Abstract
Background
Melanoma is the most common malignancy encountered during pregnancy. Conflicting data have led to ongoing confusion regarding pregnancy associated melanoma (PAM) in the media and among the public. The objective of this study was to better characterize both the clinical presentation of PAM and its prognostic implications.
Study Design
Female patients of reproductive age with stage 0-IV cutaneous melanoma were identified from our prospectively maintained database. Clinical and histopathologic factors were analyzed with appropriate statistical methods. Univariable and then multivariable analysis were utilized on matched data to compare disease free survival (DFS), overall survival (OS) and melanoma specific survival (MSS) for stage 0-III PAMs vs non-PAMs. Kaplan-Meier survival curves were then plotted for OS and MSS and compared using the log-rank test.
Results
Clinical presentation of melanoma was similar for PAM and non-PAM patients. There was no significant difference in recurrence between the two groups; for PAMs, 38.5% of patients recurred as compared to 36.6% of non-PAMs (p=0.641). For PAMs, median follow up was 14.6 years (range 0–42.6 years) and 11.1 years (0 – 48.5 years) for the non-PAMs. No significant differences in DFS, MSS, or OS were identified on univariable or multivariable analysis for PAM vs non-PAMs in stage 0/I/II and stage III cutaneous melanoma respectively (p=0.880 DFS, p=0.219 OS and p=0.670 MSS).
Conclusions
We observed no difference in DFS, OS or MSS between the two groups. Pregnant patients should be screened for melanoma in a similar manner to non-pregnant patients and should be counseled that their survival is not adversely affected by their pregnancy.
Keywords: Melanoma, Pregnancy, overall survival, melanoma specific survival, disease free survival
Introduction
Melanoma is the most common malignancy encountered during pregnancy, accounting for 31% of all malignancies in the intrapartum period. (1, 2) For many years, pregnancy has been thought to have an adverse effect on the course of melanoma. Reports beginning in the 1950s suggested pregnancy increased the risk of melanoma development, metastasis and recurrence. (3–6) Many hypotheses since then have been formulated, linking worsened outcomes to hyperpigmentation, relative immunosuppression, and hormone binding of melanocytes. (7–10) Given the overall increase in melanoma incidence in women of childbearing age in the US, this topic has become even more pertinent. (11, 12) Editorials, systemic reviews, and media coverage of PAM persist but fail to draw definitive conclusions despite many years of attention and underpowered studies. (13–21) Many of the adequately powered studies that do exist come from large, non-US based registries, with a resultant lack of granular detail and reliability. (22–25) The primary objective of this study was to query our large, single institution melanoma database to better characterize PAM with particular attention to overall survival (OS) and melanoma specific survival (MSS). Secondarily, we examined other clinical factors with regard to melanoma mortality and such as parity and gravidity, in addition to known prognostic factors such as age, stage, histologic type, Breslow thickness, and ulceration.
Methods
Female patients of reproductive age (18–50) with American Joint Committee on Cancer (AJCC) stage 0-IV cutaneous melanoma were identified from the prospectively maintained John Wayne Cancer Institute melanoma database between January 1971 and May 2016. All patient data were de-identified, and this study was independently confirmed to be exempt from Institutional Review Board review. Melanomas were staged by seventh edition AJCC criteria. (26–28) In order to ensure adequate staging, patients without lymph node staging for melanomas with Breslow thickness ≥0.75 mm were excluded from analysis (n=540 non-PAM, n=43 PAM). Pregnancy associated melanoma is a field derived either from patient questionnaire responses (self reported) or direct physician queries (physician reported). The JWCI melanoma database defines PAMs by an affirmative response to “did melanoma develop during pregnancy.” This includes cases that developed de novo during pregnancy or melanomas that arose from pre-existing lesions that changed during pregnancy. We cannot exclude the possibility that some of these lesions were identified incidentally during prenatal visits. Laboratory pregnancy confirmation is incomplete in this data set as patients receiving office based excisions would not have routinely received urine or serum pregnancy evaluations. For this reason, we are not able to comment on the women who were deemed pregnant based on preoperative bHCG alone. Clinical and histopathologic factors were examined between PAM and non-PAM groups. T-test was used to analyze age at diagnosis, parity, gravidity, and Breslow thickness. Chi-square test was used for Clark level, anatomic site, ulceration, sentinel lymph node examination status, recurrence status, type of first recurrence, stage at diagnosis, and stage first seen at JWCI. A 1:1 matched pair sample was then created using pairs of PAM and non-PAM patients who were matched for Breslow thickness, age, stage and ulceration status. With respect to age, we matched utilizing the following categories: <25, 25 – =35, ≥3. With respect to Breslow thickness, we matched for categories: ≤ 0.75, 0.75 – <2.00, 2.00 – ≤4.00, > 4.00 and unknown. For stage at diagnosis, we matched using categories: 0, I/II, and III. Finally, for ulceration, we matched using categories: yes, no and unknown. These details have been incorporated in to the manuscript on page 4. Univariable and then multivariable analyses were conducted with the matched data to analyze DFS, OS and MSS for patients with stage 0/I/II and stage III cutaneous melanoma at diagnosis. Due to the paucity of PAM patients with Stage IV disease at diagnosis (n=1), those patients were excluded from this analysis. Kaplan-Meier survival curves were then plotted for OS and MSS and compared using the log-rank test. SAS software, version 9.3 (SAS Institute) was used for all analyses. A two-sided p-value of 0.05 or lower was considered to indicate statistical significance.
