Time trends and latitudinal differences in melanoma thickness distribution in Australia, 1990–2006

Time trends and latitudinal differences in melanomathickness distribution in Australia, 1990–2006

Peter Baade1,2, Xingqiong Meng1, Danny Youlden1, Joanne Aitken1,3,4 and Philippa Youl1,3

1 Viertel Centre for Research in Cancer Control, Cancer Council Queensland, Brisbane, Australia2 School of Public Health, Queensland University of Technology, Brisbane, Australia3 Griffith Health Institute, Griffith University, Gold Coast, Australia4 School of Population Health, University of Queensland, Brisbane, Australia

This study investigated time trends and latitude differentials in the thickness distributions of invasive melanomas diagnosed

in Australia between 1990 and 2006 using data from population-based cancer registries. Trends in incidence rates were

calculated by sex, age group, thickness, year at diagnosis and latitude. For thin (<1.00mm) melanomas the increase was very

pronounced during the early 1990s (1990–1996, annual percentage change and 95% confidence interval: males

15.6(13.5,17.7); females 14.1(11.7,16.5), but then incidence rates became stable among both males (10.6(20.1,11.4))

and females (20.0(20.9,10.9)) of all ages between 1996 and 2006. In contrast, incidence of thick (>4.00 mm) melanomas

continued to increase over the entire period (males 12.6(11.9,13.4); females 11.6(10.6,12.6)). Recent reductions in the

incidence of thin melanomas were observed among young (<50 years) males and females, contrasted by an increase in thin

melanomas among older people, and increases in thick melanomas among most age groups for males and elderly (751)

females. A strong latitude gradient in incidence rates was observed, with rates being highest in northern, more tropical areas

and lowest in the most southern regions. However, the magnitude of the increase in thick melanomas was most pronounced

in southern parts of Australia. The observed trends in thin melanomas can most likely be attributed to the impact of early

detection and skin awareness campaigns. However, these efforts have not impacted on the continued increase in the

incidence of thick melanomas, although some increase may be due to earlier detection of metastasising melanomas. This

highlights the need for continued vigilance in early detection processes.

With its high rates of ultraviolet radiation, outdoor lifestyleand predominately Caucasian population, Australia continuesto have the highest incidence rates of cutaneous melanomain the world, with an estimated age-standardised incidencerate (World 2000 population) of 40.2 cases per 100,000 popu-lation in 2008, similar to New Zealand (40.1).1,2 Rates inother countries are substantially lower, with Switzerland(20.8), Denmark (19.9) and Norway (19.1) on the second tierof melanoma risk, and the USA (15.6), UK (11.6) and Ger-many (13.2) having lower incidence rates.1,2 While melanomahas been the most rapidly increasing cancer among fair-skinned populations worldwide,3 studies from Europe, Can-ada, the United States and Australia have reported slowing orstabilizing rates of increase in incidence rates starting from

the mid-1980s onwards.3–7 Much of the observed increase inmelanoma incidence has been in thinner melanomas, withthe incidence rates for thicker melanomas either increasing ata reduced rate or stabilizing. 4,6,8,9 With the strong associa-tion between tumour thickness and survival,8,10–13 anyincrease in the incidence of thick melanomas has importantimplications for the mortality burden caused by melanoma.

Although other Australian studies have reported patternsof melanoma incidence over time by thickness 4,6,8,9 thesehave all been state-specific and have used varying analyticaland reporting methods. To our knowledge, national trendson melanoma incidence by thickness have not been previ-ously reported. In addition, the pooling of state-specific inci-dence data enables us to directly examine the associationbetween incidence trends and geographical latitude, as hasbeen recently reported for Norway.14

Material and MethodsData

All cases of invasive cutaneous melanoma (ICD-0-3 codeC44, morphology codes M872-M879) diagnosed in Australiabetween 1990 and 2006 (inclusive) were obtained from theAustralia Institute of Health and Welfare (AIHW) AustralianCancer Database (ACD). The ACD receives data from indi-vidual state and territory cancer registries on all cancers

Key words: melanoma incidence, time trends, thickness, latitude

Grant sponsor: National Health and Medical Research Council;

Grant number: CDA 1005334

DOI: 10.1002/ijc.25996

History: Received 18 Oct 2010; Accepted 2 Feb 2011; Online 22 Feb

2011

Correspondence to: Peter Baade, Viertel Centre for Research in

Cancer Control, The Cancer Council Queensland, PO Box 201,

Spring Hill, QLD 4004, Australia; Facsimile, Tel: [þ61-7-32598527],

E-mail: [email protected]

Epidemiology

Int. J. Cancer: 130, 170–178 (2012) VC 2011 UICC

International Journal of Cancer

IJC

diagnosed for residents of Australia (except non-melanomaskin cancer). Notifications to these population-based regis-tries are required by law. Information on sex, 5-year agegroup (up to 85þ), calendar year of diagnosis and melanomathickness (�1.00 mm, 1.01–2.00 mm, 2.01–4.00 mm, >4.00mm or Unknown) was provided. Throughout the article, thinmelanomas refer to those < ¼ 1 mm and thick melanomasare those >4 mm. The category � 1.00 mm does not includein situ lesions. Cases with unknown age (n ¼ 5) wereexcluded from all analyses.

