic delay may diminish the benefits of adjuvant therapy, the decision to initiate chemotherapy during pregnancy is one that requires serious consideration. Unfortunately, the risks associated with chemotherapy during pregnancy are not clear. Most of the hard data have been developed in laboratory animals, in which the period of greatest risk is the initial portion of the pregnancy. Schapira and Chudley41 found the incidence of teratogenicity of chemotherapeutic agents given to humans in the first trimester of pregnancy to be only 12.7%, while Sweet and Kinzie 42 reported an 11.5% incidence. Neither investigation reported evidence for teratogenicity from the administration of chemotherapeutic agents in the second and third trimesters of pregnancy. Most of the data involve single drug use only, and little is known about the combination regimens usually given to breast cancer patients. Additionally, there are no good studies on the long-term consequences of fetal exposure to chemotherapeutic agents. In general, the use of any chemotherapeutic agents during the first trimester of pregnancy should be discouraged, but their use during the second and third trimester probably induces very few abnormalities. Nodenegative women should not be treated with chemotherapy at any stage of their pregnancy until its efficacy is more clearly defined in nonpregnant women. The following recommendations for therapy are suggested. Stages I and II (Operable) Breast Cancer Modified radical mastectomy is the treatment of choice in the pregnant patient. A second choice would be total tumor excision and axillary dissection to be followed by whole-breast irradiation after delivery. In patients with positive axillary lymph nodes, chemotherapy should certainly
1159
BREAST CANCER DURING PREGNANCY AND LACTATION
Table 1. Treatment Results in Operable Breast Cancer During Pregnancy or Lactation 5-YEAR SURVIVAL NO. OF SOURCE
PATIENTS
Peters (1962)34 Holleb & Farrow (1962)22 Rissanen (1968)38. Donegan (1978)10 Ribeiro et al (1986)38 Total
60 119 33 24 121 357
Positive Nodes (%)
Negative Nodes (%)
23 17 36 31 25 26%*
62 65 80 86 79 74%*
OVERALL(%)
37 30 43 48 37 39%*
*Mean.
be delayed until the second trimester of pregnancy. My personal recommendation would be to delay all chemotherapy in node-positive patients until after delivery. With determination of fetal age and maturity, consideration should be given to an early cesarean section to minimize the delay in beginning chemotherapy. Stage III (Locally Advanced) and Stage IV (Distant Metastases) Breast Cancer In patients with locally advanced or metastatic carcinoma diagnosed early in the pregnancy, for whom both chemotherapy and radiation therapy would be recommended if the woman were not pregnant, serious consideration must be given to termination of the pregnancy. In the later stages of pregnancy, nonsurgical therapy could be qelayed until after delivery, but one must weigh the risks of observing an advanced breast cancer for a long period of time. As mentioned earlier, chemotherapy can be used with little known risk during the second and third trimester, but, unfortunately, good data for standard breast chemotherapy combinations are lacking. If radiation therapy is imperative at any stage of pregnancy, termination of pregnancy may be the most reasonable alternative. The mother's life expectancy is often limited, and damage to the fetus is a significant possibility. There are no "right" answers to many of the therapeutic dilemmas encountered in patients developing breast cancer during pregnancy. A thorough discussion of all the risks involved with any approach must be
Table 2. Stages of Breast Cancer at Diagnosis in Patients Under 40 Years Old NO. (%)
STAGE
I (negative nodes) II (positive nodes)
III (distant disease)
Pregnant (N = 19)
Nonpregnant (N = 155)
4 (21) 14 (74) 1 (5)
84 (54) 57 (37) 14 (9)
From Nugent P, O'Connell TX: Breast cancer and pregnancy. Arch Surg 120:1221-1224, 1985; with permission.
1160
HERBERT
C.
HOOVER, JR
presented to the patient and her family, who must be a part of these oftendifficult decisions. PROGNOSIS The rarity of breast cancer in pregnancy or lactation dictates that all reports of the entity are retrospective. Most authors report on only a few patients, making statistical analysis difficult or impossible. Also, nearly all have considered pregnancy and lactation together, assuming no differences between the two. Despite all of the shortcomings of smaU, retrospective studies, some general conclusions can be drawn. Table 1 lists several series chosen as those most likely to be representative on the basis of their size and presentation of objective data. They clearly show that pregnant or lactating women who are identified early with negative axillary lymph nodes have an expected outcome similar to that of nonpregnant women. Nugent and O'Conne1l31 have shown this fact clearly in a small series comparing the distribution of stages of disease in pregnant patients with that in women under 40 years of age who were not pregnant (Table 2). The usual trend toward a higher stage of disease at diagnosis was found in pregnant patients, with 74% having positive nodes compared with 37% of the nonpregnant women. When 5-year survival rates were compared in these groups, both by stage and overall, no difference was noted (Table 3). The challenge clearly is to impact on this disease by making the diagnosis earlier. The bulk of evidence supports the opinion that pregnancy does not worsen the prognosis except by obscuring the disease, often for many months, allowing the metastatic process to progress. We can hope that increased vigilance on the part of physicians seeing these patients early in their pregnancy can reverse this trend. OTHER CONSIDERATIONS Tumor Metastatic to the Fetus Some women with breast cancer diagnosed during pregnancy fear that cancer cells have spread to their fetus and could grow there. This has been Table 3. Five-Year Survival Rates of Pregnant and Nonpregnant Women Under 40 Years Old with Breast Cancer SUjWIVAL NO. (%) STAGE
Pregnant
Nonpregnant
I II III Overall
4/4 (100) 7/14 (50) 011 (0) 11119 (57)
59/84 (70) 27/57 (48) 1114 (7) 87/155 (56)
P' 0.57 1.00 1.00 1.00
*By Fisher's exact test From Nugent P, O'Connell TX: Breast cancer and pregnancy. Arch Surg 120:1221-1224, 1985; with permission.