Results
Of the entire patient cohort (n=2025), 156 women (7.7%) with PAM were identified after selection criteria applied. No cases of transplacental transfer of melanoma were identified. Clinical presentation of melanoma was similar for PAM and non-PAM patients with no significant differences in Breslow thickness (1.30 mm vs. 1.34 mm, p=0.737), histologic type, or primary tumor site (see Table 1). Age was greater in the non-PAM patients (36.8 vs 31.7 years, p<0.001). There was also no significant difference in stage at diagnosis. (Table 1) Parity was significantly increased in the PAM group, p=0.010, as was gravidity, p<0.001 At 10-years; disease-free survival was 65.7% and 62.3% for the non-PAM and PAM groups, respectively (p=0.8934). Mean disease-free survival was also similar at 24.48 years in the non-PAM group and 20.65 years in the PAM group.
Table 1.
Whole Group Frequency
| Non-PAM | PAM | p Value | |
|---|---|---|---|
| n | 2,025 | 156 | |
| Age at diagnosis, y, mean ± SD | 36.76 ± 8.43 | 31.69 ± 8.69 | <0.001 |
| Age, n (%) | <0.001 | ||
| <25 y | 215 (10.62) | 20 (12.82) | |
| 25 – ≤35 y | 611 (30.17) | 101 (64.74) | |
| ≥35 y | 1199 (59.21) | 35 (22.44) | |
| Stage at diagnosis, n (%) | 0.186 | ||
| 0 | 138 (6.81) | 5 (3.21) | |
| I/II | 1647 (81.33) | 134 (85.90) | |
| III | 240 (11.85) | 17 (10.90) | |
| Primary site, n (%) | 0.100 | ||
| Head/neck | 193 (9.53) | 8 (5.13) | |
| Trunk | 733 (36.20) | 66 (42.31) | |
| Extremity | 1099 (54.27) | 82 (52.56) | |
| Breslow thickness, mm, n (%) | 0.311 | ||
| ≤ 0.75 | 711 (35.11) | 55 (35.26) | |
| 0.75 – <2.00 | 518 (25.58) | 43 (27.56) | |
| 2.00 – ≤4.00 | 177 (8.74) | 20 (12.82) | |
| > 4.00 | 75 (3.70) | 5 (3.21) | |
| Unknown | 544 (26.86) | 33 (21.15) | |
| Breslow thickness (mm), continuous, mean ± SD | 1.34 ± 1.58 | 1.30 ± 1.21 | 0.737 |
| Para, mean ± SD | 1.36 ± 1.46 | 1.68 ± 1.44 | <0.001 |
| Gravida, mean ± SD | 1.78 ± 1.71 | 2.29 ± 1.51 | <0.001 |
| Ulceration, n (%) | 0.849 | ||
| Yes | 173 (8.54) | 12 (7.69) | |
| No | 1381 (68.20) | 105 (67.31) | |
| Unknown | 471 (23.26) | 39 (25.00) | |
| Clark level, n (%) | 0.025 | ||
| I | 183 (9.04) | 9 (5.77) | |
| II | 546 (26.96) | 43 (27.56) | |
| III | 493 (24.35) | 37 (23.72) | |
| IV | 495 (24.44) | 54 (34.62) | |
| V | 51 (2.52) | 1 (0.64) | |
| Unknown | 257 (12.69) | 12 (7.69) | |
| Histologic type, n (%) | 0.228 | ||
| ALM/LMM | 38 (1.88) | 1 (0.64) | |
| In situ | 138 (6.81) | 5 (3.21) | |
| NM | 241 (11.90) | 21 (13.46) | |
| SSM | 972 (48.00) | 84 (53.85) | |
| Others | 78 (3.85) | 3 (1.92) | |
| Unknown | 558 (27.56) | 42 (26.92) | |
| Lymph nodes examined, n (%) | 0.540 | ||
| Positive | 248 (12.25) | 18 (11.54) | |
| Negative | 830 (40.99) | 71 (45.51) | |
| Not done | 947 (46.77) | 67 (42.95) | |
| Type of first recurrence, n (%) | 0.656 | ||
| None | 1284 (63.47) | 96 (61.54) | |
| Nodal | 412 (20.37) | 38 (24.36) | |
| Local/in-transit | 125 (6.18) | 8 (5.13) | |
| Distant | 202 (9.99) | 14 (8.97) |
PAM, pregnancy associated melanoma; ALM, acral lentiginous melanoma; LMM, lentigo maligna melanoma; NM, nodular melanoma; SSM, superficial spreading melanoma
Matched Pair Sample
In an attempt to decrease potential biases associated with delay in diagnosis of PAMs, we created a matched pair sample. Each PAM patient was matched with a non-PAM patient by Breslow thickness, age, stage and ulceration status. In this group of 310 patients (155 matched pairs), PAMs had a median follow-up of 14.6 years (range 0–42.6 years), and non-PAMs had a median follow-up of 11.1 years (0 – 48.5 years). When comparing the PAM and non-PAM matched pairs, clinical features were similar (see Table 2) except for histologic type (p=0.046), primary site (p=0.040), gravidity and parity (p<0.001, p<0.001). There were also no differences in recurrence between the PAM and non-PAM group (Table 2).
