Education and debate

Management of metastatic melanoma during pregnancy

Commentary: Pregnancy should not have affected treatment for melanoma

Commentary: Self interest is not the sole legitimate basis for making decisions

Commentary: Management of the patient

Management of metastatic melanoma during pregnancy

S R D Johnston, consultant medical oncologist, ^a K Broadley, consultant in palliative medicine, ^b G Henson, consultant obstetrician, ^d C Fisher, consultant histopathologist, ^c M Henk, consultant clinical oncologist, ^a M E Gore, head of skin and melanoma unit. ^a

^a Melanoma Unit, Royal Marsden Hospital, London SW3 6JJ, ^b Palliative Care Unit, Royal Marsden Hospital, London SW3 6JJ, ^c Department of Pathology, Royal Marsden Hospital, ^d Department of Obstetrics and Gynaecology, Whittington Hospital, London N19 5NF

Correspondence to: Dr Johnston stephen{at}icr.ac.uk

About 35% of women with melanoma are of child bearing age, and the coexistence of melanoma and pregnancy is increasing.¹ Many doctors recommend that women wait two to three years after successful treatment for melanoma before becoming pregnant as most recurrences occur during this time. This advice is inadequate. Doctors need to inform these women about the considerable problems that may arise if relapse occurs while they are pregnant. We present our recent experience of this difficult situation: malignant melanoma within the maternal intervillous space, invading into the core of the villus. Immunostaining for S100 protein and HMB45 was positive, and staining for human chorionic gonadotropin was negative in the tumour cells.

Case report

A 41 year old woman presented with left axillary lymphadenopathy. A superficial spreading melanoma had been removed from her back two years previously. Radical lymph node dissection was performed. All nodes contained metastatic melanoma that stained positive for S100 and HMB45. There was evidence of extranodal tumour towards the deep resection margin, but her chest radiograph and liver ultrasonography were both normal. The woman had a professional career and had never been pregnant.

The patient presented four months later with a short history of low back pain. Clinical examination showed no abnormalities and her neurological signs were normal. However, plain radiographs of the lumbar spine and computed tomography showed that the 8th thoracic vertebra had collapsed, associated with a soft tissue mass consistent with metastatic melanoma. The spinal cord was not affected.

The patient reported that she was 12 weeks pregnant. Termination was discussed and the patient was counselled, but she and her partner were determined to continue with the pregnancy. Since she had no neurological signs of cord compression and her pain was intense, she was given a single palliative dose of radiotherapy to the affected vertebra (8 Gy to a depth of 5 cm). The scatter dose of radiation received by the uterine fundus was calculated to be 0.0005 Gy, which represents a small increased risk of fetal malformation and leukaemia. The use of analgesic drugs in pregnancy was discussed, but she was reluctant to take these.

The patient's pain improved initially, but eight weeks later she developed bilateral leg weakness and difficulty in walking. Neurological examination showed that she had grade 3-4 weakness in her legs and loss of both ankle jerk reflexes. Magnetic resonance imaging confirmed that her spinal cord was being compressed by tumour at D8/9, and that there were several tumour deposits at other levels. She underwent a decompressive laminectomy and pediculotomy. Histological examination of the tumour confirmed metastatic melanoma. During her recovery several skin nodules developed and splenic metastases were noted on ultrasonography. Fetal growth and development seemed normal at 25 weeks' gestation. The clinical problem was how to manage progressive systemic metastases in a woman expecting her first child.

References

1. Reintgen OS, McCarthy K, Vollmer R. Malignant melanoma and pregnancy. Cancer 1985; 55: 1340-1344[Medline].

Commentary: Pregnancy should not have affected treatment for melanoma

Robert Hammond, consultant obstetrician and gynaecologist. 

University Hospital, Queen's Medical Centre, Nottingham NG7 2UH

The occurrence of malignant disease in pregnancy turns what is usually a happy experience into a potential nightmare. In no situation in medical practice is the concept of informed choice more important when considering management options. As in all problems related to pregnancy, there are two individuals to consider and it is imperative that the woman and her partner are helped to prioritise their objectives on the basis of information given in a caring and sympathetic manner.

View larger version (117K): [in this window] [in a new window]   Haematoxylin and eosin stained section of placenta showing deposit of malignant melanoma within the maternal intervillous space, invading into core of the villus. Immunostaining for S100 protein and HMB45 was positive, and human chhorionic gonadotropin negative in the tumour cells

The issues

Maternal survival is usually of paramount importance, and doctors would normally wish to offer the same treatment they would give to a woman who is not pregnant. However, other factors must be considered, such as the effect of pregnancy on the disease process and vice versa. The impact of treatment on the fetus must also be borne in mind as this may result in fetal demise, congenital abnormalities, or failure of development in utero. In addition, there may be long term risk to the fetus after birth if it is affected by metastatic tumour.

