Melanoma tumour thickness in Norway

An analysis of incidence and survival, 1983–2019

Raju Rimal

Department of Biostatistics
University of Oslo

24 November 2022

Research is what I’m doing when I don’t know what I’m doing.

Wernher von Braun (1912–1977)

A German-American physicist

Scientific research is one of the most exciting and rewarding of occupations.

Frederick Sanger (1918–2013)

An English biochemist with two Nobel Prize in Chemistry

What am I doing and why?

• Cutaneous melanoma (CM) is the most aggressive and lethal form of skin cancer.

• Increasing dramatically in fair skinned population.
• Norway is ranked fifth in incidence and third in mortality worldwide.

• It’s highly curable if caught early.
• Tumour thickness plays an important role

Outline

• Background

• Data and Methods

• Incidence rate and trend

• Perspectives

Research Objective

To study melanoma incidence and their trend by tumour thickness, overall and in important subgroups such as sex, age and anatomic sites, in a unique nationwide case series over a 35-year time period.

Background

Overall incidence and mortality in Norway

Data from Cancer Registry of Norway (CRN)

• Histologically verified data
• Tumour thickness recorded since 1980 are now digitized
• Melanoma registry established in 2008
• Here we will use data from 1983 to 2019

Melanoma and tumour thickness

T-categories

Diagnosis, thickness pattern, and T categories

Terminal digit preference, in diagnosis process. Cut-points can change the course of prognosis and follow-up. Need more exploration.

Data & Methods

Basic characteristics of data

Age at diagnosis

Age at diagnosis has increased in the recent period than earlier.

Basic characteristics of data

Tumour thickness

The thickness of tumour at diagnosis has decreased in the recent period than earlier.

Missing thickness and imputation

Proportion of cases with missing thickness

• 30 imputations
• Predictors: log-thickness, anatomic site, melanoma sub-type, age, sex, etc

Incidence rate, adjustment, and segmented regression

Incidence Rate

Age-adjustment

Segmented Regression

\[\mathcal{r} = \frac{\text{# melanoma patients}}{\text{# people at risk}}\]

Rate for each year of diagnosis, age-group and other strata are computed for further analysis.

Incidence rate, adjustment, and segmented regression

Incidence Rate

Age-adjustment

Segmented Regression

\[\mathcal{r}_a = \left[\sum_{i=1}^{a_k}\left(\mathcal{r}_i \times a_i^s\right)\right] \times 100,000\]

where,

\(\mathcal{r}_i\) = \(\mathcal{a}_i^s\) =

incidence rate of age-group \(i\) prop. of standard population for age-group \(i\)

Incidence rate, adjustment, and segmented regression

Incidence Rate

Age-adjustment

Segmented Regression

\[\log\left(\mathcal{r}_a\right)_k = \mathcal{f}\left(y_k\right) + \varepsilon\]

where,

\(y_k\) \(y_k\)

=

\(\beta_1 y_k + \beta_2 \left(y_k - \psi\right)_+\) is the year \(k\)

• Identify segments
• Annual percentage change

Results

Incidence over time

Complete Cases: only used cases with non-missing thumour thickness

Imputed Data: Earlier periods have more missing cases. More missing leads to more imputation and raise in incidence.

Segmented Regression: Change point mainly in T1 was detected in early 90s and mid 2000s.

Trend: Added spline shows a clear plateau in T1. Sharp raise in incidence in the recent years mainly in thinner cases (T1 and T2).

Trend: T1 has largest and sharply incidence of all. (Table in next slide.)

Trend: Women have higher incidence in thin cases while men have higher incidence in thicker cases.

Age-adjusted incidence rate by sex and T category

Changes in melanoma incidence rate

• Trend 1
• Trend 2
• Trend 3
• Trend 4

	Trend 1		1983–2019
	APC (95% CI)	Period	AAPC (95% CI)
Women (n=23,619)1
T1 (≤1.0 mm)	6.4 (3.1, 9.9)	1983–1989	3.1 (2.7, 3.5)
T2 (>1.0–2.0 mm)	0.7 (-0.6, 2.0)	1983–1997	2.0 (1.6, 2.5)
T3 (>2.0–4.0 mm)	-1.5 (-2.8, -0.1)	1983–1997	-0.1 (-0.5, 0.3)
T4 (>4.0 mm)	-2.8 (-6.6, 1.2)	1983–1989	0.9 (0.4, 1.4)
Men (n=22,193)1
T1 (≤1.0 mm)	10.1 (7.5, 12.7)	1983–1990	4.5 (4.1, 4.9)
T2 (>1.0–2.0 mm)	2.3 (1.5, 3.1)	1983–2002	2.9 (2.5, 3.3)
T3 (>2.0–4.0 mm)	7.1 (3.1, 11.3)	1983–1988	1.3 (0.8, 1.8)
T4 (>4.0 mm)	1.0 (0.3, 1.7)	1983–2005	1.3 (0.9, 1.7)
Trend refers to the segments separated by the join-points.
1 Average number of cases over 30 imputations