Table 2.
Matched Sample Frequency
| Non-PAM | PAM | p Value | |
|---|---|---|---|
| n | 155 | 155 | |
| Age,y, n (%) | |||
| <25 | 20 (12.90) | 20 (12.90) | |
| 25 – ≤35 | 100 (64.52) | 100 (64.52) | |
| ≥35 | 35 (22.58) | 35 (22.58) | |
| Stage at diagnosis, n (%) | |||
| 0/I/II | 139 (89.68) | 139 (89.68) | |
| III | 16 (10.32) | 16 (10.32) | |
| Breslow thickness, mm, n (%) | |||
| ≤ 0.75 | 55 (35.48) | 55 (35.48) | |
| 0.75 – <2.00 | 43 (27.74) | 43 (27.74) | |
| 2.00 – ≤4.00 | 20 (12.90) | 20 (12.90) | |
| > 4.00 | 5 (3.23) | 5 (3.23) | |
| Ulceration, n (%) | |||
| Yes | 12 (7.74) | 12 (7.74) | |
| No | 104 (67.10) | 104 (67.10) | |
| Unknown | 39 (25.16) | 39 (25.16) | |
| Age at diagnosis , y, mean ± SD | 32.17 ± 6.79 | 31.72 ±8.71 | 0.616 |
| Stage first seen, n (%) | 0.137 | ||
| 0 | 14 (9.09) | 5 (3.25) | |
| I/II | 94 (61.04) | 108 (70.13) | |
| III | 31 (20.13) | 28 (18.18) | |
| IV | 15 (9.74) | 13 (8.44) | |
| Unknown | 1 (0.01) | 1 (0.01) | |
| Primary site, n (%) | 0.040 | ||
| Head/neck | 21 (13.55) | 8 (5.16) | |
| Trunk | 61 (39.35) | 66 (42.58) | |
| Extremity | 73 (47.10) | 81 (52.26) | |
| Breslow thickness, mm, continuous, mean ± SD | 1.48 ± 1.94 | 1.30 ± 1.21 | 0.389 |
| Para, mean ± SD | 0.89 ± 1.22 | 1.68 ± 1.45 | <0.001 |
| Gravida, mean ± SD | 1.24 ± 1.45 | 2.30 ± 1.52 | <0.001 |
| Clark level, n (%) | 0.258 | ||
| I | 11 (7.10) | 9 (5.81) | |
| II | 43 (27.74) | 43 (27.74) | |
| III | 39 (25.16) | 37 (23.87) | |
| IV | 38 (24.52) | 53 (34.19) | |
| V | 4 (2.58) | 1 (0.65) | |
| Unknown | 20 (12.90) | 12 (7.74) | |
| Histologic type, n (%) | 0.046 | ||
| ALM/LMM | 2 (1.29) | 1 (0.65) | |
| In situ | 14 (9.03) | 5 (3.32) | |
| NM | 23 (14.84) | 21 (13.55) | |
| SSM | 59 (38.06) | 84 (54.19) | |
| Others | 2 (1.29) | 3 (1.94) | |
| Unknown | 55 (35.48) | 41 (26.45) | |
| Lymph nodes examined, n (%) | 0.627 | ||
| Positive | 17 (10.97) | 17 (10.97) | |
| Negative | 79 (50.97) | 71 (45.81) | |
| Not done | 59 (38.06) | 67 (43.23) |
PAM, pregnancy associated melanoma; ALM, acral lentiginous melanoma; LMM, lentigo maligna melanoma; NM, nodular melanoma; SSM, superficial spreading melanoma
Univariable Survival Analysis by Stage
For matched patients with stage 0/I/II melanoma at diagnosis, there were no differences identified in DFS (p=0.880, HR 0.97, 95% CI 0.62–1.5), OS (p=0.219, HR 0.73, 95% CI 0.44–1.21) or MSS (p=0.670, HR 0.89, 95% CI 0.51–1.53) for PAMs. For those patients with stage III melanoma at diagnosis, there were no differences in DFS (p=0.858, HR 0.94, 95% CI 0.46–1.93) OS (p=0.365, HR 0.64 95% CI 0.24–1.69) or MSS (p=0.595, HR 0.75, 95% CI 0.26–2.17) for PAMs. Concordant results were obtained when Kaplan-Meier curves were plotted and compared using log-rank test for OS and MSS for both stage 0/I/II at diagnosis (Figure 1) and stage III at diagnosis (data not shown).
Figure 1.
Kaplan-Meier survival curves from stage 0, I, and II matched date, for (A) overall survival and (B) melanoma specific survival. p Values calculated by log rank test. PAM, pregnancy-associated melanoma.
Expected differences in DFS, OS and MSS with regard to increasing Breslow thickness, stage first seen, ulceration, Clark level, histologic type and recurrence data (Tables 3–5) were identified for patients with stage 0/I/II melanoma at diagnosis. For patients with stage III melanoma at diagnosis, none of the clinicopathologic factors assessed were significantly associated with significant DFS, OS or MSS differences except for ulceration.
