The assumption on which this study is based is that tumors grow exponentially (Fig. 1).9 Although it could be argued that tumor growth follows a Gompertz, sigmoid-like, curve, we assumed that the models would be superimposable for relatively small tumors.10

Figure 1.

Theoretical exponential growth model for skin tumors.

(0.05MB).

The exponential function that links the dependent variable y (tumor size) to the independent variable x (time) is expressed as follows:

To estimate the exponential growth of a tumor, it is necessary to know y (size) and x (time) at 2 moments in time. In brief, y1 is the size of the tumor when it is first noticed by the patient and/or a relative, y2 is the size of the tumor at the time of surgery, x1 is the time when the tumor is first noticed, and x2 is the time of surgery. Asy1 (the size of the tumor when it is first noticed) is unknown, it was estimated for both SCCs and melanomas.

To estimatey1 for SCCs, we used data from a series of 6 cases that provided details of tumor size at 2 points in time (Fig. 2) (Appendix B Table 1 of the supplementary material). With this information, we calculated the exponential growth function for each of the 6 tumors (Fig. 3), where a was the diameter of the tumor at y1(when it was first noticed by the patient and/or a relative). The mean of the 6 measurements (a = 8.75 mm) was used as coefficient a to estimate the exponential growth of the SCCs in our sample. The same method was applied to estimate SCC thickness, which was known for 4 of the 6 tumors. Tumor thickness at diagnosis (y1) was assumed to be proportional to the diameter calculated for the same time point. The resulting thickness was 1.2 mm, which was used as coefficient a to calculate the exponential function for each of the SCCs in the study sample.

Figure 2.

Squamous cell carcinoma of the scalp that grew from 25 mm to 42 mm over 3 months.

(0.11MB).

Figure 3.

Exponential growth function for a squamous cell carcinoma in patient 1.

(0.14MB).

We were unable to perform similar measurements for melanomas, as we did not have details of mean Breslow thickness at 2 different time points. We therefore established a thickness of 0.2 mm for coefficient a based on the assumption that this would be the minimum vertical growth needed for the tumor to be noticed by the patient; x2 was defined as the time from when the patient noticed the tumor to the time of surgery.

Once we had estimated initial tumor size, we obtained coefficient b of the exponential function as follows:

With this information, we then estimated tumor size and thickness for SCCs and tumor thickness for melanomas after a hypothetical surgical delay of 1, 2, and 3 months.

We analyzed 200 SCCs and 1000 melanomas randomly selected from 2 databases.

The primary outcome measure was change in estimated tumor size and/or thickness at 1, 2, and 3 months. The results were categorized by size and thickness. SCCs were classified using a similar system to the SCC staging system of the American Joint Committee on Cancer (AJCC)11: T1 (diameter ≤ 2 cm), T2 (diameter 2−4 cm), and T3 (diameter > 4 cm or thickness > 6 mm). Melanomas were grouped into 6 Breslow thickness categories, similar to those proposed by the AJCC: < 0.8 mm, 0.8−1 mm, > 1−2 mm, > 2−4 mm, > 4−6 mm, and > 6 mm. The last 2 categories of thick tumors are not part of the AJCC staging system but were included as ultrathick melanomas were recently reported to have prognostic significance.12

We also performed a stratified analysis by sex and age (≤ 80 vs. > 80 years for SCC and ≤ 70 vs. > 70 years for melanoma).12

To analyze the impact of surgical delays on survival, we estimated disease-specific survival (DSS) for patients with SCC and melanoma overall and stratified by sex and age using survival curves from the Squamata database for SCCs (n = 434) (Figs. 1–5 of the supplementary material) and the melanoma database of the Instituto Valenciano de Oncología (n = 1637) (Figs. 6–10 of the supplementary material). The resulting survival rates (overall and by sex and age) were then used to estimate survival for patients with SCC and melanoma after a potential surgical delay of 1, 2, and 3 months. The Kaplan-Meier method was used to estimate overall and sex- and age-stratified DSS for the SCCs and melanomas analyzed.

The study was approved by the ethics committee of Hospital Universitario Reina Sofía Córdoba (reference 3958, file no. 280).

Of the 200 SCCS with a known time from onset to surgery analyzed, 52% had a diameter of < 2 cm (T1). Just 13 (6.5%) had a diameter of 2−4 cm (T2) and 83 (41.5 %) had a diameter of > 4 cm or a thickness of > 6 mm (T3) (Table 1). The proportion of T3 tumors increased by 58.5% for a hypothetical surgical delay of 1 month, by 70.5% for a delay of 2 months, and by 72% for a delay of 3 months.

Estimated DSS for the overall group was 86.1% at 2 years, 80.2% at 5 years, and 70.8% at 10 years. Survival rates decreased progressively over time to 79.9% for a delay of 1 month, 72% for a delay of 2 months, and 65.6% for a delay of 3 months.

In the sex-stratified analysis, female sex was associated with worse 5-year DSS (75.2% vs. 83.3%) and 10-year DSS (65.3% vs. 77.6%). Again, survival rates decreased progressively over the 3 months. At the 3-month point, 5-year survival was 68.5% for women and 72.6% for men (Table 2).

In the age-stratified analysis, DSS was worse in patients aged >80 years at all the time points analyzed, with the greatest difference observed at month 3 (5-year survival of 79% for patients aged > 80 years vs. 58.7% for those aged ≤ 80 years) (Table 3).

Most of the tumors from the melanoma database were thin; tumors measuring < 0.8 mm and 0.8−1 mm accounted for 29.8% and 10.4% of all tumors, respectively. Thick and ultrathick melanomas comprised the smallest group (9.1% for melanomas with a Breslow thickness of 4−6 mm and 6.9% for those with a Breslow thickness of > 6 mm) (Table 4). The greatest percentage increase over time was observed for tumors with a thickness of > 6 mm, which accounted for 21.9% of all tumors at 1 month, 30.2% at 2 months, and 30.2% at 3 months. Five- and 10-year DSS rates for the overall group were 85.7% and 79.4% respectively, and these decreased significantly over the 3 months analyzed, with the lowest rates observed at month 3 (71.3% for 5-year survival and 65% for 10-year survival) (Table 5).

In the sex-stratified analysis, ultrathick melanomas (> 6 mm) were somewhat more common in men (8% vs. 5.8% for women) both initially and at the 3 time points analyzed. The increase in the proportion of these tumors was significant from 1 month onwards (27% for men vs. 16.8% for women) and peaked at month 3 (41.8% vs. 29%).

Five-year DSS for the overall group was better in women (89.2% vs 82.8%), but it decreased in both groups over the following months, reaching its lowest point at month 3 (77.5% for women vs. 70.7% for men).

In the age-stratified analysis, thick melanomas (> 4−6 mm) were more common in patients aged > 70 years (16.8% vs 7.2% in patients aged ≤ 70 years), while ultrathick melanomas were present to a similar degree in both age categories (5.8% vs. 5%). A significant increase in the proportion of ultrathick tumors was observed for each of the 3 time points analyzed, and was particularly high in both age groups at month 3 (52% for patients aged > 70 years vs. 31.3% for those aged ≤ 70 years) (Table 6).

Patients in the over-70 category experienced the greatest change in survival rates from baseline to month 3 (86.3% to 75.6% for 5-year survival and 75.7% to 61.9% for 10-year survival).