Melanoma and pregnancy

INTRODUCTION — Approximately one-third of females diagnosed with melanoma are of childbearing age [1], and a 2015 Swedish population-based cancer registry study found that melanoma was the most common malignancy in pregnancy [2].

There is continuing controversy concerning the prognosis of patients diagnosed with melanoma during pregnancy. Initial concerns about pregnancy's impact on prognosis in patients diagnosed with melanoma date back to case reports which suggested that pregnancy might lead to transformation of nevi into melanomas, increase the growth rate of existing melanomas, and cause localized melanomas to metastasize [3,4]. Subsequently, multiple observations seemed to support the argument that melanoma is a hormonally responsive malignancy: changes in skin pigmentation during pregnancy, detection of hormone receptors on some melanomas using older technology, a higher incidence of melanoma after puberty, and relative immunosuppression during pregnancy [5-8].

The management of patients diagnosed with melanoma during pregnancy is likewise controversial, particularly concerning sentinel lymph node biopsy (SLNB) and decisions about the management of the patient with nodal or metastatic disease [9].

This topic will review the data that address prognosis for those diagnosed with melanoma before, during, and after pregnancy. In addition, this topic will address management, including initial biopsy of suspected melanoma, wide local excision, SLNB, and decision-making concerning the use of subsequent hormonal therapy and future pregnancy.

The general approach to cutaneous melanoma is discussed separately. (See "Surgical management of primary cutaneous melanoma or melanoma at other unusual sites" and "Overview of the management of advanced cutaneous melanoma".)

PROGNOSIS OF PREGNANCY-ASSOCIATED MELANOMA

Study limitations — Multiple studies have looked at the relationship between pregnancy and cutaneous melanoma. Factors limiting the interpretation of the literature include the following:

●Many of the case series prior to the 1980s did not account for the most important prognostic factors, such as depth of tumor or stage of disease. Subsequently, there have been a number of small case-control studies and large population-based cohort studies. While the case-control studies have the advantage of including important prognostic factors, the small numbers of patients included are an important limitation. Conversely, the larger cohort studies lack complete data on staging and Breslow depth.

●Some of the larger studies do not distinguish between diagnosis of melanoma during pregnancy and diagnosis during the postpartum period. Such studies refer to these patients as having pregnancy-associated melanoma (PAM). The definition of PAM varies in different studies and ranges from diagnosis during pregnancy to diagnosis up to five years after delivery [10].

●There is significant variability in the techniques and quality of the statistical analysis of the data between studies and in the presence of age-matched nonpregnant control groups, as well as a lack of consideration of important confounding factors, including but not limited to age, anatomic site of lesion, sun exposure or season at time of diagnosis, depth of the melanoma, absence or presence of ulceration, and presence as well as number of mitoses per mm².

Diagnosis during pregnancy — Most of the multiple small controlled studies [11-17] and large population-based cohort studies [10,18-21] do not show a negative influence of pregnancy on survival.

●An example of one population-based study is a Swedish registry-based study, which compared cause-based mortality in 1019 females with PAM (defined as melanoma diagnosis during pregnancy through the period up to two years postpartum) with females diagnosed with melanoma who did not become pregnant during a comparable period of time [10]. When focused upon females diagnosed with melanoma during pregnancy, the PAM group was comprised of 247 females, which had no significant difference in mortality (hazard ratio [HR] 0.79, 95% CI 0.44-1.41) when adjusted for pertinent prognostic factors.

●In a review of 10 case-control studies that included 185 females diagnosed with melanoma during pregnancy and 5348 females of the same childbearing age who were diagnosed with melanoma but were not pregnant, pregnancy did not have an impact on survival and did not increase the risk of a second melanoma [18]. The higher the parity and the younger the age of the mother at her first delivery, the lower the risk of melanoma. Thus, the authors concluded that there was no reason for physicians to recommend deferral of subsequent pregnancies in females who have been diagnosed with a stage I melanoma during a previous pregnancy [1].