BREAST CANCER DURING PREGNANCY AND LACTATION
1161
reported in melanoma and choriocarcinoma35 but, to my knowledge, not in breast cancer. Pregnancy After Treatment of Breast Cancer Many young women treated successfully for breast cancer desire further pregnancies. Some studies34, 20 have, in fact, shown an improved survival rate among breast cancer patients who become pregnant. Some authors have argued that this phenomenon relates to patient selection, in that only women without an early recurrence and with a good prognosis live to become pregnant or desire pregnancy, but Peters34 argues that improved survival among patients whose pregnancy occurred within a short interval after mastectomy is evidence against a selection process. She demonstrated an improvement in survival of 23% at 5 years and 28% at 10 years when 96 subsequent pregnant patients were compared with 96 patients of similar age and clinical stage in whom pregnancy did not occur. In 41 patients treated for primary operable carcinoma of the breast, Harvey and coworkers 20 found no detrimental effect of subsequent pregnancy even among patients with positive axillary nodes or those whose pregnancies occurred less than 2 years after mastectomy. Abortion produced no improvement in the survival rate. They concluded, as have most other authors of recent series, that there are no therapeutic grounds for recommending avoidance or termination of pregnancy among patients without evidence of recurrent disease after treatment for breast cancer. SUMMARY Breast cancer is the most frequently seen cancer in pregnancy and lactation, but the incidence is low, the disease being seen in approximately 0.03% of pregnancies. Only 1% to 2% of breast cancer overall is diagnosed during pregnancy or lactation. There is no evidence to implicate pregnancy or lactation in either the etiology or the progression of breast cancer. Careful breast examination early in the pregnancy is very important to find solid masses that require biopsy before breast engorgement hides them. Therapeutic options vary, depending on the stage of disease and the stage of the pregnancy. Opetable disease in the first 6 to 7 months of the pregnancy should be treated by mastectomy, as irradiation is contraindicated. Late in the pregnancy, a lumpectomy and axillary dissection can be done, with irradiation being delayed until after delivery. General anesthesia is safe if the usual precautions are taken to compensate for the physiologic changes induced by pregnancy. Unfortunately, delay in diagnosis is common, and 70% to 89% of patients with operable primary lesions have positive axillary lymph nodes. Late stage appears to be the only reason for the generally worse prognosis in these patients, as stage for stage, they have a course similar to that of nonpregnant patients. Adjuvant chemotherapy can be considered late in the pregnancy but should usually be delayed until after delivery. In patients with locally advanced or metastatic cancer diagnosed early in the pregnancy, for whom both chemotherapy and radiation therapy would normally be recommended, consideration must be
1162
HERBERT
C.
HOOVER, JR
given to termination of the pregnancy. There is no evidence that termination of pregnancy improves the outlook for the patients, but it does permit standard aggressive therapy in advanced disease.
REFERENCES 1. Anderson JM: Mammary cancers and pregnancy. Br Med J 1:1124, 1979 2. Attia RR, Ebeidan-Fischer JE: Gastrin: Placental, maternal and plasma cord levels: Its possible role in maternal residual gastric acidity. Abstr Am Soc Anesthesiologists 547, 1976 3. Blinick G, Jerez E, Wallach RC: Methadone maintenance, pregnancy and progeny. JAM A 255:477, 1973 4. Byrd BF, Bayer DS, Robertson JC, et al: Treatment of breast tumors associated with pregnancy and lactation. Ann Surg 155:940, 1962 5. Canter JW, Oliver GC, Zaloudek CJ: Surgical diseases of the breast during pregnancy. Clin Obstet Gynecol 26:853, 1983 6. Cheek JH: Cancer of the breast in pregnancy and lactation. Am J Surg 126:729, 1973 7. Cohen SE: Why is the pregnant patient different? Semin Anesthesia 1:73, 1982 8. Dekaban AS: Abnormalities in children exposed to X-radiation during various stages of gestation: Tentative time table of irradiation injury to the human fetus 1. J Nucl Med 9:471, 1968 9. Donegan WL: Mammary carcinoma in pregnancy. In Spratt JS, Donegan WL (eds): Major Problems in Clinical Surgery. Philadelphia, WB Saunders, 1967, p 170 10. Donegan WL: Mammary carcinoma in pregnancy. In Hall EJ (ed): Radiobiology for the Radiologist. New York, Harper & Row, 1978, p 397 11. Donegan WL: Cancer and pregnancy. CA 33:194, 1983 12. Garola RE, McGuire WL: An improved assay for nuclear estrogen receptor in experimental and human breast cancer. Cancer Res 37:333, 1977 13. Gleichner N, Siegel!: Common denominators of pregnancy and malignancy. In Gleichner N (ed): Reproductive Immunology. New York, Alan R Liss, 1981, P 339 14. Haagensen CD: The treatment and results in cancer of the breast at the Presbyterian Hospital, New York. Am J Roentgenol 62:328, 1949 15. Haagensen CD: Cancer of the breast in pregnancy and lactation. Am J Obstet Gynecol 98:141, 1967 16. Haagensen CD, Stout AP: Carcinoma of breast: Criteria of operability. Ann Surg 118:859, 1943 17. Hall EJ: Effects on the embryo and fetus. In Hall EJ (ed): Radiobiology for the Radiologist. Hagerstown, Harper & Row, 1973, P 231 18. Harbert JC: Efficacy of bone and liver scanning in malignant disease facts and options. In Nuclear Medicine Annual 1982. New York, Raven Press, 1982 19. Harris JR, Hellman S, Canellos GP, et al: Cancer of the breast. In DeVita VT, Hellman S, Rosenberg SA (eds): Cancer: Principles and Practice of Oncology, ed 2. Philadelphia, JB Lippincott, 1985, p 1119 20. Harvey EB, Borce JD, Honeyman M, et al: Prenatal X-ray exposure and childhood cancer in tWins. N Engl J Med 315:541, 1985 21. Hoeflken W, Lanyi M: Mammography. Philadelphia, WB Saunders, 1977 22. Holleb AI, Farrow JH: The relation of carcinoma of the breast in pregnancy in 283 patients. Surg Gynecol Obstet 115:65, 1962 23. Huseymeyer RP, Davenport HT: Prophylaxis for Mendelson's syndrome before elective cesarean section: A comparison of cimetidine and magnesium trisilicate mixtures of regimens. Br J Obstet Gynecol 87:565, 1980 24. Jablon S, Kato H: Childhood cancer in relation to pre-natal exposure to atomic bomb radiation. Lancet 2: 1000, 1970 25. Katzenellenbogen JA, Johnson HJ, Carlson KE: Studies on the uterine, cytoplasmic estrogen binding protein: Thermal stability and ligand dissociation rate: Ail assay for empty and filled sites by exchange. Biochemistry 12:4092, 1973 26. Kilgore AR: Tumors and tumor-like lesions of breast in association with pregnancy and