	Trend 2		1983–2019
	APC (95% CI)	Period	AAPC (95% CI)
Women (n=23,619)1
T1 (≤1.0 mm)	-0.0 (-1.0, 1.0)	1990–2004	3.1 (2.7, 3.5)
T2 (>1.0–2.0 mm)	3.0 (2.1, 3.9)	1998–2019	2.0 (1.6, 2.5)
T3 (>2.0–4.0 mm)	0.9 (-0.0, 1.7)	1998–2019	-0.1 (-0.5, 0.3)
T4 (>4.0 mm)	1.8 (1.2, 2.4)	1990–2019	0.9 (0.4, 1.4)
Men (n=22,193)1
T1 (≤1.0 mm)	-1.0 (-2.4, 0.3)	1991–2003	4.5 (4.1, 4.9)
T2 (>1.0–2.0 mm)	3.6 (2.4, 4.9)	2003–2019	2.9 (2.5, 3.3)
T3 (>2.0–4.0 mm)	-2.6 (-6.3, 1.2)	1989–1995	1.3 (0.8, 1.8)
T4 (>4.0 mm)	14.9 (-1.7, 34.2)	2006–2007	1.3 (0.9, 1.7)
Trend refers to the segments separated by the join-points.
1 Average number of cases over 30 imputations

	Trend 3		1983–2019
	APC (95% CI)	Period	AAPC (95% CI)
Women (n=23,619)1
T1 (≤1.0 mm)	5.0 (3.9, 6.1)	2005–2019	3.1 (2.7, 3.5)
T2 (>1.0–2.0 mm)	-	-	2.0 (1.6, 2.5)
T3 (>2.0–4.0 mm)	-	-	-0.1 (-0.5, 0.3)
T4 (>4.0 mm)	-	-	0.9 (0.4, 1.4)
Men (n=22,193)1
T1 (≤1.0 mm)	6.3 (5.1, 7.5)	2004–2019	4.5 (4.1, 4.9)
T2 (>1.0–2.0 mm)	-	-	2.9 (2.5, 3.3)
T3 (>2.0–4.0 mm)	2.5 (1.2, 3.7)	1996–2011	1.3 (0.8, 1.8)
T4 (>4.0 mm)	-0.6 (-2.7, 1.5)	2008–2019	1.3 (0.9, 1.7)
Trend refers to the segments separated by the join-points.
1 Average number of cases over 30 imputations

	Trend 4		1983–2019
	APC (95% CI)	Period	AAPC (95% CI)
Women (n=23,619)1
T1 (≤1.0 mm)	-	-	3.1 (2.7, 3.5)
T2 (>1.0–2.0 mm)	-	-	2.0 (1.6, 2.5)
T3 (>2.0–4.0 mm)	-	-	-0.1 (-0.5, 0.3)
T4 (>4.0 mm)	-	-	0.9 (0.4, 1.4)
Men (n=22,193)1
T1 (≤1.0 mm)	-	-	4.5 (4.1, 4.9)
T2 (>1.0–2.0 mm)	-	-	2.9 (2.5, 3.3)
T3 (>2.0–4.0 mm)	-1.9 (-5.0, 1.2)	2012–2019	1.3 (0.8, 1.8)
T4 (>4.0 mm)	-	-	1.3 (0.9, 1.7)
Trend refers to the segments separated by the join-points.
1 Average number of cases over 30 imputations

Incidence rate by anatomic site

• Overall
• Women
• Men

In overall, highest incidence in Trunk (see scale), followed by lower-limbs and upper-limbs.

In head and neck, thicker cases are also increasing particularly in the recent years.

In trunk, men have higher incidence than women. In upper and lower-limbs, women have higher incidence than men.

Incidence rate by melanoma sub-type

• Overall
• Women
• Men

In overall, superficial spreading melanoma is most common of all in men, women and overall.

In nodular melanoma, thicker cases have higher incidence than thinner. These are also more fatal.

In lentigo maligna melanoma, though less fatal, has sharp raise within thinner cases in the recent year.

Incidence by age, period of diagnosis and cohort

Dark color represents higher incidence. The high recent raise in incidence of thinner melanoma is found mostly in elderly of age 50 to 80.

