Marital Status and Stage at Diagnosis of Invasive Melanoma of the Skin

1
Marital Status and Stage at Diagnosis of Invasive
Melanoma of the Skin

• James (Jay) L. Fisher, Ph.D.
• Comprehensive Cancer Center and
• James Cancer Hospital at
• The Ohio State University

2
Background

• Marriage provides a prognostic advantage to those
  with cancer, including invasive melanoma of the
  skin.
• May result from immunologic benefit of
  psycho-social support associated with marriage or
  from encouragement to follow through with
  diagnosis.
• Spouses may identify suspicious or changing nevi
  (moles) or nevi on areas of the body not easily
  or routinely viewed by one's self.
• Supported by anecdotal stories and by 1 recent
  study of marital status and stage at diagnosis of
  melanoma among older (ages 65) individuals.

3
Goal

• To determine whether single, separated/divorced
  and widowed individuals are more likely to
  present with late stage melanoma, as compared to
  married individuals, and to determine whether
  associations are dependent on
• Sex
• Age at diagnosis
• Year of Diagnosis
• Histology
• Anatomic location.

4
Non-treatment Factors Related to Favorable
Melanoma Prognosis

• Female sex
• Younger age at diagnosis
• White race
• Localized stage at diagnosis
• Histology of superficial spreading and lentigo,
  versus nodular and acral lentiginous
• Anatomic location of limb, compared to trunk
• Characteristics of lesion (e.g. lower levels of
  thickness)

5
5-Year Melanoma Survival Probability by Stage at
Diagnosis, 1996-2004
Source SEER Cancer Statistics Review,
1975-2005, National Cancer Institute, 2008 Based
on SEER 17 areas.
6
5-Year Melanoma Survival Probability by Marital
Status, 1996-2004
Source SEERStat 6.4.3, 2008 Based on SEER 17
areas.
7
Methods Data Source

• Surveillance, Epidemiology, and End Results
  (SEER) Program Database, 17 Regions, National
  Cancer Institute
• Data accessed using SEERStat 6.4.3 (Released
  April 2008)
• Years of diagnosis 1973-2005
• SEER Historic Stage A used for stage at diagnosis
  
• in situ cases excluded from outset
• Early stage characterized as localized stage
• Late stage characterized as regional and distant
  stages combined
• Marital Status at Diagnosis was categorized as
• Married
• Single/Never Married
• Separated/Divorced
• Widowed

8
Methods Exclusions

• Age at Diagnosis
• Younger than 25 years of age (n 5,005)
• SEER Historic Stage A
• Unknown (n 8,728)
• Marital Status
• Unknown (n 32,256)
• 121,506 melanoma cases remained after exclusions

9
Methods Design and Statistical Analyses

• Case-control Study
• Cases defined as persons with late stage
  melanoma.
• Controls defined as persons with early stage
  melanoma.
• Logistic Regression
• Used to determine odds of specified marital
  status according to case status (case, control).
• Odds ratios (ORs) estimate relative risks.
• An OR gt 1.0 means individuals with a specified
  marital status are more likely to be cases.
• Hypotheses determined a priori.
• Alpha set at 0.05 for 2-tailed hypothesis tests.
• Potential Confounder
• Age at diagnosis (controlled by inclusion in a
  multivariate model).

10
Methods Design and Statistical Analyses

• Potential Effect Modifiers
• Sex
• Age at diagnosis
• Year of diagnosis
• Histology
• Anatomic location
• Effect modification assessed by stratification of
  regressions.
• Statistical analyses were conducted using SAS
  9.1.

11
Marital Status at Melanoma Diagnosis by Sex and
Stage at Diagnosis
12
Crude and Age-adjusted ORs and 95CIs Estimating
Risk of Late Stage (versus Early Stage) Melanoma
According to Marital Status by Sex
13
Age-Adjusted ORs and 95CIs Estimating Risk of
Late Stage Melanoma According to Marital Status
by Age of Diagnosis Among Males
14
Age-Adjusted ORs and 95CIs Estimating Risk of
Late Stage Melanoma According to Marital Status
by Age at Diagnosis Among Females
15
Major Histologic Types of Melanoma

• Superficial Spreading Melanoma
• Most common type, 70 Often found on legs and
  trunk Good prognosis.
• Nodular melanoma
• Second most common type, 15-20 Commonly occurs
  on trunk Generally poor prognosis.
• Lentigo Melanoma
• Usually least common, 4-10 Generally good
  prognosis.
• Acral Lentiginous Melanoma
• Most common type among darkly-pigmented people,
  60 only 2-8 of all melanomas Prognosis
  between superficial spreading and nodular.