BREAST CANCER DURING PREGNANCY AND LACTATION
1163
lactation (with a note by Bloodgood JC): The treatment of tumors of the breast during pregnancy and lactation. Arch Surg 18:2079, 1929 27. King RM, Welch JS, Marten JK, et al: Carcinoma of the breast associated with pregnancy. Surg Gynecol Obstet 160:228, 1985 28. Lippman ME, Lichter AS, Danforth DN: Breast cancer occurring during pregnancy. In Diagnosis and Management of Breast Cancer. Philadelphia, WB Saunders, 1988, p 414 29. Mendelson CL: Aspiration of stomach contents into the lungs during obstetric anesthesia. Am J Obstet Gynecol 53:191, 1946 30. Miller R, Mulvihill S: Small head size after atomic irradiation. Teratology 14:355, 1976 31. Nugent P, O'Connell T: Breast cancer in pregnancy. Arch Surg 120:1221, 1985 32. Nunn JF: Applied Respiratory Physiology, ed 2. Boston, Butterworths, 1977, p 397 33. Pedersen H, Finster M: Anesthetic risks in the pregnant surgical patient. Anesthesiology 51:439, 1979 34. Peters MV: Carcinoma of the breast associated with pregnancy. Radiology 78:58, 1962 35. Potter JF, Schoeneman M: Metastasis of maternal cancer to the placenta and fetus. Cancer 25:380, 1970 36. Purtilo DT, Hallgren HM, Yunis EJ: Depressed maternal lymphocyte response to phytohaemagglutinin in human pregnancy. Lancet 1:769, 1972 37, Ribeiro GG, Palmer MK: Breast carcinoma associated with pregnancy: A clinician's dilemma. Br Med J 2:1524, 1977 38, Ribeiro G, Jones DA, Jones M: Carcinoma of the breast associated with pregnancy. Br J Surg 73:607, 1986 38a. Rissanen PM: Carcinoma of the breast during pregnancy and lactation. Brit J Cancer 22:663, 1968 39. Sakai F, Saez S: Existence of receptors bound to endogenous estradiol in breast cancers in premenopausal and postmenopausal women. Steroids 27:99, 1976 40. SarrifWM, Durant JR: Evidence that estrogen-receptor-negative progesterone receptorpositive breast cancer and ovarian cancer contain estrogen receptor. Cancer 48:1215, 1981 41. Schapira DV, Chudley AE: Successful pregnancy following continuous treatment with combination chemotherapy before conception and throughout pregnancy. Cancer 54:800, 1984 42, Sweet DL, Kinzie J: Consequences of radiotherapy and antineoplastic therapy for the fetus. J Reprod Med 17:241, 1976 43. Wallack MK, Wolf JA, Bedwinek J, et al: Gestational carcinoma of the female breast. Curr Prob Cancer 7(9):1, 1983 44. White TT, White WC: Breast cancer and pregnancy: A report of 49 cases followed 5 years. Ann Surg 144:384, 1956
Address reprint requests to Herbert C, Hoover, Jr, MD Cox-1 Massachusetts General Hospital Boston, Massachusetts 02114
0039-6109/90 $0.00
+ .20
Breast Cancer During Pregnancy and lactation
Herbert C. Hoover, Jr, MD*
Breast cancer, the most common cancer, and in many geographic areas, the most frequent cause of cancer deaths in women, is also the cancer most likely to be seen in pregnancy or lactation. 11 The occurrence of breast cancer during pregnancy and lactation presents a challenging clinical problem, which is compounded by its infrequent occurrence and by the emotional issues involved. One of life's most joyous experiences can become a nightmare, often unnecessarily. Because most physicians deal with this clinical situation so infrequently, anecdotal experiences can easily propagate misinformation. We will discuss the principal questions related to the diagnosis, therapy, and prognosis of cancer of the breast in pregnancy and lactation.
HISTORICAL ASPECTS Until Halsted's pioneering work, breast cancer diagnosed during pregnancy was considered incurable. In 1896, Halsted performed a radical mastectomy on a lactating woman who was reported to be alive and well more than 30 years later. This long-term follow-up was provided by Bloodgood as a part of a report by Kilgore 26 describing a 70% 5-year survival rate in lactating women with node-negative breast cancer and a 11.5% 5year survival rate in patients with nodal involvement. These early writings rejected the idea that milk fistula would uniformly result from a breast biopsy during pregnancy or lactation. Bloodgood also mentioned that general anesthesia apparently had no harmful effect on the fetus and intimated that breast cancer during pregnancy should, in fact, be treated much as in the nonpregnant woman. However, an attitude of gloom was propagated by the 1943 publication by Haagensen and Stoutl6 stating that breast carcinoma developing during pregnancy and lactation was categori-
*Associate
Professor of Surgery, flarvard Medical School; and Chief, Surgical Oncology Research, Massachusetts General Hospital, Boston, Massachusetts
Surgical Clinics of North America-Va!' 70, No.5, October 1990
1151
1152
HERBERT
C.
HOOVER, JR
cally inoperable. This view was based on their experience with only 20 patients (19 of whom had positive nodes), all of whom had died of their breast cancer. Haagensen reversed his view later14 after more extensive experience, but the pessimistic attitude was widely prevalent and remains so today. However, studies in the past 30 years have demonstrated convincingly that the poor results are attributable to delay in diagnosis and reluctance to treat patients aggressively during pregnancy rather than to a detrimental effect of pregnancy or lactation itself.
EPIDEMIOLOGY Breast cancer diagnosed during pregnancy or lactation is relatively rare. Indeed, because the incidence of breast cancer peaks between the ages of 50 and 55, one would not expect to see the frequent coexistence of pregnancy and breast cancer. White and coworkers44 reviewed 1296 such cases, including 60 of their own, among 45,881 breast carcinoma cases, an incidence of approximately 2.8%. Of 300,860 pregnancies reviewed, there were only 93 breast cancers documented, or 3 per 10,000 pregnancies. Wallack and associates,43 in a large literature review, put the incidence even lower, with breast cancer associated with pregnancy or lactation in only 1%-2% of cancer cases. In most series, the average age of patients has been 34 to 35 years. With the current trend toward postponing childbearing to the mid to late 30s, it is expected that the incidence of breast cancer diagnosed during pregnancy and lactation will increase.
