Preliminary Fitzpatrick Results

Overview

The design of these tables is based on our research meeting with Dr. Shields on 6/2/2022. The current N for all FST tables (924) is correct given that 77 patients currently have no data for their Fitzpatrick Scale. The tables are a little crowded right now because I included results for patients with missing Fitzpatrick values and there is only a single person with FST Type V. Hopefully as we continue to gather FST data the missing column can be removed and more people will be found for the FST Type V category.

There is quite a lot to customize with the package I’m using to make these tables, so let me know if you’d like to switch up the order, change the spacing, or include different variables and I’ll try to figure it out!

Table 1

Demographics by FST (with P-values)

**Table 1. Demographics by Fitzpatrick Scale**
Demographics	Overall, N = 924¹	Fitzpatrick (FST) Scale						p-value²
Demographics	Overall, N = 924¹	Missing N = 227 (25%)¹	FST Type I N = 44 (4.8%)¹	FST Type II N = 610 (66%)¹	FST Type III N = 34 (3.7%)¹	FST Type IV N = 8 (0.9%)¹	FST Type V N = 1 (0.1%)¹	p-value²
Patient Age								<0.001
Mean (SD)	58 (14)	58 (14)	61 (14)	58 (14)	50 (14)	47 (12)	23 (NA)
Median (IQR)	60 (50, 68)	59 (48, 69)	62 (53, 69)	60 (51, 69)	53 (37, 60)	50 (36, 54)	23 (23, 23)
Range	10, 94	14, 92	29, 94	10, 90	19, 75	28, 65	23, 23
Sex								0.7
Male	479 (52%)	113 (50%)	21 (48%)	319 (52%)	21 (62%)	5 (62%)	0 (0%)
Female	445 (48%)	114 (50%)	23 (52%)	291 (48%)	13 (38%)	3 (38%)	1 (100%)
Melanocytosis								0.7
No	842 (97%)	209 (98%)	41 (98%)	554 (97%)	30 (91%)	7 (100%)	1 (100%)
Yes	29 (3.3%)	5 (2.3%)	1 (2.4%)	20 (3.5%)	3 (9.1%)	0 (0%)	0 (0%)
Heterochromia								0.072
No	855 (98%)	214 (100%)	41 (98%)	562 (98%)	30 (91%)	7 (100%)	1 (100%)
Yes	16 (1.8%)	0 (0%)	1 (2.4%)	12 (2.1%)	3 (9.1%)	0 (0%)	0 (0%)
Study Eye
Right Eye	468 (53%)	122 (54%)	22 (52%)	302 (52%)	19 (58%)	2 (29%)	1 (100%)
Left Eye	418 (47%)	105 (46%)	20 (48%)	274 (48%)	14 (42%)	5 (71%)	0 (0%)
Visual Acuity
20/20-20/50	685 (74%)	173 (76%)	32 (73%)	453 (74%)	23 (68%)	4 (50%)	0 (0%)
20/60-20/200	152 (16%)	38 (17%)	3 (6.8%)	102 (17%)	6 (18%)	3 (38%)	0 (0%)
20/400-NLP	87 (9.4%)	16 (7.0%)	9 (20%)	55 (9.0%)	5 (15%)	1 (12%)	1 (100%)
¹ n (%)
² Kruskal-Wallis rank sum test; Fisher's exact test

Tables 2 (A and B)