To assess the impact of latitude on trends in melanomathickness, we also obtained aggregated data from the AIHWon melanoma incidence for the eastern states and territories;Queensland, New South Wales, Australian Capital Territory,Victoria and Tasmania. Because of a smaller number of cases,data for Tasmania and Victoria were combined while thedata for New South Wales included incident cases from theAustralian Capital Territory. This provided 3 north-southgeographical regions: ‘‘Northern’’ (latitudes ranging from�10� to 29� South), ‘‘Central’’ (from 28� to 37� South) and‘‘Southern’’ (from 36� to 43� South). For these region-baseddata, age was collapsed by the AIHW prior to release intobroad groups (<50 years, 50–64 years, 65–74 years and 75years and over), as was period of diagnosis (1990–1996,1997–2001, 2002–2006).

Statistical methods

Directly age-standardised (World 2000) incidence rates werecalculated according to region, sex, age group, thickness andyear categories. We used Joinpoint software (version 3.4.3,National Cancer Institute, 2010)15 to assess the trends(measured by the annual percentage change, or APC), andspecifically to determine whether there were any statisticallysignificant changes in the magnitude or direction of the

trends over the study period. The APC was estimated by fit-ting a regression line to the logarithm of the age-standar-dised rates with a linear term for year of diagnosis. Toreduce the likelihood of reporting spurious changes intrends, we used a maximum of 3 joinpoints (i.e., up to 4 dif-ferent trends) with a minimum of 5 years of data each.16

Monte Carlo permutation tests were used to examine thetrend lines for each combination of join points, and thetrend line that provided the best fit to the observed data wasselected.17

Poisson regression models were used to investigate differ-ences in melanoma incidence rates by region, sex, age group,thickness and year categories. The log of the estimated resi-dent population,18 was used as the offset parameter in thePoisson model. Tests for interactions between the latituderegions and other statistically significant main effects wereused to assess differences in effects by region. Separate mod-els were used for thin and thick melanomas.

Ethics considerations

Approvals from the respective state and territory cancerregistries were obtained by the AIHW before the data wasreleased. No potentially identifying information was releasedoutside the AIHW and therefore the AIHW Ethics Commit-tee waived the requirement for Ethics Committee approval ofthis study.

ResultsBetween 1990 and 2006, 139,943 Australian residents werediagnosed with an invasive melanoma, representing an av-erage age-standardised (2000 World Population) incidencerate of 36.2 cases per 100,000 population (Table 1). Overhalf (56.2%) of the cases were males, and 67.2% of caseswere diagnosed among people aged 50 years and over.

Table 1. Proportion and age-standardised rate (ASR) for incidence of invasive cutaneous melanoma classified by tumour thickness,Australia, 1990–2006

Sex Thickness (mm) Cases Percentage ASR(1990–2006)

Male �1.00 47,564 60.5 25.3

1.01–2.00 11,165 14.2 5.9

2.01–4.00 7,485 9.5 3.8

>4.00 4,446 5.7 2.2

Unknown 7,932 10.1 4.2

Total 78,590 100.0 41.3

Female �1.00 40,153 65.4 21.2

1.01–2.00 8,701 14.2 4.3

2.01–4.00 4,823 7.9 2.1

>4.00 2,635 4.3 1.0

Unknown 5,041 8.2 2.5

Total 61,353 100.0 31.1

Total 139,943 36.2

ASR: Age-standardised rate (per 100,000 population) to the 2000 World standard population.

Epidemiology

Baade et al. 171


Incidence rates were very low among children and youngeradults (<12.2/100,000 up to the age of 24 years), then rap-idly increased up to 173.5/100,000 for those aged 85 yearsand over. Over the whole study period, nearly two-thirds(62.7%) of invasive melanomas were thin, while 5.1% werethick. The proportion of melanomas that were thinincreased from 58.0% in 1990 to a peak of 65.9% in 1999,and has ranged between 60.9% and 64.9% since then. Incontrast, the proportion of melanomas that were thickgradually increased from 4.5% in 1990 to a peak of 5.9% in2004 and 2006.