ETIOLOGY Even though pregnancy and cancer are the only two biologic conditions in which antigenic tissue is known to be tolerated by a seemingly intact immune system,13 there is no evidence to implicate pregnancy or lactation in either the etiology or the progression of breast cancer. There is evidence for depressed cell-mediated immunity, such as an abnormal lymphocyte response to phytohemagglutinin in vitro and impairment of in vivo responses to common recall antigens used in testing for delayed cutaneous hypersensitivity.36 Additionally, some of the pregnancy-induced hormonal changes, such as the elevation of circulating corticosteroids, could be immunosuppressive. 28 Theoretically, these changes could promote more rapid tumor growth but are unlikely to play any significant etiologic role. According to our knowledge of the biology of breast cancer, a cancer manifesting itself during pregnancy begins months or even years before conception. Of course, the etiology of breast cancer remains unknown, but there appears to be nothing unique about breast cancer in pregnancy.
CLINICAL PRESENTATION Delay in diagnosis appears to be the primary, and perhaps the only, reason for the generally worse prognosis overall for all patients with breast
BREAST CANCER DURING PREGNANCY AND LACTATION
1153
cancer diagnosed during pregnancy or lactation. 1. 5. 6. 15. 22. 27. 31. 34. 37. 38 The duration of symptoms before diagnosis averaged 5 to 15 months in some series, which is considerably greater than in series of nonpregnant patients. This delay can be attributed to physician neglect as readily as to the patient's. During pregnancy, the breasts become tense and often have a multinodular consistency. A discrete mass can be difficult to feel or can be obscured as the breasts hypertrophy. Because of the pregnancy, masses are often ignored even when noticed by patients or physicians, the plan being to evaluate them further if they persist after delivery. Mammography often is not recommended during pregnancy because of the radiation exposure to the fetus and the poor imaging obtained in the tense, hyperplastic breast tissue. The latter is a significant problem, but the former need be of no serious concern with modern techniques and proper abdominal and pelvic shielding. As a consequence of these delays, virtually all articles on patients with breast cancer during pregnancy report a more advanced stage than in nonpregnancy series. Ribeiro and Palmer37 reported on 88 patients with breast cancer concurrent with pregnancy, 19 of whom had inoperable, faradvanced tumors. Of the 69 patients with operable disease, 89% were found to have positive nodes in the ipsilateral axilla. Holleb and Farrow2 reported 72% with positive nodes among 117 patients. These figures are far in excess of the 40% to 50% node-positive rate reported in most series of breast cancer in nonpregnant women. 19 DIAGNOSIS Primary care physicians, especially obstetrician-gynecologists, play a pivotal role in the diagnosis of breast cancer during pregnancy and lactation. It is especially important that a careful breast examination be performed at the initial obstetric visit before the breasts become engorged and difficult to examine carefully. As mentioned earlier, studies suggest that both patients and their physicians often fail to recognize the potentially serious nature of breast masses during pregnancy. There should be nothing special about the diagnostic evaluation of pregnant or lactating patients with breast masses, with the exception that xerography tends not to be as helpful because of the parenchymal changes associated with gestation. Increased water density of the breasts reduces the discriminatory capacity of mammography.21 Radiation exposure to the fetus should be negligible with proper shielding of the abdomen. In 368 pregnant women undergoing mammography, one series reported no apparent damage to the fetus. 28 Nevertheless, in a patient with a discrete palpable mass, a mammogram would rarely influence therapy, so it need not be obtained. Any dominant mass should be evaluated promptly in pregnant or lactating patients. This evaluation normally should begin with a fine-needle aspiration to distinguish cystic from solid lesions. Any new solid mass requires serious consideration of a biopsy, which can almost always be performed under local anesthesia.
1154
HERBERT
C.
HOOVER, JR
The breast masses biopsied during pregnancy show the same spectrum of histopathology found in nonpregnant patients. Additionally, the likelihood of finding a carcinoma appears to be similar to that in breast masses biopsied in the nonpregnant population. 4 Byrd and associates 4 found that 22% of breast biopsies in pregnancy showed malignancy compared with 19% in all patients overall. Although the earlier literature 44 implied that inflammatory carcinoma was seen more frequently during pregnancy, modern series do not confirm this. For biopsies done during lactation, some surgeons prefer to suppress lactation preoperatively using bromocriptine, although there is no body of evidence to support this measure. The risk of milk fistulas is very low for peripheral lesions, although such fistulas occasionally cause a problem in deep, central lesions.
HORMONE RECEPTORS Steroid hormone receptors may be difficult to demonstrate in breast cancer tissue during pregnancy unless special methods are utilized. 12. 25, 39, 40 Pregnancy may depress levels of estrogen and progesterone receptors detectable in breast cancer cytosol fractions, resulting in false-negative studies because high levels of estrogens circulating in pregnant patients cause receptor translocation into the nucleus and occupy all the cytoplasmic receptors, leaving none available for assay. Therefore, unbound estrogen should first be removed from breast cancer specimens by treatment of the cytosol with dextran-charcoal. Exchange assays can then be performed to detect occupied estrogen receptors within the cytosol and nuclei. However, to my knowledge, there are no data available on estrogen receptors in breast cancer during pregnancy that show any meaningful prognostic significance.