The difference in the incidence of thicker melanoma between young and old is more pronounced in the recent years compared to previous year.

Wrap-up

Further study

Upcoming study

Cut-point analysis

The cut-points are important criteria for risk assessment, diagnosis and follow-up

Survival

Survival by tumour thickess gives more clear explanation on both the increase in incidence and high mortality in Norway due to melanoma

Need more exploration

Plateau

Reason behind the plateau is still unknown and further study is required.

Summary

• Awareness may have contributed to the rapid raise in melanoma incidence.

• Overdiagnosis may be considered but unable to explain the increase in thicker cases.
• Awareness focused on elderly males may be effective for early detection.

• Both long-term and short term effect may be the reason behind the plateau.
• Change in diagnosis practices, life-style, awareness may be the reason. More data and simulation studies is required.

Collaborators

Coauthors

Raju Rimal

Trude E Robsahm

Adele Green

Reza Ghiasvand

Corina S Rueegg

Assia Bassarova

Petter Gjersvik

Elisabete Weiderpass

Odd O Aalen

Bjørn Møller

Marit B Veierød

Funded by: The Research Council of Norway.

References

Gershenwald, Jeffrey E., and Richard A. Scolyer. 2018. “Melanoma Staging: American Joint Committee on Cancer (AJCC) 8th Edition and Beyond.” Annals of Surgical Oncology 25 (8): 2105–10. https://doi.org/10.1245/s10434-018-6513-7.

Kreftregisteret,. 2021. “Årsrapport 2020 Med Resultater Og Forbedringstiltak Fra Nasjonalt Kvalitetsregister for Melanom.” Oslo: Kreftregisteret.

Larsen, I. K., M. Smastuen, T. B. Johannesen, F. Langmark, D. M. Parkin, F. Bray, and B. Moller. 2009. “Data Quality at the Cancer Registry of Norway: An Overview of Comparability, Completeness, Validity and Timeliness.” Eur J Cancer 45 (7): 1218–31. https://doi.org/crj5pz.

Muggeo, Vito M. R. 2008. “Segmented: An r Package to Fit Regression Models with Broken-Line Relationships.” R News 8 (1): 20–25. https://cran.r-project.org/doc/Rnews/.

Norway, Cancer Registry of. 2022a. “Cancer in Norway 2021: Cancer Incidence, Mortality, Survival and Prevalence in Norway.” 0806-3621. Cancer Registry of Norway. https://www.kreftregisteret.no/globalassets/cancer-in-norway/2021/cin_report.pdf.

———. 2022b. “Cancer in Norway 2021: Technical Supplement: Statistical Methods.” 0806-3621. Oslo: Cancer Registry of Norway. https://www.kreftregisteret.no/globalassets/cancer-in-norway/2021/cin_2021_tech_sup_202206080716.pdf.

R Core Team. 2020. “R: A Language and Environment for Statistical Computing.” Manual. Vienna, Austria. https://www.R-project.org/.

Robsahm, T. E., P. Helsing, Y. Nilssen, L. Vos, S. M. H. Rizvi, L. A. Akslen, and M. B. Veierod. 2018. “High Mortality Due to Cutaneous Melanoma in Norway: A Study of Prognostic Factors in a Nationwide Cancer Registry.” Clin Epidemiol 10: 537–48. https://doi.org/gdkd9q.

Robsahm, Trude E., Gjøril Bergva, Unn E. Hestvik, and Bjørn Møller. 2013. “Sex Differences in Rising Trends of Cutaneous Malignant Melanoma in Norway, 1954-2008.” Melanoma Research 23 (1): 70–78. https://doi.org/t.

Veierød, M. B., C. M. Page, S. Aaserud, A. Bassarova, K. D. Jacobsen, P. Helsing, and T. E. Robsahm. 2018. “Melanoma Staging: Varying Precision and Terminal Digit Clustering in Breslow Thickness Data Is Evident in a Population-Based Study.” Journal of the American Academy of Dermatology 79 (1): 118–125.e1. https://doi.org/gdzrsw.

Welch, H. Gilbert, Benjamin L. Mazer, and Adewole S. Adamson. 2021. “The Rapid Rise in Cutaneous Melanoma Diagnoses.” New England Journal of Medicine 384 (1): 72–79. https://doi.org/gm6vs4.

Whiteman, David C., Adele C. Green, and Catherine M. Olsen. 2016. “The Growing Burden of Invasive Melanoma: Projections of Incidence Rates and Numbers of New Cases in Six Susceptible Populations Through 2031.” The Journal of Investigative Dermatology 136 (6): 1161–71. https://doi.org/f8psv4.