Survival in 31 670 patients with thin melanomas: a Swedish population-based study

Comprehensive analysis of clinical images contributions for melanoma classification using convolutional neural networks

BACKGROUND

Timely diagnosis plays a critical role in determining melanoma prognosis, prompting the development of deep learning models to aid clinicians. Questions persist regarding the efficacy of clinical images alone or in conjunction with dermoscopy images for model training. This study aims to compare the classification performance for melanoma of three types of CNN models: those trained on clinical images, dermoscopy images, and a combination of paired clinical and dermoscopy images from the same lesion.

MATERIALS AND METHODS

We divided 914 image pairs into training, validation, and test sets. Models were built using pre-trained Inception-ResNetV2 convolutional layers for feature extraction, followed by binary classification. Training comprised 20 models per CNN type using sets of random hyperparameters. Best models were chosen based on validation AUC-ROC.

RESULTS

Significant AUC-ROC differences were found between clinical versus dermoscopy models (0.661 vs. 0.869, p < 0.001) and clinical versus clinical + dermoscopy models (0.661 vs. 0.822, p = 0.001). Significant sensitivity differences were found between clinical and dermoscopy models (0.513 vs. 0.799, p = 0.01), dermoscopy versus clinical + dermoscopy models (0.799 vs. 1.000, p = 0.02), and clinical versus clinical + dermoscopy models (0.513 vs. 1.000, p < 0.001). Significant specificity differences were found between dermoscopy versus clinical + dermoscopy models (0.800 vs. 0.288, p < 0.001) and clinical versus clinical + dermoscopy models (0.650 vs. 0.288, p < 0.001).

CONCLUSION

CNN models trained on dermoscopy images outperformed those relying solely on clinical images under our study conditions. The potential advantages of incorporating paired clinical and dermoscopy images for CNN-based melanoma classification appear less clear based on our findings.

Survival trends for patients diagnosed with cutaneous malignant melanoma in the Nordic countries 1990-2016: The NORDCAN survival studies

BACKGROUND

The survival in patients diagnosed with cutaneous malignant melanoma (CMM) has improved in the Nordic countries in the last decades. It is of interest to know if these improvements are observed in all ages and for both women and men.

METHODS

Patients diagnosed with CMM in the Nordic countries in 1990-2016 were identified in the NORDCAN database. Flexible parametric relative survival models were fitted, except for Iceland where a non-parametric Pohar-Perme approach was used. A range of survival metrics were estimated by sex, both age-standardised and age-specific.

RESULTS

The 5-year relative survival improved in all countries, in both women and men and across age. While the improvement was more pronounced in men, women still had a higher survival at the end of the study period. The survival was generally high, with age-standardised estimates of 5-year relative survival towards the end of the study period ranging from 85% in Icelandic men to 95% in Danish women. The age-standardised and reference-adjusted 5-year crude probability of death due to CMM ranged from 5% in Danish and Swedish women to 13% in Icelandic men.

CONCLUSION

Although survival following CMM was relatively high in the Nordic countries in 1990, continued improvements in survival were observed throughout the study period in both women and men and across age.

Dermoscopy as a Tool for Identifying Potentially Metastatic Thin Melanoma: A Clinical-Dermoscopic and Histopathological Case-Control Study

Despite being early-stage tumors, thin cutaneous melanomas contribute significantly to mortality and have a rising incidence. A retrospective case-control study was performed to identify clinical-dermoscopic and histopathological variables linked to local and distant metastases in melanomas ≤0.8 mm. Data from 1 January 2000 to 22 June 2022 were analyzed from two Italian skin cancer referral centers. Sixteen patients with ≤0.8 mm melanomas developing metastases were studied compared to controls without metastases over 5 years. Statistical analysis involved Pearson's chi-squared test or Fisher's exact test. Of the 1396 cases, 1.1% progressed. The median diagnosis age was 49 (range 28-83), with 56.3% men and 43.7% women. The torso was the primary tumor site (43.7%). Clinically, lesions were pigmented (>10 mm diameter: 73.3%, ≥3 colors: 80%). Dermoscopically, the common features were white patches (73.3%), atypical vascular patterns (66.5%), blue-gray areas (60%) and absent pigment networks (60%). Histopathologically, all cases had adverse features like regression (87.4%), dermal mitoses (50%), a vertical growth phase (62.5%) and ulceration (12.5%). These findings were statistically significant compared to controls (p < 0.05). In ≤0.8 mm melanomas, specific clinical-dermoscopic traits might indicate higher metastatic potential when paired with adverse histopathological features.