Table 3.
Overall Survival Stage 0/I/II at diagnosis: Matched Data
| Univariable | Multivariable | |||
|---|---|---|---|---|
| p Value | Hazard ratio (95% CI) | p Value | Hazard ratio (95% CI) | |
| Age in years (increasing) | 0.103 | 1.03 (0.99, 1.06) | ||
| Location | ||||
| Extremity (Reference) | ||||
| Head/Neck | 0.440 | 1.52 (0.53, 4.37) | ||
| Trunk | 0.215 | 1.43 (0.81, 2.50) | ||
| Breslow thickness (mm) ≤ 0.75 (Reference) | ||||
| 0.75 – <2.00 | 0.020 | 2.68 (1.17, 6.14) | 0.067 | 2.05 (0.95, 4.39) |
| 2.00 – ≤4.00 | <0.001 | 7.59 (3.23, 17.83) | <0.001 | 4.73 (2.10, 10.66) |
| >4.00 | <0.001 | 40.03 (8.2,4 194.46) | <0.001 | 49.20 (10.24, 236.39) |
| Unknown | 0.018 | 2.75 (1.19, 6.36) | 0.067 | 1.98 (0.95, 4.12) |
| Breslow thickness (mm), continuous | <0.001 | 1.74 (1.40, 2.18) | ||
| Para | 0.562 | 1.05 (0.88, 1.26) | ||
| Gravida | 0.042 | 1.19 (1.01, 1.40) | ||
| PAM | 0.219 | 0.73 (0.44, 1.21) | ||
| Stage first seen | ||||
| 0 | NA* | NA* | ||
| I/II (Reference) | ||||
| III | <0.001 | 4.91 (2.53, 9.56) | ||
| IV | <0.001 | 17.45 (9.05, 33.62) | ||
| Ulceration | ||||
| Yes | <0.001 | 8.49 (3.79, 19.02) | <0.001 | 5.19 (2.18, 12.34) |
| No (Reference) | ||||
| Unknown | 0.009 | 2.20 (1.22, 3.97) | <0.001 | 3.18 (1.82, 5.55) |
| Clark Level | ||||
| I (Reference) | ||||
| II | 0.113 | 5.23 (0.68, 40.50) | ||
| III | 0.018 | 11.26 (1.50, 84.28) | ||
| IV | 0.007 | 16.50 (2.17, 125.44) | ||
| V | NA* | |||
| Unknown | <0.001 | 45.88 (5.74, 366.92) | ||
| Histologic type | ||||
| ALM/LMM | 0.86 | 0.83 (0.11, 6.54) | ||
| In Situ | NA* | |||
| NM (Reference) | ||||
| SSM | 0.003 | 0.32 (0.15, 0.67) | ||
| Others | 0.448 | 0.45 (0.06, 3.53) | ||
| Unknown | 0.413 | 0.73 (0.34, 1.56) | ||
NA, Not applicable: analysis not available due to small sample size.
PAM, pregnancy associated melanoma; ALM, acral lentiginous melanoma; LMM, lentigo maligna melanoma; NM, nodular melanoma; SSM, superficial spreading melanoma
Table 5.
Disease-Free Survival Stage 0/II/II at Diagnosis: Matched Data
| Univariable | Multivariable | |||
|---|---|---|---|---|
| p Value | Hazard ratio (95% CI) | p Value | Hazard ratio (95% CI) | |
| Age in years (increasing) | 0.739 | 1.01 (0.97 – 1.04) | ||
| Location | ||||
| Extremity (Reference) | ||||
| Head/neck | 0.007 | 2.57 (1.30–5.09) | ||
| Trunk | 0.027 | 1.69 (1.06–2.70) | 0.007 | 1.91 (1.19–3.06) |
| Breslow thickness (mm) (continuous) | <0.001 | 1.74 (1.43–2.13) | ||
| Breslow category (mm) | ||||
| ≤ 0.75 (Reference) | ||||
| 0.75 – <2.00 | 0.021 | 2.18 (1.12–4.23) | 0.009 | 2.47 (1.25–4.89) |
| 2.00 – ≤4.00 | <0.001 | 5.51 (2.72–11.15) | <0.001 | 4.28 (1.87–9.85) |
| > 4.00 | <0.001 | 18.60 (4.17–82.85) | <0.001 | 22.50 (4.86–104.21) |
| Unknown | <0.001 | 4.22 (2.26–7.87) | <0.001 | 3.95 (2.01–7.74) |
| Para | 0.330 | 1.07 (0.931–1.24) | ||
| Gravida | 0.026 | 1.18 (1.02–1.36) | ||
| PAM | 0.880 | 0.97 (0.62–1.50) | ||
| Stage first seen | ||||
| 0 | 0.500 | 0.50 (0.07–3.71) | ||
| I/II (Reference) | ||||
| III | <0.001 | 12.78 (7.5–21.77) | ||
| IV | <0.001 | 12.15 (7.00–21.11) | ||
| Ulceration | ||||
| Yes | <0.001 | 4.19 (1.86–9.45) | ||
| No (Reference) | ||||
| Unknown | <0.001 | 2.92 (1.86–4.60) | ||
| Clark Level | ||||
| I | NA* | |||
| II (Reference) | ||||
| III | 0.068 | 2.01 (0.95–4.26) | ||
| IV | <0.001 | 3.90 (1.95–7.82) | ||
| V | 0.176 | 4.12 (0.53–32.08) | ||
| Unknown | <0.001 | 8.88 (4.34–18.15) | ||
| Histologic type | ||||
| ALM/LMM | NA* | |||
| NM (Reference) | ||||
| In Situ | 0.022 | 0.09 (0.01–0.71) | ||
| SSM | <0.001 | 0.22 (0.12–0.43) | ||
| Others | 0.070 | 3.15 (0.91–10.82) | ||
| Unknown | 0.619 | 0.87 (0.49–1.53) | ||
NA, sample size too small for comparison.