Trends by sex and thickness

Over the 17 years (1990–2006), incidence rates for melanomagenerally increased within all known thickness categories(Table 2; Fig. 1). During the early to mid-1990s, incidencerates for thin melanomas increased at approximately twicethe rate of increase of thick melanomas for both males andfemales (Table 2). However, in contrast with the generallyongoing increasing trends for intermediate (1.01–2.00 mmand 2.01–4.00 mm) and thick melanomas, the Joinpoint anal-ysis suggested that the increasing trends for thin melanomasby sex plateaued (non-significant trend) during the period1996–2006 (Table 2).

Trends by age group and thin/thick melanomas

Until the mid-1990s significantly increasing trends in thinmelanomas were observed across all 4 age groups for malesand females (Table 3). The gradients of the trend generallyincreased with age among males and were much steeper thanthe corresponding trends among females aged 50 years andolder. However from 1997 onwards, there was a significantdecrease in the incidence of thin melanomas among younger(<50 years) males and females. The magnitude of the rate ofincrease also reduced for males in each of the other agegroups from around the mid-1990s.

While there was some evidence of consistently increasingtrends for thick melanomas across all age groups, for some ofthe age-sex cohorts (males aged 50–64 years and femalesaged <75 years) these trends were not statistically significant(Table 3).

Incidence by latitude

There was a strong inverse relationship in melanoma inci-dence by latitude, with incidence rates between 1990 and2006 being highest in the Northern region which is closest tothe equator (average of 52.7/100,000), then lower in the Cen-tral and Southern regions (Central: 37.5/100,000; Southern:30.9/100,000). This pattern was also true for each time period

Table 2. Annual percentage change (APC) in age standardised incidence rates for invasive melanoma by trend period, Australia, 1990–2006

ThicknessTrend 1 Trend 2 Trend 3 Trend 4

Sex (mm) Year APC (95% CI) Year APC (95% CI) Year APC (95% CI) Year APC (95% CI)

Male �1.00 1990–1996 þ5.6(þ3.5, þ7.7)*

1996–2006 þ0.6(�0.1, þ1.4)

1.01–2.00 1990–2006 þ1.4(þ0.9, þ1.9)*

2.01–4.00 1990–2006 þ1.5(þ0.8, þ2.2)*

>4.00 1990–2006 þ2.6(þ1.9, þ3.4)*

Unknown 1990–1994 þ1.2(�6.7, þ9.8)

1994–1998 �8.0(�19.0, þ4.5)

1998–2002 þ7.8(�5.6, þ23.2)

2002–2006 �5.3(�12.0, þ2.0)

Total 1990–1996 þ3.4(þ1.6, þ5.3)*

1996–2006 þ0.9(þ0.2, þ1.6)*

Female �1.00 1990–1996 þ4.1(þ1.7, þ6.5)*

1996–2006 �0.0(�0.9, þ0.9)

1.01–2.00 1990–2006 þ0.7(�0.0, þ1.3)

2.01–4.00 1990–2006 þ1.0(þ0.2, þ1.7)*

>4.00 1990–2006 þ1.6(þ0.6, þ2.6)*

Unknown 1990–1998 �5.4(�7.0, �3.8)*

1998–2002 þ7.5(�1.3, þ17.1)

2002–2006 �6.0(�10.4, �1.5)*

Total 1990–2006 þ0.9(þ0.4, þ1.4)*

*The Annual Percent Change (APC) is statistically significant from zero. APC derived from JoinPoint regression. The age-standard rates used in theJoinpoint regression were standardised to the 2000 World Standard Population.

Epidemiology

172 Time trends in melanoma thickness


(Table 4). Compared with melanoma incidence rates in theNorthern region, rates were 28% lower in the Central regionand 43% lower in the Southern region after adjustment for

sex, age group, diagnosis year and thickness (Table 5). Thecorresponding rates for thin melanomas by region followed asimilar, although even more pronounced, pattern. In contrast,

Table 3. Annual percentage change (APC) in age standardised incidence rates for thin- and thick-invasive melanomas by sex, age atdiagnosis and trend period, Australia, 1990–2006

Trend 1 Trend 2

Thickness (mm) Sex Age (years) Year APC (95% CI) Year APC (95% CI)

Thin Male <50 1990–1997 þ4.1 (þ2.1, þ6.3)* 1997–2006 �2.1 (�3.4, �0.8)*

(�1.00) 50–64 1990–1994 þ8.6 (þ4.4, þ13.0)* 1994–2006 þ1.7 (þ1.1, þ2.3)*

65–74 1990–1997 þ6.4 (þ3.8,þ 9.0)* 1997–2006 þ1.7 (þ0.4, þ3.1)*

75þ 1990–1994 þ10.4 (þ5.4, þ15.6)* 1994–2006 þ3.7 (þ3.1, þ4.3)*

Female <50 1990–1997 þ4.0 (þ2.1, þ6.0)* 1997–2006 �1.8 (�2.9, �0.6)*

50–64 1990–2006 þ1.9 (þ1.2, þ2.6)*

65–74 1990–2006 þ2.8 (þ2.0, þ3.7)*

75þ 1990–2006 þ4.0 (þ3.5,þ 4.6)*

Thick Male <50 1990–2006 þ2.0 (þ0.6, þ3.3)*

(>4.00) 50–64 1990–2006 þ1.1 (�0.5, þ2.7)