ANESTHETIC CONSIDERATIONS There is no evidence that a breast biopsy poses any significant risk to either the mother or the fetus even if general anesthesia is required, assuming that proper precautions are used. Byrd and associates 4 report only one fetal loss in 134 breast biopsies on pregnant patients under general anesthesia, and this loss was in an older patient not known to be pregnant at the time of the operation, which therefore was done without the usual precautions. There are some unique management considerations because of the phy~iologic changes of pregnancy and the physical effects of an enlarged, gravid uterus. The special problems posed by general anesthesia in pregnant patients are reviewed in several recent publications, 2, 7, 23, 29, 32, 33 which have been extensively discussed by Wallack and coworkers 43 and will be summarized here. A number of cardiovascular and hematologic alterations during pregnancy necessitate anesthetic adjustments. Blood volume increases by ap-
BREAST CANCER DURING PREGNANCY AND LACTATION
1155
proximately 35%, but this is primarily in the plasma rather than red cell volume, giving a physiologic dilutional anemia. By term, the heart rate increases 10 to 15 beats per minute. Along with the increase in blood volume and heart rate is a 30% to 50% increase in cardiac output. There is a slight decrease in the systemic vascular resistance as well. The lower hemoglobin content leads to a decrease in oxygen-carrying capacity, and the task of maintaining oxygen saturation and cardiac output becomes critical. A high inspired oxygen concentration (at least 50%) and the avoidance of agents that depress cardiac output are advised. Pregnancy also alters coagulation. The platelet count increases nearly 50%, as does the fibrinogen level. The resulting hypercoagulable state can increase the risk of thromboembolism. Serum cholinesterase activity is decreased by approximately 20%, prolonging the metabolism of succinylcholine. With the patient supine, a gravid uterus may interfere with blood flow because of pressure on the distal aorta and inferior vena cava, making positioning during anesthesia of great importance. Tilting the operating table slightly to the left and putting a cushion under the patient's right hip can help to displace the uterus to the left and away from major vascular structures. A large gravid uterus raises the level of the diaphragm significantly, thus compromising lung volumes. Decreased functional residual capacity along with a pregnancy-associated increase in the breathing rate can lead to a more rapid build-up of anesthetic vapor within the aveoli and thus a more rapid induction of anesthesia. Forced hyperventilation and maternal respiratory alkalosis can affect the fetus adversely by causing progressive fetal acidosis. The anesthetist must exercise great care in maintaining a "normal for pregnancy" ventilatory balance. Because of the decreased functional residual capacity and diminished ability to store oxygen, coupled with the increased oxygen consumption in pregnancy, the pregnant patient is predisposed to hypoxemia during even brief apneic episodes. Preoxygenation for several minutes should always precede endotracheal intubation. Intubation itself should be carefully performed because of the hypervascularity in the respiratory tract mucosa in most pregnant women. Gastric emptying time is prolonged in pregnancy, being as much as 60% above normal as the patient approaches term. The pressure on the stomach makes reflux much more likely as well. Hypergastrinemia not uncommonly leads to more acidic gastric secretions. Therefore, pregnant patients require special intubation techniques to avoid aspiration of this very damaging acidic material. There are no published studies to confirm any deleterious effects of anesthetic agents on the developing fetus. There are many studies of the offspring of narcotic addicts, which have shown no long-term effect of morphine, heroin, or methadone other than low birth weight. 3 Fetal monitoring should be used whenever possible so that early distress can be recognized and appropriate adjustments made. Otherwise, it is felt that general anesthesia poses no real hazard as long as careful attention is given to the unique requirements of the situation.
1156
HERBERT
C. HOOVER, JR
STAGING EVALUATION Once the diagnosis of breast cancer is made, staging of the disease is critical before therapeutic decisions are made. Because all of the staging procedures, with the exception of blood studies, involve radiation exposure, there is considerable controversy concerning the extent of evaluation that should be performed in breast cancer patients who are pregnant. Liver function tests, calcium measurements, and carcinoembryonic antigen assays can give some useful information but do not permit a definitive diagnosis or localization of metastatic disease. A thorough understanding of the radiation risk to the fetus is helpful to the clinician deciding on the diagnostic tests to pursue. Hall 17 has published much useful information in this regard. Radiation delivered during the preimplantation phase, from fertilization to embryonic implantation in the uterine wall (days 0 to 10), can result in the death of the embryo. As little as 10 rad can be lethal in the mouse. He did find that those embryos that survived irradiation during this period appeared to develop normally. Irradiation during organogenesis (days 11 to 56) can lead to serious developmental malformations. In humans, the most frequently noted defect is microcephaly. Again, doses as low as 10 rad may be sufficient to produce a measurable increase in birth defects, and doses exceeding 100 rad are felt to produce congenital abnormalities in 100% of cases on the basis of the atomic bomb experience. 27 • 30 Mter day 57, during the stage when additional fetal growth is occurring, larger doses of radiation are required to cause significant effects. After weeks 25 to 30, the fetus is far more resistant to the damaging effects of radiation. 8 The majority of the reports on the effects of ionizing radiation on the developing organs are extrapolations from animals to man, with doses that are much higher than are pertinent to the radiologic evaluation for metastases in breast cancer, although they are more relevant in considering therapeutic breast irradiation. However, there are insufficient data in the literature to quantify the long-term risk to the fetus from the low radiation doses involved in most scanning procedures. For example, the radiation dose to the fetus from a bone scan is approximately 0.1 rad. Even though we conclude that any radiologic staging procedure is probably safe in a pregnant patient, we cannot say this with certainty. Therefore, bone scans generally are not recommended unless significant therapeutic decisions would be based on their outcome. This is similar to our policy in stage I and II nonpregnant patients with carcinoma of the breast, where the diagnostic yield in an asymptomatic patient is not felt to be high enough to justify the expense on a routine basis. Harbert18 reviewed a large number of studies showing a 3% true-positive yield from bone scans in 533 patients with stage I breast cancer, 7% in 696 patients with stage II disease, and 25% in 278 with stage III disease. When pregnant patients with stage I or II disease have skeletal pain, a bone scan is preferred to radiographic evaluation, as the radiation exposure from the scan would be considerably less than that of a skeletal survey. In clinical stage III disease, where a significantly higher positive scan rate would be expected, one would still prefer to defer the bone scan until the later stages of pregnancy
BREAST CANCER DURING PREGNANCY AND LACTATION
1157
unless one would decide to treat with systemic therapy at that stage of the pregnancy if bone metastases were demonstrated. This raises complex issues as to which patients should be considered for termination of pregnancy, a topic that will be discussed in the therapeutic section. Carcinogenesis from radiation exposure is another possible risk to the fetus. Although the association between fetal radiation exposure and cancer was not clearly shown in the atomic bomb experience, a recent study of twins showed that those who both developed cancer or leukemia were twice as likely to have been exposed to x-rays in utero as were control sets of twins.20 Rather than take chances, it is best to avoid any radiation exposure during pregnancy unless the data to be gained are critically important. This would not apply to routine chest radiographs, which can be done with essentially no risk to the fetus with proper abdominal and pelvic shielding. However, in stage I breast cancer, the chance of a positive chest study is sufficiently low that one could justify deferring the chest film until delivery of the fetus. There is no role for radionuclide brain or liver scans in these patients. Any patient with CNS symptoms would be better served by a CT scan, and the liver can be evaluated with ultrasound at no risk to the fetus in patients with liver function abnormalities. TREATMENT Although Haagensen and Stout concluded in 1943 that the prognosis was so poor that mastectomy was not justified in patients with breast cancer during pregnancy or lactation,16 Haagensen did not persist with that point of view for long. He soon realized that his initial experience with 20 patients, who all died, was not representative of the prognosis. 17 Virtually all reports since then have stressed that the surgical treatment should be essentially that of the nonpregnant patient in all stages of disease. More controversial has been whether termination of pregnancy should be advised. Obviously, the therapeutic decisions will be influenced by the stage of pregnancy and the stage of disease, so we must consider the treatment plan accordingly. Because the risk of spontaneous abortion during mastectomy is low, 4 mastectomy or other breast procedures should be performed promptly upon diagnosis at any stage of pregnancy. A mastectomy allows a normal pregnancy to continue with little risk to the mother or the fetus. This is especially true in that no modem series suggests that abortion favorably affects the course of patients with breast cancer. 9, 10 Therefore, any decision to terminate the pregnancy should be based on considerations other than abortion's influence on the progression of the disease. Termination of pregnancy has no role in the management of patients with stage I or II breast cancer. Likewise, there is no substantial evidence that oophorectomy influence's the course of breast cancer during pregnancy and lactation. With the trend toward breast conservation and radiation, some pregnant women with early stages of breast cancer will want to consider that option. It should be strongly discouraged except in cases where radiation therapy could reasonably be withheld until after delivery, because the
1158
HERBERT
C.