●A controversial study is a single-institution study that compared 41 females diagnosed with PAM with a control group of females of childbearing age who were not pregnant within one year of diagnosis [17]. PAM was defined as melanoma diagnosis either during pregnancy or within one year after delivery. After adjustment for stage, age, and location, the PAM group showed a five-, seven-, and ninefold increase in mortality, metastasis, and recurrence, respectively, when compared with controls.

However, in this study, the number of relevant patients in the PAM group becomes much smaller since melanoma in situ cases comprised 50 percent of the PAM group [17]. More importantly, there is inconsistent reporting of staging information in the paper such that the authors report double the cases of advanced disease in the PAM group later in the manuscript compared with an earlier statement in the text, thus profoundly affecting their analysis due to the small number of cases.

•A single-institution study identified patients diagnosed with melanoma during pregnancy from a prospectively maintained database from January 1971 through May 2016. When 155 patients with PAM were compared with matched (by Breslow thickness, age, stage, and ulceration status), non-PAM controls, there were no significant differences in disease-free survival, overall survival, or melanoma-specific survival [22].

Diagnosis postpartum — Multiple observational studies [2,10,15,20,21,23] have generally found no influence on prognosis when melanoma is diagnosed up to five years following delivery. However, other studies suggested an increased risk of death from melanoma in the first year postpartum, which may be due to delayed diagnosis during pregnancy [2,23,24].

●In a Swedish registry study, 1019 females were diagnosed with PAM, which is defined as a melanoma diagnosis during pregnancy and up to five years postpartum [10]. When the melanoma diagnosis in each postpartum year is evaluated independently from diagnosis during pregnancy, there was no evidence of a negative impact on survival when evaluating HRs for each year postpartum; PAM groups in each year included approximately 200 to 300 cases and more than 10,000 controls.

●By contrast, a retrospective English study that linked data from a national cancer registry and hospital discharge data evaluated patients diagnosed with melanoma up to five years postpartum [23]. There was a significantly increased death rate in the first year after delivery (HR 1.92, 95% CI 1.32-2.79) but not in the four subsequent years postpartum. Another study found a lower incidence of melanoma diagnosed during pregnancy than expected compared with the first six months postpartum [2]. The spike in melanoma diagnosis and death in the early postpartum period may be caused by a delay in diagnosis.

Diagnosis prior to pregnancy — Few studies have addressed the impact on prognosis when melanoma is diagnosed before a patient becomes pregnant, but based upon the available data, there does not appear to be an effect on prognosis [11,15,18]. In a large Swedish retrospective cohort study [18], 966 females who had pregnancies after a diagnosis of a primary melanoma were compared with 4567 females who did not become pregnant after diagnosis. After adjustment for Breslow depth, tumor site, Clark level, and age, pregnancy did not significantly affect survival (HR 0.58, 95% CI 0.32-1.05).

For patients with a history of melanoma and multiple dysplastic nevi, we suggest more frequent dermatology examinations during pregnancy.

MANAGEMENT OF MELANOMA DURING PREGNANCY — The evaluation and management of the pregnant patient are similar to that of the nonpregnant patient and are based upon stage of disease. However, there are potential concerns that arise even in the initial biopsy of suspected melanoma. As the stage of disease becomes more advanced, evaluation and management decisions become more complex in order to ensure safety of the mother and the fetus.

Initial biopsy — A changing pigmented lesion during pregnancy that is clinically and dermatoscopically of concern as a possible melanoma should be biopsied immediately, as it would be in a nonpregnant patient. Excisional biopsy is the optimal way to evaluate a primary cutaneous melanoma.

The use of local anesthesia should be considered more carefully in the pregnant female. Lidocaine is considered safe for use in pregnancy "in small amounts" by the American Academy of Dermatology (AAD) 2016 Guidelines for the Use of Local Anesthesia in Office-Based Dermatologic Surgery [25]. Under the new US Food and Drug Administration (FDA) pregnancy guidelines for lidocaine, lidocaine "may use during pregnancy; risk of fetal harm not expected based on limited human data."