The gestational age at which the problem arises is important, not just from the point of view of risks of treatment but because the doctor may wish to delay treatment until fetal viability is reachedso long as this does not compromise maternal survival. There may be circumstances in which the couple would wish to place fetal survival above maternal outcomeparticularly if they have moral objections to termination or if the prognosis for the mother is so poor that further treatment is unlikely to influence the course of the disease.

Fortunately, malignant disease in pregnancy is uncommon, but this rarity may cause problems when it does arise because doctors may not have reliable information about some of the issues. In this case, the mother presented with metastatic malignant melanoma at 12 weeks of pregnancy. Sutherland et al reported that malignant melanoma was affected by hormones, and was stimulated by increasing oestrogen and progesterone concentrations in pregnancy.¹ However, MacKie et al did not find any adverse effect of pregnancy on the disease.² On balance, there does not seem to be sufficient evidence to recommend termination of pregnancy in these cases.

The mother's median survival with treatment was about six months, and the time until fetal viability would be reached was between four and five months. The risks to the fetus of thoracic radiotherapy and general anaesthesia for surgical treatment were small, and data suggested that it would not be affected adversely by the preferred drug. Indeed, it is possible that this treatment would reduce the risk of the fetus being affected by metastatic disease.

Recommendation

In this case information on both parents' views on termination of pregnancy, as well as the father's feelings about the probability that he would be a single parent from early in the child's life, is not reported. Assuming that the parents found this scenario acceptable, and they were prepared to take the small but statistically significant risk that the fetus might develop metastatic disease after birth, I would not have recommended termination of pregnancy to them. I would have treated the patient as I would a woman who was not pregnant in the hope that she would survive for long enough to experience some joy from her baby.

References

1. Sutherland CM, Wittliff J, Mabie WC. The effect of pregnancy on hormone levels and receptors in malignant melanoma. J Surg Oncol 1983; 22: 191-192[Medline].
2. MacKie RM, Bufalino R, Morabito A, Sutherland C, Cascinelli N. Lack of effect of pregnancy on outcome of melanoma. Lancet 1991; 337: 653-655[Medline].

Commentary: Self interest is not the sole legitimate basis for making decisions

Charles Weijer, bioethicist. 

Mount Sinai Hospital, Toronto, Ontario M5G 1X5, Canada

cweijer{at}mtsinai.on.ca

Treating cancer in pregnancy presents the patient and doctor with difficult medical, ethical, and, at times, even legal questions. Should the pregnancy be terminated? How best should the cancer be treated? How should potential benefits of treatment to the mother be balanced against risks to the fetus?

Texts on bioethics help little

Little guidance is found in publications on bioethics. Discussion to date has focused on cases in which a mother has refused a potentially life saving treatment for the fetus, for example, caesarian section.¹ The best known case is that of Angela Carter, a terminally ill cancer patient who was 26.5 weeks pregnant.² Close to death, Carter refused a caesarian section and demanded that comfort measures be continued, even if the delay in surgery resulted in the death of her fetus. The hospital consulted the court, which ordered that a caesarian section be performed and, unfortunately, neither Carter nor her child survived. Review by a higher court found that the court erred in its decision: "We hold that in virtually all cases the question of what is to be done is to be decided by the patientthe pregnant womanon behalf of herself and the fetus. If the patient is incompetent ... her decision must be ascertained through ... substituted judgement.³"

The case at hand is the mirror image of the Carter case. We are told that the patient determined to continue with the pregnancy. Her actions bespeak a willingness to sacrifice her own comfort, and perhaps her chances of survival, to protect her unborn child. Even when she developed bone metastases she was reluctant to use analgesics. Continuing the pregnancy precluded treatment options such as high dose or multiagent chemotherapy, which is associated with higher response rates than chemotherapy with a single agent.⁴

Selfless decisions in others make us uncomfortable

Decisions in which one person sacrifices her own wellbeing for the sake of a loved one, a fetus in this case, make us uncomfortable. The temptation is to regard these decisions as not voluntary, as the result of "internal coercion."⁵ This temptation ought to be resisted. The problem is not with the decision as such but rather with the prevalent view of human agency in bioethics that proceeds as if self interest is the only legitimate basis for decision making. The fact is that people make sacrifices for loved ones all the time: parents get by with less so their children may go to a better school, and spouses turn down lucrative job offers to remain with their mate. Our interests, and those of the patient in this case, therefore comprise both self interest and interest with regard to others. The patient's decision to proceed with the pregnancy is thus voluntary and valid.

A muted victory

Whatever moral discomfort we might feel in such cases in general is alleviated by the particular circumstances of this case. There is no convincing evidence that therapeutic abortion improves the cure rate of melanoma in pregnancy.⁶ Dacarbazine is the most effective single agent regimen, and can be safely given in the second and third trimesters. Furthermore, while high dose and multiagent regimens are promising, none have yet been proved better than dacarbazine alone in a prospective randomised controlled trial.⁴ Finally, no currently available regimen is likely to provide cure or even long term survival. In this context, salvaging a healthy child seems like a victory, albeit a muted one, for the mother and her doctors.