Increasing melanoma incidence and survival trend shifts with improved melanoma-specific survival between 1990 and 2020 in Sweden

BACKGROUND

Melanoma-specific survival (MSS) is heterogenous between stages and is highly dependent on the T stage for primary localized disease. New systemic therapies for metastatic cutaneous melanoma (CM) have been introduced since 2012 in Sweden.

OBJECTIVES

To analyse the incidence and MSS time trends between 1990 and 2020 in Sweden.

METHODS

Nationwide, population-based and prospectively collected clinico-pathological data on invasive CM from the Swedish Melanoma Registry (SweMR) were analysed for survival trends between 1990 and 2020 using Kaplan-Meier curves and Cox proportional hazard ratios (HRs).

RESULTS

In total, 77 036 primary invasive CMs were diagnosed in 70 511 patients in Sweden between 1990 and 2020. The 5-year MSS [95% confidence interval (CI)] was 88.9% (88.3-89.4) for 1990-2000, 89.2% (88.7-89.6) for 2001-2010 and 93.0% (92.7-93.9) for 2011-2020. The odds ratios for being diagnosed with nodular melanoma (vs. superficial spreading melanoma) was significantly reduced by 20% (2001-2010) and by 46% (2011-2020) vs. the reference period 1990-2000. Overall, the MSS improved over both diagnostic periods (2001-2010 and 2011-2020) vs. the reference period 1990-2000 among men and women, respectively [HRmen: 2001-2010: 0.89 (95% CI 0.82-0.96) and 2011-2020: 0.62 (95% CI 0.56-0.67); HRwomen: 2001-2010: 0.82 (95% CI 0.74-0.91) and 2011-2020: 0.62 (95% CI 0.56-0.70)]. The risk of death from CM was significantly lower in all age groups for both men and women in the most recent diagnostic period (2011-2020 vs.1990-2000).

CONCLUSIONS

The results emphasize the improved MSS among men and women in Sweden. The MSS improvements, specifically for the period 2011-2020, may be correlated to the introduction of new systemic therapies and are here shown for the first time in detail for Sweden.

Survival in patients diagnosed with melanoma in situ compared to the general population. A Swedish population-based matched cohort study

Background

The incidence of melanoma in situ (MIS) is increasing even more rapidly than the incidence of cutaneous malignant melanoma (CMM). No previous studies have in detail investigated the survival in individuals diagnosed with MIS compared to the general population.

Methods

This population-based study included individuals with MIS diagnosed in Sweden between 2001 and 2010 and randomly selected MIS-free comparators matched on age, sex and county of residence. Exclusion criterion was a previous CMM. Data on socioeconomic status (SES) including educational level, income and marital status, comorbidity and cause of death were obtained from population-based registers. Overall survival (OS) was estimated by the Kaplan-Meier method. The mortality risk adjusted for SES and comorbidity was assessed by multivariable Cox regression analyses.

Findings

The survival analyses included 7963 cases and 39,662 comparators. Median age at MIS diagnosis were 63 (IQR 50-75) and 67 (IQR 57-76) years in women and men respectively. Median follow-up time was 120 months (IQR 102-152 months). In individuals with MIS, the ten-year OS was 77% (95% CI 0.76-0.78) compared to 72% (95% CI 0.72-0.73) in comparators. The MIS patients had a higher SES and lower comorbidity burden than the comparators. In a fully adjusted multivariable analysis, including 7772 cases and 38,103 comparators, the mortality was significantly lower in women with MIS (HR 0.88, 95% CI 0.82-0.94) compared to the background population. The corresponding estimate in men was HR 0.94 (95% CI 0.88-1.0). The risk of melanoma-related deaths during the study period was ten-fold higher in MIS patients.