PAM, pregnancy associated melanoma; ALM, acral lentiginous melanoma; LMM, lentigo maligna melanoma; NM, nodular melanoma; SSM, superficial spreading melanoma
Univariable Reproductive History and Survival Differences
Interestingly, increasing gravidity was associated with worse DFS (p=0.026, HR 1.18, 95% CI 1.02–1.36), OS (p=0.042 HR 1.19 95% CI 1.01–1.40) and MSS (p=0.034, HR 1.21 95% CI 1.02–1.44) for stage 0/I/II melanomas, but these differences were not seen for stage III patients. Conversely, when parity was examined, there were no significant differences in DFS (p=0.330, HR 1.07, 95% CI 0.93–1.24), OS (p=0.562, HR 1.05 95% CI 0.88–1.26) or MSS (p=0.556 HR 1.06 95% CI 0.88–1.28) for stage 0/I/II melanomas. There were also no significant differences in OS or MSS for stage III patients with respect to parity as an independent prognostic factor.
Multivariable Survival Analysis by Stage
For patients with stage 0/I/II or stage III melanoma at diagnosis, there were no significant differences identified in DFS, OS or MSS for PAMs in the model (Tables 3–5, not shown). The only factors associated with significant differences in OS were ulceration (yes or unknown) and increasing Breslow thickness for stage 0/I/II melanoma. (Table 3) For MSS, factors associated with significant differences in OS were trunk as primary site (p=0.016, HR 2.22, 95% CI 1.16–4.23), Breslow thickness 2-≤4 and >4 (p<0.001 HR 5.62, 2.05–15.35, and p<0.001, HR141.76, 95% CI 21.47–936.05), ulceration (yes: p<0.001, HR 6.83 95% CI 2.38–9.96; unknown: p<0.01, HR 4.87, 95% CI 1.27–5.15) and increasing gravidity (p=0.027, HR 1.24, 95% CI 1.03–1.51). For DFS, the only factors associated with significant differences on multivariable analysis were also expected: increasing Breslow thickness, trunk as the primary site (p=0.007, HR 1.91, 95% CI 1.19–3.06) and histologic type (SSM, In situ, and other) for stage 0/I/II melanoma. (See Table 5) There were no factors associated with significant differences in Stage III patients with respect to DFS, OS and MSS on multivariable analysis.
Discussion
Despite previous study, controversy remains for women of childbearing age at risk for or diagnosed with cutaneous melanoma. In our institutional analysis, the incidence of PAM was 7.7%, congruent with previous work from this institution identifying a 8.7 percent incidence of PAM. (29) In early 2016, Tellez and colleagues reported that women with melanoma arising during or within one year of childbirth had a significantly worse prognosis when compared to their non-pregnant counterparts. (30) This study was limited by its small, retrospective nature; additionally, only 19 of 41 patients in the study had melanoma diagnosed during pregnancy. Based on multivariable analysis of this limited sample, the authors reported a 9-fold increase in recurrence (odds ratio (OR) 9.30, p=0.01), a 7-fold increase in metastasis (OR 6.70, p=0.01) and a 5-fold increase in mortality (OR 5.10, p=0.03). (30)
Larger studies have been published, however, that refute this notion of diminished survival for PAMs. In 2004, O’Meara et al. evaluated 412 women with PAM from California maternal and neonatal discharge records linked to California Cancer Registry data. After controlling for age, race, stage, and tumor thickness, pregnancy had no impact on survival in women with melanoma. This held true for those during pregnancy (p=0.570) or in the postpartum period (p=0.162). (23) Johansson and colleagues used Swedish Cancer and Multi-Generation Registers to conduct a population-based cohort study examining 1019 PAMs. However, the definition of PAM in this study was broad, including patients with melanoma arising during pregnancy or up to two years following delivery. (24) This group showed no significant difference in survival (p=0.47) for PAMs vs non-PAMs in this study of women aged 15–44. (24) Similarly, in 2004, Lens et al. examined 185 PAMs using data from the Swedish National and Regional Registries. This group also found no significant difference in overall survival between PAMs and non-PAMs (p=0.84, log rank test). At multivariable cox regression, pregnancy at the time of melanoma diagnosis was not related to increased risk of death (p=0.804, HR 1.08, 95% CI 0.60–1.93). Prognostic factors for these women were Breslow thickness (p<0.001, HR 2.16 95% CI 1.80–2.58) and axial vs limb site of the primary melanoma (p<0.001, HR 2.51 95% CI 1.78–3.56). (22)