65–74 1990–2006 þ2.2 (þ0.5, þ3.9)*

75þ 1990–2006 þ4.6 (þ3.2, þ5.9)*

Female <50 1990–2006 þ1.7 (�1.4, þ5.0)

50–64 1990–2006 þ1.2 (�1.1, þ3.5)

65–74 1990–2006 þ1.5 (�0.7, þ3.7)

75þ 1990–2006 þ2.3 (þ1.1, þ3.5)*

*The Annual Percent Change (APC) is statistically significant from zero. APC derived from JoinPoint regression. The age-standard rates used in theJoinpoint regression were standardised to 2000 World Standard Population.

Figure 1. Time trends of age-standardised rates (ASR, standardised to 2000 World Standard Population) for incidence of invasive

melanoma by thickness and sex, Australia, 1990-2006. Fitted lines were derived from Joinpoint regression.

Epidemiology

Baade et al. 173


the incidence rate of thick melanomas in the Central regionwas equivalent to the Northern region, while the rate for theSouthern region was 41% lower.

As evidenced by the significant (p < 0.01) tests for inter-action between region and time, the trends over time weresignificantly different across the 3 regions for all melanomasand both thin and thick melanomas. For example, comparedwith the region-specific incidence of thick melanoma between1990 and 1996, there was a 19% increase in the Northernregion by 2002 and 2006, while the Central region increasedby 31%, and the Southern region increased by 63% over thesame period (Table 5).

DiscussionThis large study, utilising data from almost 140,000 Australiansdiagnosed with melanoma, demonstrated that the previouslysharp rate of increase in thin melanomas has plateaued. How-ever, rates of thick melanoma have continued to increase amongboth males and females, particularly in areas of higher latitude.

In contrast to these national results, a recent Queenslandstudy4 examining trends between 1991 and 2002 found noevidence to suggest that the increasing trends for thin mela-nomas were levelling off. Similarly, ongoing increasing trendshave been reported previously in the United Kingdom(1993–2003),8,19 the United States (1988–1994),20 New SouthWales (1989–1996)6 and (1993–2003),8 Puerto Rico (1987–2002),21 Southern Germany (1976–2003),22 and Northern Ire-land (1984–2006).23 However the short time periods and ana-

lytical methods used in some of these studies may not haveallowed changes in trends to be detected.

It has been suggested that widespread increases in thinmelanomas during the 1990s were predominately due toheightened levels of melanoma awareness and improved earlydetection, rather than a real increase in the underlying mela-noma incidence.4,24 The recent levelling off of this increaseobserved in the Australian context is consistent with this hy-pothesis—improved detection leads to an initial large increasein the incidence of early disease after which incidence ratesplateau, albeit at a higher level than previously, as the pool ofundetected lesions diminishes.25,26 It remains to be seenwhether this stabilization in the incidence of thin melanomascontinues, or even whether incidence trends will eventuallystart to decrease across the whole population.

To the extent that the apparent increased detection ofthin melanomas observed here is explained by earlier diagno-sis (as opposed to over-diagnosis of non-progressive lesions),it would be expected to be followed in time by a reducedincidence of thick melanomas. These Australian results sug-gest the opposite has occurred. That is, the improvements inearly diagnostic methods have not been sufficient to counteran increase in the incidence in the underlying incidence ofthicker tumours. One explanation is that some tumours growrapidly and are thus less likely to be detected before theybecome thick. Therefore advances in early detection activitiessuch as skin screening may be less likely to have an impacton the incidence rates of these types of tumours.27

Table 4. Differences in age-standardised incidence rates of melanoma classified by tumour thickness and latitude, Australia, 1990–2006

Region Thickness (mm) Cases % ASR (1990–96) ASR (1997–01) ASR (2002–06) ASR (1990–06)