HOOVER, JR
standard technique of whole-breast irradiation with a boost dose to the tumor bed would produce a dose scattered to the fetus that would be unacceptably high. 28 There are two sources of radiation dosage outside the beam: scatter within the body, and leakage from the head of the machine. With proper lead blocks, the leakage rate can be reduced to a low range, less than 2 rad. However, internal scatter is dependent on field size, photon energy, and proximity to the field edge; and shielding cannot modify this scatter. Early in pregnancy, when the distance between the breast and uterus is great, the internal scatter dose may be only 10 to 20 rad for a total course of treatment. However, when the top of the uterus approaches the xiphoid, fetal doses of as much as 100 rad or even more may be given. 28 Obviously, special treatment arrangements can be made for women who insist on breast conservation. This could involve primary excision using a small boost of irradiation to the area with delivery of full breast irradiation after the patient gives birth. The patient would have to be aware of the potential risk to the fetus even with this low dose. Every effort should be made to discourage such an approach. A more reasonable alternative would be simple surgical removal of the primary tumor and axillary lymph nodes with observation until delivery, when further therapy could be given. Adjuvant chemotherapy has become standard for node-positive premenopausal women with breast cancer and is now considered in nodenegative patients. Because it is felt that therapeutic delay may diminish the benefits of adjuvant therapy, the decision to initiate chemotherapy during pregnancy is one that requires serious consideration. Unfortunately, the risks associated with chemotherapy during pregnancy are not clear. Most of the hard data have been developed in laboratory animals, in which the period of greatest risk is the initial portion of the pregnancy. Schapira and Chudley41 found the incidence of teratogenicity of chemotherapeutic agents given to humans in the first trimester of pregnancy to be only 12.7%, while Sweet and Kinzie 42 reported an 11.5% incidence. Neither investigation reported evidence for teratogenicity from the administration of chemotherapeutic agents in the second and third trimesters of pregnancy. Most of the data involve single drug use only, and little is known about the combination regimens usually given to breast cancer patients. Additionally, there are no good studies on the long-term consequences of fetal exposure to chemotherapeutic agents. In general, the use of any chemotherapeutic agents during the first trimester of pregnancy should be discouraged, but their use during the second and third trimester probably induces very few abnormalities. Nodenegative women should not be treated with chemotherapy at any stage of their pregnancy until its efficacy is more clearly defined in nonpregnant women. The following recommendations for therapy are suggested. Stages I and II (Operable) Breast Cancer Modified radical mastectomy is the treatment of choice in the pregnant patient. A second choice would be total tumor excision and axillary dissection to be followed by whole-breast irradiation after delivery. In patients with positive axillary lymph nodes, chemotherapy should certainly
1159
BREAST CANCER DURING PREGNANCY AND LACTATION
Table 1. Treatment Results in Operable Breast Cancer During Pregnancy or Lactation 5-YEAR SURVIVAL NO. OF SOURCE
PATIENTS
Peters (1962)34 Holleb & Farrow (1962)22 Rissanen (1968)38. Donegan (1978)10 Ribeiro et al (1986)38 Total
60 119 33 24 121 357
Positive Nodes (%)
Negative Nodes (%)
23 17 36 31 25 26%*
62 65 80 86 79 74%*
OVERALL(%)
37 30 43 48 37 39%*
*Mean.
be delayed until the second trimester of pregnancy. My personal recommendation would be to delay all chemotherapy in node-positive patients until after delivery. With determination of fetal age and maturity, consideration should be given to an early cesarean section to minimize the delay in beginning chemotherapy. Stage III (Locally Advanced) and Stage IV (Distant Metastases) Breast Cancer In patients with locally advanced or metastatic carcinoma diagnosed early in the pregnancy, for whom both chemotherapy and radiation therapy would be recommended if the woman were not pregnant, serious consideration must be given to termination of the pregnancy. In the later stages of pregnancy, nonsurgical therapy could be qelayed until after delivery, but one must weigh the risks of observing an advanced breast cancer for a long period of time. As mentioned earlier, chemotherapy can be used with little known risk during the second and third trimester, but, unfortunately, good data for standard breast chemotherapy combinations are lacking. If radiation therapy is imperative at any stage of pregnancy, termination of pregnancy may be the most reasonable alternative. The mother's life expectancy is often limited, and damage to the fetus is a significant possibility. There are no "right" answers to many of the therapeutic dilemmas encountered in patients developing breast cancer during pregnancy. A thorough discussion of all the risks involved with any approach must be
Table 2. Stages of Breast Cancer at Diagnosis in Patients Under 40 Years Old NO. (%)
STAGE
I (negative nodes) II (positive nodes)
III (distant disease)
Pregnant (N = 19)
Nonpregnant (N = 155)
4 (21) 14 (74) 1 (5)
84 (54) 57 (37) 14 (9)
From Nugent P, O'Connell TX: Breast cancer and pregnancy. Arch Surg 120:1221-1224, 1985; with permission.