Interpretation

Despite being at increased risk of developing CMM, MIS patients had a better OS compared to their matched comparators from the background population, findings which could not fully be explained by differences in SES and comorbidity. Our results are reassuring and should be communicated to patients who have been diagnosed with MIS.

Funding

Stiftelsen Onkologiska Klinikens i Uppsala Forskningsfond, Mats and Stefan Paulsson Trust, Medicon Village, Lund and Uppsala University Hospital (ALF).

Topographical and Chronological Analysis of Thin Cutaneous Melanoma's Progressions: A Multicentric Study

A great portion of cutaneous melanoma's diagnoses nowadays is attributed to thin tumors with up to 1 mm in Breslow thickness (hereafter thin CMs), which occasionally metastasize. The objective of this study was to identify thin CM's metastatic patterns from a topographical and chronological standpoint. A total of 204 cases of metastatic thin CMs from five specialized centers were included in the study, and corresponding data were collected (clinical, epidemiological, histopathological information of primary tumor and the number, anatomical site, and time intervals of their progressions). First progressions occurred locally, in regional lymph nodes, and in a distant site in 24%, 15% and 61% of cases, respectively, with a median time to first progression of 3.10 years (IQR: 1.09-5.24). The median elapsed time between the first and second progression and between the second and third progression was 0.82 (IQR: 0.34-1.97) and 0.49 (IQR: 0.21-2.30) years, respectively, while the median survival time was about 4 years since first progression. Furthermore, the sequences of locations and time intervals of the progressions were associated with the clinicopathological and demographic features of the primary tumors along with the features of the preceding progressions. In conclusion, the findings of this study describe the natural history of thin CMs, thus highlighting the necessity to identify subgroups of thin CMs at a higher risk for metastasis and contributing to the optimization of the management and follow-up of thin CM patients.

The Importance of Immunohistochemistry in the Evaluation of Tumor Depth of Primary Cutaneous Melanoma

Primary cutaneous melanoma (PCM) is the most aggressive skin malignancy, with an increasing incidence and significant mortality. Tumoral invasion, expressed as Breslow thickness, is routinely assessed on hematoxylin and eosin (HE), although this stain may sometimes underestimate the tumoral depth. The aim of this study was to compare the efficiency of the immunohistochemical (IHC) markers S-100, SOX10, Melan-A, and HMB-45 with HE for the evaluation of the Breslow thickness and staging of PCM. This retrospective study included 46 cases of PCM diagnosed between 2015 and 2022; for each case, the Breslow thickness using HE, S-100, SOX10, Melan-A, and HMB-45 was measured and the appropriate T category was recorded. The highest values of the Breslow thickness were observed for S-100. However, S-100, SOX10, and Melan-A provided statistically significant higher values of the Breslow thickness compared to HE, but no difference was noted between HMB-45 and HE. S-100 was most frequently involved in increasing the T category (26.1%), the majority of cases being upstaged from T1a to T1b. The IHC markers S-100, SOX10, and Melan-A contributed to better evaluation of the melanoma invasion, especially in thin melanomas, but their impact on staging and consecutive treatment remains to be confirmed by future studies.

Conditional survival in patients with stage IB-IIIA melanoma undergoing sentinel node biopsy in Queensland: A longitudinal study

BACKGROUND

Tumour characteristics such as thickness and ulceration, along with sentinel lymph node (SLN) status, have been essential in predicting survival in patients with locally invasive melanomas at the time of diagnosis. It is unclear if these prognostic factors are relevant 1, 2 or 5 years after diagnosis.

OBJECTIVES

The key aim of this project was to analyse conditional survival in a cohort of Queensland patients with stage IB to IIIA melanomas (American Joint Committee on Cancer's staging system, 8th version) and to test the relevance of clinicopathological prognostic factors for melanoma outcome after varying intervals of survival time.