In this study, as expected, Breslow thickness was associated with worsened OS and MSS in Stage 0/I/II patients. (Tables 3 and 4) A criticism of previous studies linking melanoma with abject survival outcomes was the preponderance of thicker tumors in PAM patients, possibly due to delay in diagnosis. Travers et al., in 1995, queried patient records from the Massachusetts General Hospital and identified 45 women with PAM. This group reported significantly greater mean tumor thickness in PAMs than non-PAMs (2.28 vs 1.22 mm, p<0.007). (31) While the authors stated that “the mechanisms by which pregnancy may lead to increased thickness of melanoma remain unclear…” they do acknowledge that a delay in diagnosis may have led to the thicker lesions identified in PAM cases. (31) When examining incidence and outcomes of 577 PAMs from registry data in New South Wales, Bannister-Tyrrell and colleagues showed that melanomas diagnosed in pregnancy were thicker (median = 0.75 mm) than melanomas diagnosed postpartum (median = 0.60 mm, p = 0.002). (25) When we controlled for this potential confounder by matching for Breslow thickness, no difference in survival outcomes were identified for PAMs as compared to non-PAMs. Additionally, by excluding patients without lymph node staging for lesions ≥ 0.75 mm in Breslow thickness, our group limited the proportion of patients receiving non-standard care. Mackie et al. reported similar findings with regard to Breslow thickness in a study of 388 women of childbearing age with melanoma (92 treated during pregnancy) from the World Health Organization melanoma programme. These authors concluded that Breslow thickness was greater in women who were diagnosed while pregnant (p=0.002), however once Breslow thickness was controlled for, survival rates did not differ (p=0.1). (32)
Table 4.
Melanoma-Specific Survival Stage 0/I/II at Diagnosis: Matched Data
| Univariable | Multivariable | |||
|---|---|---|---|---|
| p Value | Hazard ratio (95% CI) | p Value | Hazard ratio (95% CI) | |
| Age in years (increasing) | 0.189 | 1.03 (0.99, 1.06) | ||
| Location | ||||
| Extremity (Reference) | ||||
| Head/neck | 0.261 | 1.85 (0.63, 5.42) | ||
| Trunk | 0.157 | 1.54 (0.85, 2.81) | 0.016 | 2.22 (1.16, 4.23) |
| Breslow thickness (mm), continuous | <0.001 | 1.77 (1.39, 2.24) | ||
| Breslow category (mm) | ||||
| ≤ 0.75 (Reference) | ||||
| 0.75 – <2.00 | 0.015 | 3.07 (1.24, 7.62) | ||
| 2.00 – ≤4.00 | <0.001 | 8.45 (3.32, 21.49) | <0.001 | 5.62 (2.05, 15.38) |
| > 4.00 | <0.001 | 44.69 (8.89, 224.68) | <0.001 | 141.76 (21.47, 936.05) |
| Unknown | 0.015 | 3.20 (1.26, 8.13) | ||
| Para | 0.556 | 1.06 (0.88, 1.28) | ||
| Gravida | 0.034 | 1.21 (1.02, 1.44) | 0.027 | 1.24 (1.03, 1.51) |
| PAM | 0.670 | 0.89 (0.51, 1.53) | ||
| Stage first seen | ||||
| 0 | NA* | |||
| I/II (Reference) | ||||
| III | <0.001 | 6.63 (3.27, 13.48) | ||
| IV | <0.001 | 21.18 (10.59, 42.36) | ||
| Ulceration | ||||
| Yes | <0.001 | 10.12 (4.42, 23.17) | <0.001 | 6.83 (2.38 9.96) |
| No (Reference) | ||||
| Unknown | 0.007 | 2.38 (1.26, 4.47) | <0.001 | 4.87 (1.27, 5.15) |
| Clark Level | ||||
| I | NA* | |||
| II (Reference) | ||||
| III | 0.040 | 2.35 (1.04, 5.32) | ||
| IV | 0.003 | 3.56 (1.52, 8.34) | ||
| V | NA* | |||
| Unknown | <0.001 | 11.30 (4.35, 29.34) | ||
| Histologic type | ||||
| NM (Reference) | ||||
| SSM | 0.004 | 0.31 (0.14, 0.68) | ||
| Others | 0.534 | 0.52 (0.07, 4.10) | ||
| Unknown | 0.487 | 0.76 (0.34, 1.68) | ||
PAM, pregnancy associated melanoma; ALM, acral lentiginous melanoma; LMM, lentigo maligna melanoma; NM, nodular melanoma; SSM,
superficial spreading melanoma
To our knowledge, this is the first study associating increased gravidity to worsened survival outcomes in women of childbearing age with stage 0/I/II melanoma. Mechanisms underlying this finding could be associated with hormonal differences or other un-characterized risk factors associated with gravidity. Future study will be needed to fully examine this association in a prospective manner. Also of note, while we do not have consistent or reliably documented follow up for babies born to these mothers with PAM, no cases of transplacental melanoma have been identified in the JWCI melanoma database comprising approximately 15,600 patients to date. Compared to larger registry-based studies, our data comes from a well-maintained and reliable single institution database with extensive follow up. Additionally, we excluded patients with melanomas ≥ 0.75 mm thick who did not receive nodal staging, thereby limiting non-standard care and inaccurately staged lesions. In contrast to previous studies that were not matched for pertinent factors such as Breslow thickness, age, and stage; our study was not clouded by these known prognostic factors.