A) Clinical Features by FST

**Table 2A. Clinical Features by Fitzpatrick Scale**
Clinical Features	Overall, N = 924¹	Fitzpatrick (FST) Scale						p-value²
Clinical Features	Overall, N = 924¹	Missing N = 227 (25%)¹	FST Type I N = 44 (4.8%)¹	FST Type II N = 610 (66%)¹	FST Type III N = 34 (3.7%)¹	FST Type IV N = 8 (0.9%)¹	FST Type V N = 1 (0.1%)¹	p-value²
Distance to optic disc (mm)								0.050
Mean (SD)	4.6 (4.2)	3.9 (3.6)	4.9 (4.1)	4.8 (4.4)	4.2 (4.7)	7.1 (4.7)	0.0 (NA)
Median (IQR)	3.7 (1.0, 7.0)	3.0 (1.0, 6.0)	4.5 (1.8, 7.0)	4.0 (1.0, 7.0)	3.0 (0.0, 6.0)	5.0 (4.5, 8.0)	0.0 (0.0, 0.0)
Range	0.0, 20.0	0.0, 16.0	0.0, 15.0	0.0, 20.0	0.0, 16.0	3.0, 17.0	0.0, 0.0
Unknown	54	1	4	46	2	1	0
Distance to foveola (mm)								<0.001
Mean (SD)	4.3 (4.2)	3.5 (3.7)	5.4 (3.9)	4.6 (4.4)	4.0 (5.0)	5.9 (3.9)	0.0 (NA)
Median (IQR)	3.0 (1.0, 6.5)	2.0 (0.0, 6.0)	5.0 (3.0, 7.6)	3.0 (1.0, 7.0)	2.2 (0.0, 6.0)	5.0 (4.0, 6.0)	0.0 (0.0, 0.0)
Range	0.0, 18.4	0.0, 16.5	0.0, 17.0	0.0, 18.4	0.0, 17.0	2.0, 14.0	0.0, 0.0
Unknown	55	1	4	47	2	1	0
Largest basal diameter (mm)								0.006
Mean (SD)	12.2 (4.4)	11.7 (4.1)	14.2 (4.3)	12.2 (4.4)	12.2 (4.7)	15.2 (4.1)	17.0 (NA)
Median (IQR)	12.0 (9.0, 16.0)	11.0 (9.0, 14.0)	15.5 (10.0, 16.2)	12.0 (9.0, 16.0)	12.5 (8.2, 16.0)	15.5 (14.2, 18.2)	17.0 (17.0, 17.0)
Range	1.0, 24.0	4.0, 22.0	6.0, 24.0	1.0, 24.0	3.5, 20.0	7.0, 19.2	17.0, 17.0
Thickness at DFS (mm)								<0.001
Mean (SD)	5.7 (3.3)	4.7 (2.4)	7.4 (4.0)	5.9 (3.4)	6.1 (3.7)	8.3 (3.2)	4.7 (NA)
Median (IQR)	4.7 (3.0, 7.8)	3.9 (3.0, 5.8)	6.6 (4.7, 9.4)	5.0 (3.0, 8.2)	4.7 (2.9, 8.8)	9.2 (5.2, 10.9)	4.7 (4.7, 4.7)
Range	0.7, 20.4	1.3, 13.3	2.2, 20.4	0.7, 15.3	1.4, 16.0	3.7, 12.0	4.7, 4.7
Thickness at DLS (mm)								<0.001
Mean (SD)	3.19 (1.99)	2.82 (1.50)	3.90 (3.08)	3.19 (1.93)	4.06 (2.70)	6.91 (3.41)	6.50 (NA)
Median (IQR)	2.60 (1.90, 3.80)	2.40 (1.80, 3.60)	3.30 (2.25, 4.40)	2.65 (2.00, 3.80)	2.95 (2.28, 5.53)	8.20 (3.95, 9.25)	6.50 (6.50, 6.50)
Range	0.00, 19.80	1.00, 9.10	1.50, 19.80	0.00, 15.30	1.12, 12.00	2.70, 11.10	6.50, 6.50
Unknown	94	4	5	82	2	1	0
Change in thickness (DLS-DFS								0.005
Mean (SD)	-2.17 (2.23)	-1.80 (1.79)	-2.97 (2.50)	-2.32 (2.34)	-1.69 (1.88)	-1.39 (3.39)	1.80 (NA)
Median (IQR)	-1.50 (-3.38, -0.50)	-1.30 (-2.40, -0.60)	-2.90 (-4.35, -1.15)	-1.60 (-3.60, -0.60)	-1.00 (-3.33, -0.30)	-0.90 (-1.80, 0.00)	1.80 (1.80, 1.80)
Range	-13.50, 5.10	-10.50, 1.20	-8.90, 2.10	-13.50, 5.10	-6.30, 0.00	-8.10, 2.90	1.80, 1.80
Unknown	94	4	5	82	2	1	0
Percent change in thickness (DLS-DFS/DFS)								0.007
Mean (SD)	-0.35 (0.27)	-0.35 (0.22)	-0.39 (0.31)	-0.35 (0.29)	-0.26 (0.21)	-0.11 (0.39)	0.38 (NA)
Median (IQR)	-0.36 (-0.54, -0.18)	-0.35 (-0.53, -0.19)	-0.43 (-0.58, -0.29)	-0.37 (-0.55, -0.18)	-0.21 (-0.39, -0.11)	-0.08 (-0.26, 0.00)	0.38 (0.38, 0.38)
Range	-1.00, 2.12	-0.88, 0.26	-0.76, 0.81	-1.00, 2.12	-0.83, 0.00	-0.74, 0.55	0.38, 0.38
Unknown	94	4	5	82	2	1	0
Tumor epicenter
Choroid	838 (91%)	220 (97%)	40 (91%)	539 (89%)	32 (94%)	6 (75%)	1 (100%)
Ciliary Body	68 (7.4%)	7 (3.1%)	4 (9.1%)	53 (8.7%)	2 (5.9%)	2 (25%)	0 (0%)
Iris	16 (1.7%)	0 (0%)	0 (0%)	16 (2.6%)	0 (0%)	0 (0%)	0 (0%)
¹ n (%)
² Kruskal-Wallis rank sum test