1160
HERBERT
C.
HOOVER, JR
presented to the patient and her family, who must be a part of these oftendifficult decisions. PROGNOSIS The rarity of breast cancer in pregnancy or lactation dictates that all reports of the entity are retrospective. Most authors report on only a few patients, making statistical analysis difficult or impossible. Also, nearly all have considered pregnancy and lactation together, assuming no differences between the two. Despite all of the shortcomings of smaU, retrospective studies, some general conclusions can be drawn. Table 1 lists several series chosen as those most likely to be representative on the basis of their size and presentation of objective data. They clearly show that pregnant or lactating women who are identified early with negative axillary lymph nodes have an expected outcome similar to that of nonpregnant women. Nugent and O'Conne1l31 have shown this fact clearly in a small series comparing the distribution of stages of disease in pregnant patients with that in women under 40 years of age who were not pregnant (Table 2). The usual trend toward a higher stage of disease at diagnosis was found in pregnant patients, with 74% having positive nodes compared with 37% of the nonpregnant women. When 5-year survival rates were compared in these groups, both by stage and overall, no difference was noted (Table 3). The challenge clearly is to impact on this disease by making the diagnosis earlier. The bulk of evidence supports the opinion that pregnancy does not worsen the prognosis except by obscuring the disease, often for many months, allowing the metastatic process to progress. We can hope that increased vigilance on the part of physicians seeing these patients early in their pregnancy can reverse this trend. OTHER CONSIDERATIONS Tumor Metastatic to the Fetus Some women with breast cancer diagnosed during pregnancy fear that cancer cells have spread to their fetus and could grow there. This has been Table 3. Five-Year Survival Rates of Pregnant and Nonpregnant Women Under 40 Years Old with Breast Cancer SUjWIVAL NO. (%) STAGE
Pregnant
Nonpregnant
I II III Overall
4/4 (100) 7/14 (50) 011 (0) 11119 (57)
59/84 (70) 27/57 (48) 1114 (7) 87/155 (56)
P' 0.57 1.00 1.00 1.00
*By Fisher's exact test From Nugent P, O'Connell TX: Breast cancer and pregnancy. Arch Surg 120:1221-1224, 1985; with permission.
BREAST CANCER DURING PREGNANCY AND LACTATION
1161
reported in melanoma and choriocarcinoma35 but, to my knowledge, not in breast cancer. Pregnancy After Treatment of Breast Cancer Many young women treated successfully for breast cancer desire further pregnancies. Some studies34, 20 have, in fact, shown an improved survival rate among breast cancer patients who become pregnant. Some authors have argued that this phenomenon relates to patient selection, in that only women without an early recurrence and with a good prognosis live to become pregnant or desire pregnancy, but Peters34 argues that improved survival among patients whose pregnancy occurred within a short interval after mastectomy is evidence against a selection process. She demonstrated an improvement in survival of 23% at 5 years and 28% at 10 years when 96 subsequent pregnant patients were compared with 96 patients of similar age and clinical stage in whom pregnancy did not occur. In 41 patients treated for primary operable carcinoma of the breast, Harvey and coworkers 20 found no detrimental effect of subsequent pregnancy even among patients with positive axillary nodes or those whose pregnancies occurred less than 2 years after mastectomy. Abortion produced no improvement in the survival rate. They concluded, as have most other authors of recent series, that there are no therapeutic grounds for recommending avoidance or termination of pregnancy among patients without evidence of recurrent disease after treatment for breast cancer. SUMMARY Breast cancer is the most frequently seen cancer in pregnancy and lactation, but the incidence is low, the disease being seen in approximately 0.03% of pregnancies. Only 1% to 2% of breast cancer overall is diagnosed during pregnancy or lactation. There is no evidence to implicate pregnancy or lactation in either the etiology or the progression of breast cancer. Careful breast examination early in the pregnancy is very important to find solid masses that require biopsy before breast engorgement hides them. Therapeutic options vary, depending on the stage of disease and the stage of the pregnancy. Opetable disease in the first 6 to 7 months of the pregnancy should be treated by mastectomy, as irradiation is contraindicated. Late in the pregnancy, a lumpectomy and axillary dissection can be done, with irradiation being delayed until after delivery. General anesthesia is safe if the usual precautions are taken to compensate for the physiologic changes induced by pregnancy. Unfortunately, delay in diagnosis is common, and 70% to 89% of patients with operable primary lesions have positive axillary lymph nodes. Late stage appears to be the only reason for the generally worse prognosis in these patients, as stage for stage, they have a course similar to that of nonpregnant patients. Adjuvant chemotherapy can be considered late in the pregnancy but should usually be delayed until after delivery. In patients with locally advanced or metastatic cancer diagnosed early in the pregnancy, for whom both chemotherapy and radiation therapy would normally be recommended, consideration must be
1162
HERBERT
C.
HOOVER, JR
given to termination of the pregnancy. There is no evidence that termination of pregnancy improves the outlook for the patients, but it does permit standard aggressive therapy in advanced disease.