The addition of epinephrine as a vasoconstrictor to local anesthetic solutions has been endorsed as safe in the AAD guidelines. The local vasoconstriction minimizes serum levels of epinephrine, making a significant effect on placental vessels very unlikely [26]. Under the FDA's new pregnancy classification, epinephrine's "benefits outweigh risks during pregnancy; possible risk of teratogenicity based on limited or conflicting human data; risk of teratogenicity based on animal data at up to 30 times maximal recommended human dose; possible risk of decreased uterine perfusion based on drug's mechanism of action." However, the guidelines do caution to delay "urgent surgery" until the second trimester, if possible, and if "large amounts" of anesthesia are necessary, consultation with the patient's obstetrician should be made [25].

Wide local excision — If a wide local excision is performed under local anesthesia, the same considerations about choice of local anesthesia hold, as discussed for the initial biopsy.

If a wide local excision is to be performed under general anesthesia, the procedure should be coordinated with the obstetrician, anesthesiologist, and neonatologist. Issues concerning the role of general anesthesia during pregnancy are discussed in detail separately. (See "Anesthesia for nonobstetric surgery during pregnancy".)

Sentinel lymph node biopsy — If the pregnant patient is considered a candidate for sentinel lymph node biopsy (SLNB), there is controversy about the technique and timing of the procedure.

●Technique – There is controversy about use of blue dye alone, use of a radiocolloid (technetium-99) alone, or use of both for SLNB. Some centers refrain from using blue dye due to risk for anaphylaxis [27,28]. Likewise, the European Society for Medical Oncology (ESMO) recommends avoidance of blue dye and the use of technetium-99 alone [29]. There is accumulating evidence that the radiation exposure to the fetus is minimal, even when the injection site is as close as 5 cm to the uterus [27]. Two small series have shown safety for the fetus when utilizing a radiocolloid alone, blue dye alone, or the combination [30,31].

●Timing – The timing of SLNB is widely debated. Some advocate waiting until the second trimester to minimize exposure of the fetus to surgery and medication during pregnancy, especially during organogenesis. (See "Evaluation and management of regional nodes in primary cutaneous melanoma", section on 'SLNB timing and technique' and "Evaluation and management of regional nodes in primary cutaneous melanoma".)

In a European survey in which 290 questionnaires were completed by "melanoma physicians," almost half of respondents preferred to delay SLNB until after pregnancy [9]. This approach may be supported by a study that observed no difference in survival if the interval between wide local excision and SLNB was a median of 47 days [32]. The Yale Melanoma Unit supports this position, and recommends wide local excision under local anesthesia during pregnancy but delay of the SLNB until after delivery if general anesthesia is required [33]. However, at other institutions, this procedure is not delayed as it is viewed as important for staging, with SLNB performed during pregnancy in the second trimester and beyond [27].

If a wide local excision is to be performed under general anesthesia, the procedure should be coordinated with the obstetrician, anesthesiologist, and neonatologist. Issues concerning the role of general anesthesia during pregnancy are discussed in detail separately. (See "Anesthesia for nonobstetric surgery during pregnancy".)

Another option suggested is to perform SLNB during pregnancy under local anesthesia, although this may be difficult in certain anatomic locations, such as deep nodes in the axilla close to the thoracodorsal nerve or deep nodes in the inguinal pelvic regions [33]. Complete lymph node dissections have been performed as well, if the SLNB is positive, without reported harm to the fetus or mother [34].

Imaging studies — In the case of a patient with advanced melanoma, imaging studies may be considered. According to a Committee Opinion Summary published by the American College of Obstetrics and Gynecologists' Committee on Obstetric practice, chest radiograph with appropriate shielding, ultrasonography, and magnetic resonance imaging (MRI; preferably without gadolinium) are the techniques of choice for imaging of the pregnant patient [35]. In addition, studies such as other radiography, computed tomography (CT) scan (without contrast), and nuclear medicine imaging studies can be utilized since they are typically administered at doses that do not lead to fetal harm [35]. (See "Diagnostic imaging in pregnant and lactating patients".)