B) Clinical Features by TCGA

**Table 2B. Clinical Features by TCGA**
Clinical Features	Overall, N = 1,001¹	TCGA Classification					p-value²
Clinical Features	Overall, N = 1,001¹	Missing N = 1 (<0.1%)¹	Group A N = 485 (48%)¹	Group B N = 141 (14%)¹	Group C N = 260 (26%)¹	Group D N = 114 (11%)¹	p-value²
Distance to optic disc (mm)							<0.001
Mean (SD)	4.5 (4.1)	8.0 (NA)	3.9 (3.9)	4.7 (4.1)	5.5 (4.3)	4.9 (4.3)
Median (IQR)	3.5 (1.0, 7.0)	8.0 (8.0, 8.0)	3.0 (0.5, 5.8)	4.0 (1.0, 6.6)	5.0 (2.0, 8.0)	5.0 (0.0, 8.0)
Range	0.0, 20.0	8.0, 8.0	0.0, 20.0	0.0, 18.0	0.0, 17.0	0.0, 18.0
Unknown	55	0	22	9	17	7
Distance to foveola (mm)							<0.001
Mean (SD)	4.3 (4.2)	8.0 (NA)	3.6 (3.9)	4.1 (3.7)	5.4 (4.6)	4.7 (4.3)
Median (IQR)	3.0 (1.0, 6.0)	8.0 (8.0, 8.0)	2.0 (0.3, 5.8)	3.2 (1.0, 6.0)	4.0 (2.0, 8.3)	3.0 (1.0, 8.0)
Range	0.0, 18.4	8.0, 8.0	0.0, 18.4	0.0, 15.0	0.0, 18.0	0.0, 17.0
Unknown	56	0	22	9	17	8
Largest basal diameter (mm)							<0.001
Mean (SD)	12.1 (4.4)	8.0 (NA)	10.5 (3.9)	12.7 (4.5)	13.6 (4.1)	15.3 (3.5)
Median (IQR)	12.0 (9.0, 16.0)	8.0 (8.0, 8.0)	10.0 (8.0, 13.0)	13.0 (9.0, 16.0)	14.0 (10.0, 16.0)	16.0 (14.0, 18.0)
Range	1.0, 24.0	8.0, 8.0	1.0, 22.0	2.0, 22.0	2.0, 24.0	6.0, 24.0
Thickness at DFS (mm)							<0.001
Mean (SD)	5.6 (3.3)	5.5 (NA)	4.4 (2.5)	6.2 (3.5)	6.7 (3.4)	7.6 (3.4)
Median (IQR)	4.7 (3.0, 7.7)	5.5 (5.5, 5.5)	3.5 (2.5, 5.5)	5.2 (3.2, 9.0)	6.0 (4.0, 9.2)	7.0 (4.5, 10.0)
Range	0.7, 20.4	5.5, 5.5	1.0, 14.1	1.3, 15.0	0.7, 16.0	2.1, 20.4
Thickness at DLS (mm)							<0.001
Mean (SD)	3.19 (1.99)	2.70 (NA)	2.64 (1.41)	3.46 (2.21)	3.86 (2.31)	3.82 (2.44)
Median (IQR)	2.60 (1.90, 3.80)	2.70 (2.70, 2.70)	2.30 (1.70, 3.10)	2.70 (2.00, 4.15)	3.30 (2.20, 4.88)	3.40 (2.50, 4.20)