REFERENCES 1. Anderson JM: Mammary cancers and pregnancy. Br Med J 1:1124, 1979 2. Attia RR, Ebeidan-Fischer JE: Gastrin: Placental, maternal and plasma cord levels: Its possible role in maternal residual gastric acidity. Abstr Am Soc Anesthesiologists 547, 1976 3. Blinick G, Jerez E, Wallach RC: Methadone maintenance, pregnancy and progeny. JAM A 255:477, 1973 4. Byrd BF, Bayer DS, Robertson JC, et al: Treatment of breast tumors associated with pregnancy and lactation. Ann Surg 155:940, 1962 5. Canter JW, Oliver GC, Zaloudek CJ: Surgical diseases of the breast during pregnancy. Clin Obstet Gynecol 26:853, 1983 6. Cheek JH: Cancer of the breast in pregnancy and lactation. Am J Surg 126:729, 1973 7. Cohen SE: Why is the pregnant patient different? Semin Anesthesia 1:73, 1982 8. Dekaban AS: Abnormalities in children exposed to X-radiation during various stages of gestation: Tentative time table of irradiation injury to the human fetus 1. J Nucl Med 9:471, 1968 9. Donegan WL: Mammary carcinoma in pregnancy. In Spratt JS, Donegan WL (eds): Major Problems in Clinical Surgery. Philadelphia, WB Saunders, 1967, p 170 10. Donegan WL: Mammary carcinoma in pregnancy. In Hall EJ (ed): Radiobiology for the Radiologist. New York, Harper & Row, 1978, p 397 11. Donegan WL: Cancer and pregnancy. CA 33:194, 1983 12. Garola RE, McGuire WL: An improved assay for nuclear estrogen receptor in experimental and human breast cancer. Cancer Res 37:333, 1977 13. Gleichner N, Siegel!: Common denominators of pregnancy and malignancy. In Gleichner N (ed): Reproductive Immunology. New York, Alan R Liss, 1981, P 339 14. Haagensen CD: The treatment and results in cancer of the breast at the Presbyterian Hospital, New York. Am J Roentgenol 62:328, 1949 15. Haagensen CD: Cancer of the breast in pregnancy and lactation. Am J Obstet Gynecol 98:141, 1967 16. Haagensen CD, Stout AP: Carcinoma of breast: Criteria of operability. Ann Surg 118:859, 1943 17. Hall EJ: Effects on the embryo and fetus. In Hall EJ (ed): Radiobiology for the Radiologist. Hagerstown, Harper & Row, 1973, P 231 18. Harbert JC: Efficacy of bone and liver scanning in malignant disease facts and options. In Nuclear Medicine Annual 1982. New York, Raven Press, 1982 19. Harris JR, Hellman S, Canellos GP, et al: Cancer of the breast. In DeVita VT, Hellman S, Rosenberg SA (eds): Cancer: Principles and Practice of Oncology, ed 2. Philadelphia, JB Lippincott, 1985, p 1119 20. Harvey EB, Borce JD, Honeyman M, et al: Prenatal X-ray exposure and childhood cancer in tWins. N Engl J Med 315:541, 1985 21. Hoeflken W, Lanyi M: Mammography. Philadelphia, WB Saunders, 1977 22. Holleb AI, Farrow JH: The relation of carcinoma of the breast in pregnancy in 283 patients. Surg Gynecol Obstet 115:65, 1962 23. Huseymeyer RP, Davenport HT: Prophylaxis for Mendelson's syndrome before elective cesarean section: A comparison of cimetidine and magnesium trisilicate mixtures of regimens. Br J Obstet Gynecol 87:565, 1980 24. Jablon S, Kato H: Childhood cancer in relation to pre-natal exposure to atomic bomb radiation. Lancet 2: 1000, 1970 25. Katzenellenbogen JA, Johnson HJ, Carlson KE: Studies on the uterine, cytoplasmic estrogen binding protein: Thermal stability and ligand dissociation rate: Ail assay for empty and filled sites by exchange. Biochemistry 12:4092, 1973 26. Kilgore AR: Tumors and tumor-like lesions of breast in association with pregnancy and
BREAST CANCER DURING PREGNANCY AND LACTATION
1163
lactation (with a note by Bloodgood JC): The treatment of tumors of the breast during pregnancy and lactation. Arch Surg 18:2079, 1929 27. King RM, Welch JS, Marten JK, et al: Carcinoma of the breast associated with pregnancy. Surg Gynecol Obstet 160:228, 1985 28. Lippman ME, Lichter AS, Danforth DN: Breast cancer occurring during pregnancy. In Diagnosis and Management of Breast Cancer. Philadelphia, WB Saunders, 1988, p 414 29. Mendelson CL: Aspiration of stomach contents into the lungs during obstetric anesthesia. Am J Obstet Gynecol 53:191, 1946 30. Miller R, Mulvihill S: Small head size after atomic irradiation. Teratology 14:355, 1976 31. Nugent P, O'Connell T: Breast cancer in pregnancy. Arch Surg 120:1221, 1985 32. Nunn JF: Applied Respiratory Physiology, ed 2. Boston, Butterworths, 1977, p 397 33. Pedersen H, Finster M: Anesthetic risks in the pregnant surgical patient. Anesthesiology 51:439, 1979 34. Peters MV: Carcinoma of the breast associated with pregnancy. Radiology 78:58, 1962 35. Potter JF, Schoeneman M: Metastasis of maternal cancer to the placenta and fetus. Cancer 25:380, 1970 36. Purtilo DT, Hallgren HM, Yunis EJ: Depressed maternal lymphocyte response to phytohaemagglutinin in human pregnancy. Lancet 1:769, 1972 37, Ribeiro GG, Palmer MK: Breast carcinoma associated with pregnancy: A clinician's dilemma. Br Med J 2:1524, 1977 38, Ribeiro G, Jones DA, Jones M: Carcinoma of the breast associated with pregnancy. Br J Surg 73:607, 1986 38a. Rissanen PM: Carcinoma of the breast during pregnancy and lactation. Brit J Cancer 22:663, 1968 39. Sakai F, Saez S: Existence of receptors bound to endogenous estradiol in breast cancers in premenopausal and postmenopausal women. Steroids 27:99, 1976 40. SarrifWM, Durant JR: Evidence that estrogen-receptor-negative progesterone receptorpositive breast cancer and ovarian cancer contain estrogen receptor. Cancer 48:1215, 1981 41. Schapira DV, Chudley AE: Successful pregnancy following continuous treatment with combination chemotherapy before conception and throughout pregnancy. Cancer 54:800, 1984 42, Sweet DL, Kinzie J: Consequences of radiotherapy and antineoplastic therapy for the fetus. J Reprod Med 17:241, 1976 43. Wallack MK, Wolf JA, Bedwinek J, et al: Gestational carcinoma of the female breast. Curr Prob Cancer 7(9):1, 1983 44. White TT, White WC: Breast cancer and pregnancy: A report of 49 cases followed 5 years. Ann Surg 144:384, 1956
Address reprint requests to Herbert C, Hoover, Jr, MD Cox-1 Massachusetts General Hospital Boston, Massachusetts 02114