Northern �1.00 23,142 67 31.9 39.7 37.4 36.3

1.01–2.00 4,470 13 6.3 6.6 7.2 6.7

2.01–4.00 2,629 8 3.4 3.6 4.1 3.7

>4.00 1,403 4 1.8 1.9 2.1 1.9

Unknown 2,728 8 5.0 3.9 3.2 4.0

Total 34,372 100 48.4 55.6 54.1 52.7

Central �1.00 28,950 59 20.4 23.4 24.8 22.9

1.01–2.00 7,446 15 5.4 5.5 6.1 5.7

2.01–4.00 4,888 10 3.3 3.4 3.7 3.5

>4.00 2,874 6 1.8 1.9 2.2 2.0

Unknown 4,924 10 3.8 3.0 4.0 3.6

Total 49,082 100 34.6 37.1 40.9 37.5

Southern �1.00 18,681 60 17.2 20.1 20.3 19.2

1.01–2.00 4,499 15 4.0 4.3 4.9 4.4

2.01–4.00 2,776 9 2.3 2.4 2.8 2.5

>4.00 1,599 5 1.1 1.4 1.7 1.4

Unknown 3,508 11 3.2 3.3 3.7 3.4

Total 31,063 100 27.8 31.4 33.5 30.9

Total 114,517 36.9 41.4 42.8 40.4

ASR: Age standardised rate (2000 World Standard Population).

Epidemiology



Increased tumour thickness has been consistently shown tobe the strongest predictor of poorer survival prognosis, bothinternationally8,10–13 and in Australia.28–30 When combinedwith the limited effectiveness of treatment for thick melano-mas,31 it would be expected that any trends in thick melano-mas, after allowing for some lead time, would be reflected in

melanoma mortality outcomes. Therefore, the consistentlyincreasing incidence of thick melanomas observed in thisstudy should be accompanied by increasing mortality rates ofmelanoma. This has not been observed. A previous studyreporting on Australian melanoma mortality rates since 1950,found that, following consistent increases, melanoma

Table 5. Incidence rate ratios (with 95% confidence intervals) of melanoma by region, sex, age group, year and thickness

All1 Thin2 (�1.00 mm) Thick2 (>4.00 mm)

Region

Northern (reference) 1.00 1.00 1.00

Central *0.72 (0.70, 0.74) *0.64 (0.62, 0.66) 0.97 (0.87, 1.09)

Southern *0.57 (0.56, 0.59) *0.54 (0.52, 0.56) *0.59 (0.52, 0.68)

p < 0.001 p < 0.001 p < 0.001

Sex

Male (reference) 1.00 1.00 1.00

Female *0.71 (0.70, 0.72) *0.79 (0.78, 0.80) *0.49 (0.46, 0.51)

p < 0.001 p < 0.001 p < 0.001

Age group (years)

<50 (reference) 1.00 1.00 1.00

50–69 *4.54 (4.48, 4.60) *3.97 (3.90, 4.04) *8.85 (8.12, 9.64)

70þ *8.06 (7.94, 8.18) *5.39 (5.28, 5.49) *40.46 (37.37, 43.80)

p < 0.001 p < 0.001 p < 0.001

Year of Diagnosis

Northern

1990–1996 (reference) 1.00 1.00 1.00

1997–2001 *1.15 (1.12, 1.18) *1.25 (1.21, 1.29) 1.03 (0.90, 1.18)

2002–2006 *1.13 (1.11, 1.16) *1.20 (1.16, 1.24) †1.19 (1.05, 1.35)

Central

1990–1996 (reference) 1.00 1.00 1.00

1997–2001 *1.09 (1.06, 1.11) *1.17 (1.13, 1.20) 1.07 (0.97, 1.17)

2002–2006 *1.22 (1.19, 1.24) *1.26 (1.23, 1.30) *1.31 (1.20, 1.43)

Southern

1990–1996 (reference) 1.00 1.00 1.00

1997–2001 *1.14 (1.11, 1.17) *1.17 (1.13, 1.22) *1.33 (1.17, 1.51)

2002–2006 *1.23 (1.20, 1.27) *1.21 (1.17, 1.25) *1.63 (1.45, 1.84)

Interaction: p < 0.001 Interaction: p < 0.001 Interaction: p ¼ 0.004

Thickness (mm)

<¼1.00 (reference) 1.00

1.01–2.00 *0.23 (0.23, 0.24)

2.01–4.00 *0.15 (0.14, 0.15)

>4.00 *0.08 (0.08, 0.09)

Unknown *0.16 (0.15, 0.16)

p < 0.001

1Poisson regression for all melanoma (including all thickness categories) adjusted for sex, age group, year of diagnosis, thickness, region, andinteraction of year of diagnosis and region. 2Poisson regression for thin (�1.00 mm) or thick (>4.00 mm) melanoma adjusted for sex, age group,year of diagnosis, region and interaction of year of diagnosis and region.*p < 0.001.†p < 0.01.

Epidemiology

Baade et al. 175


mortality rates stabilized among males between 1989 and2002 and decreased among females.31 Additional analyses(results not shown) suggest these patterns have continued upto 2006.