References

1. Mahowald MB. Women and children in health care: an unequal majority. , New York: Oxford University Press, 1993:131-147.
2. Annas GJ. Foreclosing the use of force: AC reversed. Hastings Center Report 1990; 20: 27-29.[Medline]
3. In re: AC. District of Columbia Court of Appeal, April 26, 1990. 
4. Houghton AN, Meyers ML, Chapman PB. Medical treatment of metastatic melanoma. Surg Clin N Am 1996; 76: 1343-1344.
5. Siegler M, Lantos JD. Commentary: ethical justification for living donor transplantation. Cambridge Q Healthcare Ethics 1992; 1: 320-325.
6. Kjems E, Krag C. Melanoma and pregnancy. Acta Oncologica 1993; 32: 371-378[Medline].

Commentary: Management of the patient

S R D Johnston, consultant medical oncologist, ^a K Broadley, consultant in palliative medicine, ^b G Henson, consultant obstetrician, ^d C Fisher, consultant histopathologist, ^c M Henk, consultant clinical oncologist, ^a M E Gore, head of skin and melanoma unit. ^a

^a Melanoma Unit, Royal Marsden Hospital, London SW3 6JJ, ^b Palliative Care Unit, Royal Marsden Hospital, ^c Department of Pathology, Royal Marsden Hospital, ^d Department of Obstetrics and Gynaecology, Whittington Hospital, London N19 5NF

Because of the progressive metastases, systemic chemotherapy was started with single agent dacarbazine (250 mg/m² given intravenously for three days every 21 days). After two courses, which were tolerated well, metastatic spread halted. However, at 31 weeks' gestation the patient's back pain increased and was not relieved by opioid analgesics. She became confused and was hypercalcaemic (calcium 3.04 mmol/l). An elective caesarian section was performed and a live girl (1.22 kg) was delivered. Unfortunately, the mother's condition deteriorated and she died four days later. Histological examination of the placenta showed evidence of melanoma deposits, which again stained positive for S100 and HMB45. Metastases seemed to breach the single layer of syncytial cells and extend into the chorionic villous space in places. One year later the infant remains well and has no signs of metastatic melanoma.

In retrospectethical and clinical issues

Whether pregnancy influences the natural history of malignant melanoma has been debated. Primary melanomas in pregnant women were thicker than melanomas in matched control women who were not pregnant, but whether this was because of late diagnosis or hormonal or growth factor stimulation during pregnancy is unclear.¹ When corrected for tumour thickness, no adverse effect of pregnancy on overall survival in women of childbearing age treated for stage I primary cutaneous disease has been shown.¹ However, patients with metastatic melanoma confined to the regional lymph nodes (stage II) who have disease recurrence or who need treatment during pregnancy may have a shorter survival than either nulliparous or parous women with no disease recurrence during pregnancy.²

Difficult ethical and clinical issues arise. Firstly, the question of terminating the pregnancy may be raised in light of a median survival of only six months in these patients. The expectant mother may become seriously ill before term, putting the life of the fetus at risk, or she may die soon after delivery, leaving a child without a mother. Secondly, cancer treatment may affect the fetus. In our patient the risk of radiotherapy was considered low. In addition, dacarbazine is the most active chemotherapeutic agent in melanoma and can be given safely during the second and third trimesters. In Hodgkin's disease, where dacarbazine was used as part of the regimen, no long term effects were seen in children followed for up to 19 years.³ Finally, the risk of transplacental spread of melanoma to the fetus has been perceived to be low up to now.⁴ In a recent report of 16 cases of placental metastases, however, 25% of the infants developed metastatic melanoma (mainly skin and liver) at 1-8 months of age, with a mortality of 100%.⁵ Maternal factors associated with an unfavourable outcome for the infant were: age less than 30 years, nodal metastases before pregnancy, primiparity, onset of disease more than three years before pregnancy, more than three sites of metastases before the third trimester, and maternal death within one month of birth.

All these issues need to be considered when counselling women who develop regional (stage III) or metastatic (stage IV) recurrence of malignant melanoma during the early stages of pregnancy, especially in view of the difficult clinical and ethical decisions which may need to be taken if they develop progressive disease.

References

1. MacKie RM, Bufalino R, Morabito A, Sutherland C, Cascinelli N. Lack of effect of pregnancy on outcome of melanoma. Lancet 1991; 337: 653-655.
2. Shiu MH, Schottenfeld D, Maclean B, Fortner JG. Adverse effect of pregnancy on melanoma. Cancer 1976; 37: 181-187[Medline].
3. Aviles A, Diaz-Maqueo JC, Talavera A, Guzman R, Garcia EL. Longterm follow-up of 43 children exposed to chemotherapy in utero. Am J Hematol 1991; 36: 243-248[Medline].
4. Ross MI. Melanoma and pregnancy: prognostic and therapeutic considerations. Cancer Bull 1994; 46: 412-417.
5. Anderson JF, Kent S, Machin GA. Maternal malignant melanoma with placental metastases: a case report with literature review. Pediatric Pathol 1989; 9: 35-42.