16
Age-Adjusted ORs and 95CIs Estimating Risk of
Late Stage Melanoma According to Marital Status
by Histology Among Males
1. Superficial Spreading 2. Acral Lentiginous
17
Age-Adjusted ORs and 95CIs Estimating Risk of
Late Stage Melanoma According to Marital Status
by Histology Among Females
1. Superficial Spreading 2. Acral Lentiginous
18
Anatomic Location

• Individuals with melanomas located on areas of
  the body not easily or routinely viewed by ones
  self may benefit more as the result of early
  spousal recognition.
• Ideal comparison would be easily visible areas of
  the body versus areas not easily visible (e.g.
  posterior versus anterior areas). However, this
  information is not available.
• These groupings of anatomic locations were used
  
• face/head/scalp/neck
• trunk
• upper limb/shoulder
• lower limb/hip

19
Age-Adjusted ORs and 95CIs Estimating Risk of
Late Stage Melanoma According to Marital Status
by Anatomic Location Among Males
20
Age-Adjusted ORs and 95CIs Estimating Risk of
Late Stage Melanoma According to Marital Status
by Anatomic Location Among Females
21
Summary of Findings

• Marriage affords the advantage of diagnosis at an
  earlier stage.
• Associations between late stage melanoma and
  single marital status were stronger among males
  associations with separated/divorced and widowed
  were stronger among females.
• No clear increasing/decreasing trend in late
  stage melanoma with age for any non-married
  marital status.
• For males, being single was associated with late
  stage superficial spreading and nodular melanoma
  for females, being both single and widowed were
  associated with late stage nodular melanoma.
• No anatomic location for which associations with
  late stage melanoma were strongest/weakest for
  each non-married marital status.

22
Alternative Explanations Pertaining to Melanoma
Prognosis According to Marital Status

• Alternative explanations pertaining to melanoma
  prognosis according to marital status
• Higher income and health insurance among the
  married
• Psychological stress among the non-married
• Reduced immunologic response to melanoma among
  the widowed
• Lower quality care among older widows.

23
Limitations

• Inability to examine late stage melanoma among
  monogamously coupled, yet unmarried, individuals
• No knowledge of length of marital status at time
  of diagnosis
• High proportion of unknown marital status (20)
• Possible uncontrolled confounding (or
  explanation) by factors associated with marriage
  (e.g. socioeconomic factors, physical appearance,
  personality characteristics)

24
The Point

• Because melanoma incidence is rapidly increasing
  and because localized melanoma has a very
  favorable prognosis, it is important to identify
  groups at greater risk of late stage melanoma.
• Marriage affords a melanoma survival benefit due,
  at least in part, to earlier stage at diagnosis.
• Unmarried individuals should be targeted in skin
  screening programs and educational efforts to
  improve early detection of melanoma.

25
Co-authors

• Holly L. Engelhardt, M.S.
• Cancer Epidemiologist, Ohio Cancer Incidence
  Surveillance System, Ohio Department of Health
• Julie A. Stephens, M.S.
• Senior Consulting Research Statistician, Center
  for Biostatistics, The Ohio State University
• Robert W. Indian, M.S.
• Chief, Chronic Disease and Behavioral
  Epidemiology Section, Ohio Department of Health
• Electra D. Paskett, Ph.D.
• Marion N. Rowley Professor of Cancer Research,
  Division of Epidemiology, College of Public
  Health Associate Director of Population Sciences
  and Program Co-Leader, Cancer Control Program,
  Comprehensive Cancer Center, The Ohio State
  University

26
Questions? Contact Information Jay.Fisher_at_osumc.
edu [phone]