One possible explanation is that trends in the proportion(9%) of melanomas with unknown thickness, nearly twice ascommon as thick melanomas (5%), could compromise thereported trends in thick melanomas, particularly since thepercentage of melanomas with unknown thickness hasdecreased over time, from 12% in 1990 to 8% in 2006.Reductions in the percentage of melanomas with unknownthickness have also been reported in the United States.24

Although we were unable to obtain national data, unpub-lished data from the Queensland Cancer Registry showedthat over well over half (60%) of the melanomas withunknown thickness were diagnosed on the basis of histologyof metastasis. If previously metastasising melanomas are nowbeing detected earlier, most likely as thick melanomas, thenthis would increase the number of thick melanomas beingdetected, and hence directly impact on observed incidencetrends by thickness. However the relationship between thickmelanomas and melanomas with unknown thickness is notconsistent by latitude; the rates of thick melanomas haveincreased in each of the 3 regions whereas the rates of mela-nomas with unknown thickness has either decreased (North-ern), remained stable (Central) or increased (Southern). Alsothe trends in melanomas of unknown thickness are highlyvariable, suggesting against a consistent improvement indetection practices. Therefore the possible implications ofchanges in the rates of melanomas with unknown thicknesson the observed incidence trends by thickness can only bespeculative. Although some of the observed increase in thickmelanomas may be due to improved detection of previouslymetastasising melanomas, it would be presumptuous to sug-gest that a real increase has not occurred. Clearly furtherinvestigation is required.

Our finding that the incidence of thin melanoma isdecreasing among younger people is promising. The impor-tance of early life sun exposure and the generally long latencyperiod for the development of melanoma is well established.32

Public primary prevention campaigns directed at reducingsun exposure have been ongoing in Australia since the early1980s.4,33 The observed reduction in thin melanomas inyounger age groups, and a similar stabilising of rates of non-melanoma skin cancer,34 could lend cautious support to thesuccess of these campaigns, particularly if these lower ratescontinue into the future.4,34,35

As suggested by others,24 the current analysis highlights aneed for new early detection strategies to be developed, par-ticularly for segments of the population previously shown tobe at higher risk of thick melanomas, including men, olderpeople and those with low education.36 These strategiesshould include identification of the clinical features of mela-nomas that are more likely to have a rapid growth phase,and encouraging regular clinical skin examinations among

high risk individuals. A recent case-control study of mela-noma screening demonstrated that having a whole-body clin-ical skin examination in the 3 years prior to a diagnosis ofmelanoma was associated with a 40% reduction in the inci-dence of melanoma �3.00mm thick.37 Therefore, moreresearch is needed to understand differences in the develop-ment between thin and thick melanoma which would thenallow specific targeting of the various underlying factors infuture public health campaigns.

We found a strong association between melanoma inci-dence and latitude, with overall incidence risks being signifi-cantly lower in the Central and Southern regions comparedwith the Northern region. A similar pattern was observed forthin melanomas, but for thick melanomas there was only asignificant differential between the Southern and Northernregions. We also found strong evidence that the changes inincidence over time varied by latitude, with the increase inincidence of thick melanomas over time particularly pro-nounced in Southern states. Reasons for this differentialincrease are unclear. The authors of a European study38 sug-gested possible reasons for an observed differential in theincidence of thick melanoma by latitude were variations inoverall awareness of melanoma and frequency of campaignsaimed at early detection. However, it is unlikely that factorssuch as lower prevalence of melanoma, impacting on clinicaldiagnostic abilities,39 or lower community awareness of mela-noma, are explanations for the observed Australian trends.Per capita expenditure on sun protection programs is report-edly much higher in Victoria (Southern region) than Queens-land (Northern region),40 however the increased incidence ofnon-melanoma skin cancer in Northern Australia34 may indi-rectly increase awareness through greater utilisation of doc-tors for treatment.

The strengths of this study include the use of data frompopulation-based cancer registries providing a complete enu-meration of all Australians diagnosed with invasive mela-noma between 1990 and 2006, including information aboutmelanoma thickness at diagnosis. All information used inthis study has been collected prospectively for administrativepurposes independently of the study hypotheses, thus remov-ing recall or information bias. In contrast to other researchstudies that have considered constant linear changes overtheir entire study period, the use of Joinpoint regressionenabled us to detect changes in the magnitude and directionof trends over time. Since the selection of joinpoints can beinfluenced by random fluctuations in rates, particularly thoseat the end points, we chose conservative parameters to reducethe chance of detecting spurious changes in trends. We wereunable to obtain region-specific data by calendar year due toconfidentiality restrictions, so the aggregating of incidencedata across year groups may have influenced our ability torecognise trends over time by latitude.