Pathology

Maternal tumor — Some studies have suggested that melanomas diagnosed during pregnancy are more often of greater Breslow depth [15,36,37], but a larger proportion of studies have not observed a significant difference [12-17,21]. Likewise, a retrospective review analyzed both clinical and pathologic characteristics of 34 melanomas diagnosed during pregnancy and up to one year after delivery, and compared these with melanomas from age- and disease-matched controls. There was no significant difference in Breslow depth, ulceration, mitotic rate, stage of disease, anatomic location of the primary tumor, histologic subtype, Clark level, regression, necrosis, or vascular invasion [38].

Placenta/fetus — While melanoma is the most common cancer to metastasize to the fetus, metastasis across the placenta to the fetus is rare and is only observed in patients with widely metastatic disease [39]. Even if placental involvement with melanoma is identified, it has been estimated that the fetus is affected in only 25 percent of these cases [40]. In cases of maternal advanced disease, it is important to alert the pathologist to perform meticulous sectioning of the placenta since many sections may be needed to detect small foci of melanoma.

Treatment — The general approach to the treatment of pregnancy-associated melanoma is based upon the same prognostic factors as for nonpregnant patients. (See "Surgical management of primary cutaneous melanoma or melanoma at other unusual sites".)

Management becomes more complex once the need for SLNB is established or if the patient has more advanced disease, and should be individualized. In patients with advanced or metastatic melanoma, certain systemic agents may be teratogenic, such as combination BRAF (eg, vemurafenib, dabrafenib, encorafenib) plus MEK (eg, cobimetinib, trametinib, binimetinib) inhibitors and checkpoint inhibitor immunotherapy (eg, nivolumab, pembrolizumab, ipilimumab) [41]. (See "Systemic treatment of metastatic melanoma lacking a BRAF mutation" and "Systemic treatment of metastatic melanoma with BRAF and other molecular alterations".)

Patients receiving these agents should avoid lactation and pregnancy during and for variable periods of time after therapy is completed, based on the agent(s) used, as follows:

●Combined BRAF plus MEK inhibitors – Patients receiving the combination of vemurafenib and cobimetinib should avoid pregnancy and lactation during therapy and for up to two weeks after the last dose of both agents.

Those receiving the combination of dabrafenib plus trametinib should avoid pregnancy and lactation during therapy, up to two weeks after the last dose of dabrafenib and up to four months after the last dose of trametinib.

Those receiving the combination of encorafenib plus binimetinib should avoid pregnancy during therapy, up to two weeks after the last dose of encorafenib and up to one month after the last dose of binimetinib. Those patients receiving this combination should also avoid lactation during therapy, up to three days after the last dose of binimetinib (based on the shorter half-life of this MEK inhibitor) and up to two weeks after the last dose of encorafenib.

●Checkpoint inhibitor immunotherapy – Patients receiving checkpoint inhibitor immunotherapy should avoid lactation and pregnancy during therapy and up to three months after the last dose of ipilimumab, during therapy and up to four months after the last dose of pembrolizumab, and during therapy and up to five months after the last dose of nivolumab.

Counseling

Future pregnancies — A future pregnancy need not be delayed in a patient who has a thin localized melanoma with an excellent prognosis.

For a patient with more advanced disease, we recommend waiting at least two to three years before becoming pregnant since this is the time period in which recurrence is most likely [42]. However, this question should be addressed on a case-by-case basis since a patient later in her childbearing years may be concerned about her fertility if she defers pregnancy. The issue becomes even more controversial in a patient who has widespread disease because her life expectancy is unclear.

The physician must provide as much information as possible about prognosis and treatment to help guide the patient and her family to the best option.

Oral contraceptives — There does not appear to be an increased risk of melanoma from the use of estrogen-progestin contraceptives, even for long durations of use, so there is no reason to withhold this type of birth control in a patient who has been diagnosed with melanoma [43].