METHODS

Patients with primary invasive cutaneous melanoma who were referred to a tertiary melanoma clinic and underwent SLN biopsy between 1994 and 2011 were ascertained. The effect of patient and tumour characteristics on melanoma survival were calculated using multivariate Cox proportional hazard models at diagnosis and at variable times after diagnosis.

RESULTS

The final analysis included 651 patients (average age 49 years, 55.5% male) with stage IB to IIIA melanoma. At diagnosis, and after 1 and 2 years survived, SLN positivity, thickness and ulceration were predictive of 10-year survival since diagnosis. However, once patients survived 5 years, only SLN status was predictive. Overall conditional melanoma survival improved with increasing time survived. Five years after diagnosis, 10-year conditional melanoma survival (MSS) was 91% (95% CI 86%-95%) compared with 85% (82%-88%) predicted at diagnosis. The improvement in MSS was observed mainly for Stage II melanoma patients and not for those with a positive SLN biopsy.

CONCLUSIONS

This study confirms the improvement of prognosis according to time survived since diagnosis suggesting that after 5 years survival the classic prognostic indicators may not have the same influence.

Predictors of Sentinel Lymph Node Metastasis in Patients with Thin Melanoma: An International Multi-institutional Collaboration

BACKGROUND

Consideration of sentinel lymph node biopsy (SLNB) is recommended for patients with T1b melanomas and T1a melanomas with high-risk features; however, the proportion of patients with actionable results is low. We aimed to identify factors predicting SLNB positivity in T1 melanomas by examining a multi-institutional international population.

METHODS

Data were extracted on patients with T1 cutaneous melanoma who underwent SLNB between 2005 and 2018 at five tertiary centers in Europe and Canada. Univariable and multivariable logistic regression analyses were performed to identify predictors of SLNB positivity.

RESULTS

Overall, 676 patients were analyzed. Most patients had one or more high-risk features: Breslow thickness 0.8-1 mm in 78.1% of patients, ulceration in 8.3%, mitotic rate > 1/mm² in 42.5%, Clark's level ≥ 4 in 34.3%, lymphovascular invasion in 1.4%, nodular histology in 2.9%, and absence of tumor-infiltrating lymphocytes in 14.4%. Fifty-three patients (7.8%) had a positive SLNB. Breslow thickness and mitotic rate independently predicted SLNB positivity. The odds of positive SLNB increased by 50% for each 0.1 mm increase in thickness past 0.7 mm (95% confidence interval [CI] 1.05-2.13) and by 22% for each mitosis per mm² (95% CI 1.06-1.41). Patients who had one excised node (vs. two or more) were three times less likely to have a positive SLNB (3.6% vs. 9.6%; odds ratio 2.9 [1.3-7.7]).

CONCLUSIONS

Our international multi-institutional data confirm that Breslow thickness and mitotic rate independently predict SLNB positivity in patients with T1 melanoma. Even within this highly selected population, the number needed to diagnose is 13:1 (7.8%), indicating that more work is required to identify additional predictors of sentinel node positivity.

Interobserver Agreement on Dermoscopic Features and their Associations with In Situ and Invasive Cutaneous Melanomas

Several melanoma-specific dermoscopic features have been described, some of which have been reported as indicative of in situ or invasive melanomas. To assess the usefulness of these features to differentiate between these 2 categories, a retrospective, single-centre investigation was conducted. Dermoscopic images of melanomas were reviewed by 7 independent dermatologists. Fleiss' kappa (κ) was used to analyse interobserver agreement of predefined features. Logistic regression and odds ratios were used to assess whether specific features correlated with melanoma in situ or invasive melanoma. Overall, 182 melanomas (101 melanoma in situ and 81 invasive melanomas) were included. The interobserver agreement for melanoma-specific features ranged from slight to substantial. Atypical blue-white structures (κ=0.62, 95% confidence interval 0.59-0.65) and shiny white lines (κ=0.61, 95% confidence interval 0.58-0.64) had a substantial interobserver agreement. These 2 features were also indicative of invasive melanomas >1.0 mm in Breslow thickness. Furthermore, regression/peppering correlated with thin invasive melanomas. The overall agreement for classification of the lesions as invasive or melanoma in situ was moderate (κ=0.52, 95% confidence interval 0.49-0.56).