There are encouraging trends in the incidence of thinmelanomas; however, the continued increase in the incidenceof thick melanomas, more pronounced in regions of high

Epidemiology



latitude, is cause for concern. Although improvements inearly detection could mean that some melanomas that maypreviously have metastasised are now being diagnosed asthick melanomas, there is still a strong need to develop newearly detection strategies, ideally with an emphasis on thosegroups known to have greater risk of being diagnosed withthick melanomas, to improve the efficacy of detecting mela-noma early and address this increasing trend of thickmelanomas.

AcknowledgementsThe authors acknowledge the assistance of staff from the Cancer and Screen-ing Unit, Australian Institute of Health andWelfare who conducted the dataextraction required for this study.Dr. Peter Baade is supported by the National Health and Medical

Research Council (CDF 1005334).The authors appreciate the assistance provided by Jacques Ferlay, Interna-

tional Agency for Research on Cancer (IARC) who re-calculated the interna-tional age standardized melanoma incidence rates from GLOBOCAN usingtheWorld 2000 Standard population as described in the Introduction.

References

1. Ferlay J, Shin HR, Bary F, Forman D,Mathers C, Parkin DM. GLOBOCAN2008, Cancer incidence and mortalityworldwide: IARC CancerBase No. 10[Internet] Lyon, France: InternationalAgency for Research on Cancer, 2010.

2. Ferlay J. Age-standarised incidence ratesfrom the GLOBOCAN 2008 database(http://www-dep.iarc.fr/) using World 2000population as standard provided onrequest, 2010.

3. Garbe C, Leiter U. Melanomaepidemiology and trends. Clin Dermatol2009;27:3–9.

4. Coory M, Baade P, Aitken J, Smithers M,McLeod GR, Ring I. Trends for in situ andinvasive melanoma in Queensland,Australia, 1982–2002. Cancer CausesControl 2006;17:21–7.

5. de Vries E, Bray FI, Coebergh JW, ParkinDM. Changing epidemiology of malignantcutaneous melanoma in Europe 1953–1997:rising trends in incidence and mortalitybut recent stabilizations in western Europeand decreases in Scandinavia. Int J Cancer2003;107:119–26.

6. Marrett LD, Nguyen HL, Armstrong BK.Trends in the incidence of cutaneousmalignant melanoma in New South Wales,1983–1996. Int J Cancer 2001;92:457–62.

7. Stang A, Pukkala E, Sankila R, SodermanB, Hakulinen T. Time trend analysis of theskin melanoma incidence of Finland from1953 through 2003 including 16,414 cases.Int J Cancer 2006;119:380–4.

8. Downing A, Yu XQ, Newton-Bishop J,Forman D. Trends in prognostic factorsand survival from cutaneous melanoma inYorkshire, UK and New South Wales,Australia between 1993 and 2003. Int JCancer 2008;123:861–6.

9. Garbe C, McLeod GR, Buettner PG. Timetrends of cutaneous melanoma inQueensland, Australia and Central Europe.Cancer 2000;89:1269–78.

10. Metelitsa AI, Dover DC, Smylie M, deGara CJ, Lauzon GJ. A population-basedstudy of cutaneous melanoma in Alberta,Canada (1993–2002). J Am Acad Dermatol2010;62:227–32.

11. Payette MJ, Katz M, III, Grant-Kels JM.Melanoma prognostic factors found in thedermatopathology report. Clin Dermatol2009;27:53–74.

12. Balch CM, Gershenwald JE, Soong SJ,Thompson JF, Atkins MB, Byrd DR, BuzaidAC, Cochran AJ, Coit DG, Ding S,Eggermont AM, Flaherty KT, et al. Finalversion of 2009 AJCC melanoma staging andclassification. J Clin Oncol 2009;27:6199–206.

13. Haddad FF, Stall A, Messina J, Brobeil A,Ramnath E, Glass LF, Cruse CW, BermanCG, Reintgen DS. The progression ofmelanoma nodal metastasis is dependenton tumor thickness of the primary lesion.Ann Surg Oncol 1999;6:144–9.

14. Cicarma E, Juzeniene A, Porojnicu AC,Bruland OS, Moan J. Latitude gradient formelanoma incidence by anatomic site andgender in Norway 1966–2007. J PhotochemPhotobiol B 2010;101:174–8.

15. Statistical Research and ApplicationsBranch. Joinpoint Regression Program,Version 3.4.3 - April 2010: NationalCancer Institute, 2010.

16. Yu B, Barrett MJ, Kim H-J, Feuer EJ.Estimating joinpoints in continuous timescale for multiple change-point models.Computational Statistics Data Anal 2007;51:2420–7.

17. Kim HJ, Fay MP, Feuer EJ, Midthune DN.Permutation tests for joinpoint regressionwith applications to cancer rates. Stat Med2000;19:335–51.

18. Australian Bureau of Statistics. 3201.0Population by Age and Sex, AustralianStates and Territories. Canberra: AustralianBureau of Statistics (Catalogue number3201.0) (www.abs.gov.au), 2008.