Limitations of this study include its retrospective nature and treatment-related biases inherent to our referral center. Additionally, the majority of patients evaluated exhibited favorable prognostic factors and few PAMs had stage III disease or thick primary tumors (n=17 and n=5 respectively). Finally, due to the time period over which this data was collected, some subjects did not have complete data with regard to histology, ulceration and lymph node staging exam, which introduces some amount of uncertainty with regards to their initial prognosis. Despite these limitations, we believe this work accurately reflects PAM incidence and outcomes.
Conclusions
This, the largest, single-institution study examining the characteristics and outcomes associated with melanoma arising during pregnancy demonstrates no significant difference in DFS, OS or MSS. Pregnant patients should be screened for melanoma in a similar manner to non-pregnant patients and should be counseled that their prognosis is not adversely affected by pregnancy.
Acknowledgments
The authors are grateful for the editorial assistance of Gwen Berry.
Support: This study was supported by grants (CA189163 and CA29605) from the National Cancer Institute and by funding from the Amyx Foundation, Inc, the Borstein Family Foundation, Dr Miriam and Sheldon G Adelson Medical Research Foundation, and the John Wayne Cancer Institute Auxiliary. Dr Maris S Jones is the Harold McAlister Charitable Foundation Fellow.
Footnotes
Disclosure Information: Nothing to disclose.
Presented at the Western Surgical Association 124th Scientific Session, Coronado, CA, November 2016.
Disclaimer: The content of this report is solely the responsibility of the authors and does not necessarily represent the official view of the National Cancer Institute or the NIH.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Stensheim H, Moller B, van Dijk T, Fossa SD. Cause-specific survival for women diagnosed with cancer during pregnancy or lactation: a registry-based cohort study. J Clin Oncol. 2009;27:45–51. doi: 10.1200/JCO.2008.17.4110. [DOI] [PubMed] [Google Scholar]
- 2.Lee YY, Roberts CL, Dobbins T, et al. Incidence and outcomes of pregnancy-associated cancer in Australia, 1994–2008: a population-based linkage study. BJOG. 2012;119:1572–1582. doi: 10.1111/j.1471-0528.2012.03475.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Pack GT, Scharnagel IM. The prognosis for malignant melanoma in the pregnant woman. Cancer. 1951;4:324–334. doi: 10.1002/1097-0142(195103)4:2<324::aid-cncr2820040218>3.0.co;2-g. [DOI] [PubMed] [Google Scholar]
- 4.Shiu MH, Schottenfeld D, Maclean B, Fortner JG. Adverse effect of pregnancy on melanoma: a reappraisal. Cancer. 1976;37:181–187. doi: 10.1002/1097-0142(197601)37:1<181::aid-cncr2820370126>3.0.co;2-f. [DOI] [PubMed] [Google Scholar]
- 5.Byrd BF, Jr, Mc GW. The effect of pregnancy on the clinical course of malignant melanoma. Southern Med J. 1954;47:196–200. doi: 10.1097/00007611-195403000-00002. [DOI] [PubMed] [Google Scholar]
- 6.Slingluff CL, Jr, Reintgen DS, Vollmer RT, Seigler HF. Malignant melanoma arising during pregnancy. A study of 100 patients. Ann Surg. 1990;211:552–557. doi: 10.1097/00000658-199005000-00005. discussion 558–559. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Enninga EA, Holtan SG, Creedon DJ, et al. Immunomodulatory effects of sex hormones: requirements for pregnancy and relevance in melanoma. Mayo Clinic Proc. 2014;89:520–535. doi: 10.1016/j.mayocp.2014.01.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Luppi P. How immune mechanisms are affected by pregnancy. Vaccine. 2003;21:3352–3357. doi: 10.1016/s0264-410x(03)00331-1. [DOI] [PubMed] [Google Scholar]
- 9.Gupta A, Driscoll MS. Do hormones influence melanoma? Facts and controversies. Clin Dermatol. 2010;28:287–292. doi: 10.1016/j.clindermatol.2010.04.003. [DOI] [PubMed] [Google Scholar]
- 10.Lea CS, Holly EA, Hartge P, et al. Reproductive risk factors for cutaneous melanoma in women: a case-control study. Am J Epidemiol. 2007;165:505–513. doi: 10.1093/aje/kwk040. [DOI] [PubMed] [Google Scholar]