Trend Shifts in Age-Specific Incidence for In Situ and Invasive Cutaneous Melanoma in Sweden

BACKGROUND

The incidence of invasive cutaneous melanoma (CM) is increasing in Sweden. The aim was to present age- and sex-specific trends of the age-standardised incidence and the average annual percentage change (AAPC) for in situ and invasive CM.

METHODS

Joinpoint regression models were used to analyse data from the Swedish Cancer Register and the Swedish Melanoma Registry 1997-2018 (N = 35,350 in situ CM; 59,932 CM).

RESULTS

The AAPC of CM for women was 4.5 (4.1-5.0; p < 0.001) for the period 1997-2018. For men, the APCC was 4.2 (3.0-5.4; p < 0.001), with a significantly higher annual percentage change (APC) for the period 2000-2018 (5.0; 4.6-5.4; p < 0.001) compared to 1997-1999. An increasing annual incidence of CM ≤ 0.6 mm and 0.7 mm Breslow tumour thickness was found for men with a significant incidence shift for the period 2006-2015, respectively. Similarly for women, with a significantly higher APC for CM ≤ 0.6 mm from 2005. The incidence of intermediate thick CM (2.1-4.0 mm) has not increased since 2011. The incidence of CM > 4.0 mm has been increasing among both sexes, with a significantly lower APC among women from 2005.

CONCLUSIONS

The incidence of in situ and low-risk CM ≤ 1.0 mm in tumour thickness has been rising among both sexes since the 2000s.

Development and Validation of Nomograms to Predict Local, Regional, and Distant Recurrence in Patients With Thin (T1) Melanomas

PURPOSE

Although the prognosis of patients with thin primary cutaneous melanomas (T1, ≤ 1.0 mm) is generally excellent, some develop recurrence. We sought to develop and validate a model predicting recurrences in patients with thin melanomas.

METHODS

A Dutch population-based cohort (n = 25,930, development set) and a cohort from an Australian melanoma treatment center (n = 2,968, validation set) were analyzed (median follow-up 6.7 and 12.0 years, respectively). Multivariable Cox models were generated for local, regional, and distant recurrence-free survival (RFS). Discrimination was assessed using Harrell's C-statistic for each outcome. Each nomogram performance was evaluated using calibration plots defining low-risk and high-risk groups as the lowest and top 5% of the nomogram risk score, respectively. The nomograms' C-statistics were compared with those of a model including the current American Joint Committee on Cancer staging parameters (T-stage and sentinel node status).

RESULTS

Local, regional, and distant recurrences were found in 209 (0.8%), 503 (1.9%), and 203 (0.8%) Dutch patients, respectively, and 23 (0.8%), 61 (2.1%), and 75 (2.5%) Australian patients, respectively. C-statistics of 0.79 (95% CI, 0.75 to 0.82) for local RFS, 0.77 (95% CI, 0.75 to 0.78) for regional RFS, and 0.80 (95% CI, 0.77 to 0.83) for distant RFS were obtained for the development model. External validation showed C-statistics of 0.80 (95% CI, 0.69 to 0.90), 0.76 (95% CI, 0.70 to 0.82), and 0.74 (95% CI, 0.69 to 0.80), respectively. Calibration plots showed a good match between predicted and observed rates. Using the nomogram, the C-statistic was increased by 9%-12% for the development cohort and by 11%-15% for the validation cohort, compared with a model including only T-stage and sentinel node status.

CONCLUSION

Most patients with thin melanomas have an excellent prognosis, but some develop recurrence. The presented nomograms can accurately identify a subgroup at high risk. An online calculator is available at www.melanomarisk.org.au.