19. Downing A, Newton-Bishop JA, FormanD. Recent trends in cutaneous malignantmelanoma in the Yorkshire region ofEngland; incidence, mortality and survivalin relation to stage of disease, 1993–2003.Br J Cancer 2006;95:91–5.

20. Dennis LK. Analysis of the melanomaepidemic, both apparent and real: datafrom the 1973 through 1994 surveillance,epidemiology, and end results programregistry. Arch Dermatol 1999;135:275–80.

21. Valentin SM, Sanchez JL, Figueroa LD,Nazario CM. Epidemiology of melanomain Puerto Rico, 1987–2002. PR Health Sci J2007;26:343–8.

22. Lasithiotakis KG, Leiter U, Gorkievicz R,Eigentler T, Breuninger H, Metzler G,Strobel W, Garbe C. The incidence andmortality of cutaneous melanoma inSouthern Germany: trends by anatomic siteand pathologic characteristics, 1976 to2003. Cancer 2006;107:1331–9.

23. Montella A, Gavin A, Middleton R, AutierP, Boniol M. Cutaneous melanomamortality starting to change: a study oftrends in Northern Ireland. Eur J Cancer2009;45:2360–6.

24. Criscione VD, Weinstock MA. Melanomathickness trends in the United States, 1988–2006. J Invest Dermatol 2010;130:793–7.

25. Burton RC, Armstrong BK. Currentmelanoma epidemic: a nonmetastasizing formof melanoma?World J Surg 1995;19:330–3.

26. Rees JL. The melanoma epidemic: realityand artefact. BMJ 1996;312:137–8.

27. Liu W, Dowling JP, Murray WK, McArthurGA, Thompson JF, Wolfe R, Kelly JW. Rateof growth in melanomas: characteristics andassociations of rapidly growing melanomas.Arch Dermatol 2006;142:1551–8.

28. Coory M, Smithers M, Aitken J, Baade P,Ring I. Urban-rural differences in survivalfrom cutaneous melanoma in Queensland.Aust NZJ Public Health 2006;30:71–4.

29. SACR. Epidemiology of cancer in SouthAustralia: incidence, mortality and survival1977–1997. Adelaide: South AustralianCancer Registry, 1998.

30. Barnhill RL, Fine JA, Roush GC, Berwick M.Predicting five-year outcome for patients withcutaneous melanoma in a population-basedstudy. Cancer 1996;78:427–32.

31. Baade P, Coory M. Trends in melanomamortality in Australia: 1950–2002 and theirimplications for melanoma control. AustNZ J Public Health 2005;29:383–6.

32. Whiteman DC, Whiteman CA, Green AC.Childhood sun exposure as a risk factor formelanoma: a systematic review ofepidemiologic studies. Cancer Causes Control2001;12:69–82.

Epidemiology

Baade et al. 177


33. Montague M, Borland R, Sinclair C. Slip!Slop! Slap! and SunSmart, 1980–2000: skincancer control and 20 years of population-based campaigning. Health Educ Behav 2001;28:290–305.

34. Staples MP, Elwood M, Burton RC,Williams JL, Marks R, Giles GG. Non-melanoma skin cancer in Australia: the2002 national survey and trends since1985. Med J Aust 2006;184:6–10.

35. Whiteman DC, Bray CA, Siskind V, GreenAC, Hole DJ, Mackie RM. Changes in theincidence of cutaneous melanoma in the

west of Scotland and Queensland,Australia: hope for health promotion? EurJ Cancer Prev 2008;17:243–50.

36. Youl PH, Baade PD, Parekh S, English D,Elwood M, Aitken JF. Association betweenmelanoma thickness, clinical skinexamination and socioeconomic status:results of a large population-based study. IntJ Cancer 2010 Jul 6. [Epub ahead of print].

37. Aitken JF, Elwood M, Baade PD, Youl P,English D. Clinical whole-body skinexamination reduces the incidence of thickmelanomas. Int J Cancer 2010;126:450–8.

38. de Vries E, Boniol M, Dore JF, CoeberghJW. Lower incidence rates but thickermelanomas in Eastern Europe before 1992:a comparison with Western Europe. Eur JCancer 2004;40:1045–52.

39. Baade P, Carriere P, Fritschi L. Trends intesticular germ cell cancer incidence inAustralia.Cancer Causes Control 2008;19:1043–9.

40. Shih ST, Carter R, Sinclair C,Mihalopoulos C, Vos T. Economicevaluation of skin cancer prevention inAustralia. Prev Med 2009;49:449–53.

Epidemiology



Time trends and latitudinal differences in melanoma thickness distribution in Australia, 1990–2006

Documents

Transcript of Time trends and latitudinal differences in melanoma thickness distribution in Australia, 1990–2006