- 11.Broer N, Buonocore S, Goldberg C, et al. A proposal for the timing of management of patients with melanoma presenting during pregnancy. J Surg Oncol. 2012;106:36–40. doi: 10.1002/jso.23035. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Purdue MP, Freeman LE, Anderson WF, Tucker MA. Recent trends in incidence of cutaneous melanoma among US Caucasian young adults. J Invest Dermatol. 2008;128:2905–2908. doi: 10.1038/jid.2008.159. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Gandini S, Iodice S, Koomen E, et al. Hormonal and reproductive factors in relation to melanoma in women: current review and meta-analysis. Eur J Cancer. 2011;47:2607–2617. doi: 10.1016/j.ejca.2011.04.023. [DOI] [PubMed] [Google Scholar]
- 14.Lens M, Rosdahl I, Newton-Bishop J. Cutaneous melanoma during pregnancy: is the controversy over? J Clin Oncol. 2009;27:e11–12. doi: 10.1200/JCO.2009.22.2588. author reply e13–14. [DOI] [PubMed] [Google Scholar]
- 15.Byrom L, Olsen CM, Knight L, et al. Does pregnancy after a diagnosis of melanoma affect prognosis? Systematic review and meta-analysis. Dermatol Surg. 2015;41:875–882. doi: 10.1097/DSS.0000000000000406. [DOI] [PubMed] [Google Scholar]
- 16.Jhaveri MB, Driscoll MS, Grant-Kels JM. Melanoma in pregnancy. Clin Obstet Gynecol. 2011;54:537–545. doi: 10.1097/GRF.0b013e318236e18b. [DOI] [PubMed] [Google Scholar]
- 17.Byrom L, Olsen C, Knight L, et al. Increased mortality for pregnancy-associated melanoma: systematic review and meta-analysis. J Eur Acad Dermatol Venereol. 2015;29:1457–1466. doi: 10.1111/jdv.12972. [DOI] [PubMed] [Google Scholar]
- 18.Ribero S, Longo C, Dika E, et al. Pregnancy and melanoma: a European-wide survey to assess current management and a critical literature overview. J Eur Acad Dermatol Venereol. 2016 May 27; doi: 10.1111/jdv.13722. [DOI] [PubMed] [Google Scholar]
- 19.Driscoll MS, Martires K, Bieber AK, et al. Pregnancy and melanoma. J Am Acad Dermatol. 2016;75:669–678. doi: 10.1016/j.jaad.2016.01.061. [DOI] [PubMed] [Google Scholar]
- 20.Martires KJ, Pomeranz MK, Stein JA, et al. Pregnancy-associated melanoma (PAMM): Is there truly a worse prognosis? Would not sound alarm bells just yet. J Am Acad Dermatol. 2016;75:e77. doi: 10.1016/j.jaad.2016.03.055. [DOI] [PubMed] [Google Scholar]
- 21.Richtig G, Byrom L, Kupsa R, et al. Pregnancy as driver for melanoma. Br J Dermatol. 2016 Oct 8; doi: 10.1111/bjd.15124. [DOI] [PubMed] [Google Scholar]
- 22.Lens MB, Rosdahl I, Ahlbom A, et al. Effect of pregnancy on survival in women with cutaneous malignant melanoma. J Clin Oncol. 2004;22:4369–4375. doi: 10.1200/JCO.2004.02.096. [DOI] [PubMed] [Google Scholar]
- 23.O'Meara AT, Cress R, Xing G, et al. Malignant melanoma in pregnancy.A population-based evaluation. Cancer. 2005;103:1217–1226. doi: 10.1002/cncr.20925. [DOI] [PubMed] [Google Scholar]
- 24.Johansson AL, Andersson TM, Plym A, et al. Mortality in women with pregnancy-associated malignant melanoma. J Am Acad Dermatol. 2014;71:1093–1101. doi: 10.1016/j.jaad.2014.09.018. [DOI] [PubMed] [Google Scholar]
- 25.Bannister-Tyrrell M, Roberts CL, Hasovits C, et al. Incidence and outcomes of pregnancy-associated melanoma in New South Wales 1994–2008. Austral N Zealand J Obstet Gynecol. 2015;55:116–122. doi: 10.1111/ajo.12279. [DOI] [PubMed] [Google Scholar]
- 26.Balch CM, Gershenwald JE, Soong SJ, et al. Final version of 2009 AJCC melanoma staging and classification. J Clin Oncol. 2009;27:6199–6206. doi: 10.1200/JCO.2009.23.4799. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Greene FL American Joint Committee on Cancer., American Cancer Society. AJCC cancer staging manual. 6. New York: Springer; 2002. [Google Scholar]
- 28.Edge SBBD, Compton CC, Fritz AG, et al., editors. American Joint Committee on Cancer Staging Manual. 7. New York: Springer; 2009. [Google Scholar]
- 29.Wong JH, Sterns EE, Kopald KH, et al. Prognostic significance of pregnancy in stage I melanoma. Arch Surg. 1989;124:1227–1230. doi: 10.1001/archsurg.1989.01410100133023. discussion 1230–1231. [DOI] [PubMed] [Google Scholar]
- 30.Tellez A, Rueda S, Conic RZ, et al. Risk factors and outcomes of cutaneous melanoma in women less than 50 years of age. J Am Acad Dermatol. 2016;74:731–738. doi: 10.1016/j.jaad.2015.11.014. [DOI] [PubMed] [Google Scholar]
- 31.Travers RL, Sober AJ, Berwick M, et al. Increased thickness of pregnancy-associated melanoma. Br J Dermatol. 1995;132:876–883. doi: 10.1111/j.1365-2133.1995.tb16942.x. [DOI] [PubMed] [Google Scholar]
- 32.MacKie RM, Bufalino R, Morabito A, et al. Lack of effect of pregnancy on outcome of melanoma. For The World Health Organisation Melanoma Programme. Lancet. 1991;337:653–655. doi: 10.1016/0140-6736(91)92462-b. [DOI] [PubMed] [Google Scholar]