Many observational studies have assessed the risk of developing melanoma after oral contraceptive use, and the majority observed no effect when the patients who had ever used oral contraceptives were compared with controls who never used an oral contraceptive [43]. As examples:

●A pooled analysis that included 10 case-control studies and 3796 cases of melanoma in patients who had ever used oral contraceptives found no increased risk for melanoma when compared with 9442 controls who had never taken an oral contraceptive [44]. There was no relationship found between the risk of developing melanoma and duration of prior oral contraceptive use, age at first use, or past use of oral contraceptives, even in those who had taken oral contraceptives for extended periods of time. Similar results have been seen in subsequent observational studies [45].

●However, one prospective study involving 10 European countries demonstrated a borderline-significant association between use of oral contraceptives and risk of melanoma [46].

Hormonal replacement therapy — Hormone replacement therapy (HRT) does not appear to increase the risk of melanoma and should not be withheld from a patient previously diagnosed with melanoma when no other reasonable alternative exists.

Compared with studies of melanoma and oral contraceptives, fewer studies have examined the relationship between HRT and melanoma. A majority of studies have not found an increased risk of melanoma from the use of HRT [43], but these studies have been mainly conducted in patients in the United States as opposed to Europe. These conflicting data may be related to the different formulations of estrogen utilized in Europe compared with the United States [46].

The largest body of evidence from the United States is from the Women's Health Initiative trial, which analyzed the incidence of melanoma in 27,347 postmenopausal females. The incidence of melanoma did not differ in those taking HRT (either in patients with an intact uterus who took estrogen with medroxyprogesterone or in those who had a hysterectomy and took estrogen alone) compared with a placebo group with a six-year follow-up period [47].

In contrast, multiple cohort studies conducted in Europe have demonstrated a modest increased risk in melanoma risk associated with menopausal hormone therapy [46,48-50].

In vitro fertilization — There is a paucity of high-quality data concerning the risk of melanoma associated with in vitro fertilization (IVF).One review did not observe a consistent association between IVF and malignant melanoma among all infertile patients. However, in ever-parous patients treated with IVF, the authors suggested there may be increased risk for melanoma [51].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Melanoma screening, prevention, diagnosis, and management".)

SUMMARY

●Patients with pregnancy-associated melanoma do not seem to have a worse prognosis than those with melanoma not diagnosed around the time of pregnancy. Their prognosis is based upon established prognostic factors. (See 'Prognosis of pregnancy-associated melanoma' above.)

●A changing pigmented lesion during pregnancy that is clinically and dermatoscopically of concern as a possible melanoma should be biopsied immediately, as it would be in a nonpregnant patient. (See 'Initial biopsy' above.)

●The evaluation and treatment of the pregnant patient with melanoma should generally be similar to that of patients who are not pregnant. Special monitoring may be needed when performing wide local excision late in pregnancy, and the timing of sentinel lymph node biopsy may need to be altered. (See 'Wide local excision' above and 'Sentinel lymph node biopsy' above.)

●Melanomas in pregnant patients do not seem to have clinicopathologic characteristics that differ from the nonpregnant patient. While melanoma is the most common malignancy to metastasize to the fetus, it is still very rare and typically seen in patients with widely metastatic disease. (See 'Pathology' above.)

●The patient with a thin melanoma with excellent prognosis need not delay future pregnancies or avoid use of oral contraceptives or hormone replacement therapy (HRT), if the latter are indicated. (See 'Counseling' above.)

•Decision-making becomes much more complex in the patient with a more uncertain prognosis where a delay in future pregnancy may be considered, but this should be evaluated on a case-by-case basis.

•There are no conclusive data concerning in vitro fertilization (IVF) and risk of melanoma.

●Systemic agents used in advanced or metastatic melanoma may be teratogenic, such as combination BRAF plus MEK inhibitors and checkpoint inhibitor immunotherapy. Patients should avoid pregnancy and lactation during therapy and for variable periods of time after such therapy is completed, based on the agent(s) used. (See 'Treatment' above.)