Range	0.00, 19.80	2.70, 2.70	0.80, 12.00	0.80, 12.50	0.00, 15.30	1.30, 19.80
Unknown	169	0	72	26	50	21
Change in thickness (DLS-DFS							<0.001
Mean (SD)	-2.17 (2.23)	-2.80 (NA)	-1.57 (1.68)	-2.40 (2.48)	-2.74 (2.60)	-3.27 (2.35)
Median (IQR)	-1.50 (-3.33, -0.50)	-2.80 (-2.80, -2.80)	-1.10 (-2.20, -0.40)	-1.70 (-3.85, -0.75)	-2.25 (-4.00, -0.90)	-2.80 (-5.10, -1.50)
Range	-13.50, 5.10	-2.80, -2.80	-8.20, 1.70	-10.70, 5.10	-13.50, 2.90	-8.30, 1.50
Unknown	169	0	72	26	50	21
Percent change in thickness (DLS-DFS/DFS)							<0.001
Mean (SD)	-0.35 (0.27)	-0.51 (NA)	-0.31 (0.25)	-0.35 (0.36)	-0.37 (0.27)	-0.44 (0.24)
Median (IQR)	-0.36 (-0.54, -0.18)	-0.51 (-0.51, -0.51)	-0.33 (-0.48, -0.16)	-0.41 (-0.59, -0.18)	-0.38 (-0.57, -0.20)	-0.49 (-0.60, -0.31)
Range	-1.00, 2.12	-0.51, -0.51	-0.85, 1.31	-0.77, 2.12	-1.00, 0.81	-0.84, 0.45
Unknown	169	0	72	26	50	21
Tumor epicenter
Choroid	914 (91%)	1 (100%)	458 (94%)	126 (90%)	226 (87%)	103 (90%)
Ciliary Body	68 (6.8%)	0 (0%)	20 (4.1%)	11 (7.9%)	26 (10%)	11 (9.6%)
Iris	17 (1.7%)	0 (0%)	7 (1.4%)	3 (2.1%)	7 (2.7%)	0 (0%)
¹ n (%)
² Kruskal-Wallis rank sum test

Table 3

Clinical Treatment/Management by FST

**Table 3. Clinical Treatment/Management by Fitzpatrick Scale**
Clinical Treatment/Management	Overall, N = 924¹	Fitzpatrick (FST) Scale						p-value²
Clinical Treatment/Management	Overall, N = 924¹	Missing N = 227 (25%)¹	FST Type I N = 44 (4.8%)¹	FST Type II N = 610 (66%)¹	FST Type III N = 34 (3.7%)¹	FST Type IV N = 8 (0.9%)¹	FST Type V N = 1 (0.1%)¹	p-value²
Time from DFS to date of treatment (mons)								0.12
Mean (SD)	1.58 (8.91)	1.21 (7.49)	0.48 (1.21)	1.88 (9.95)	0.32 (0.65)	0.64 (0.81)	3.83 (NA)
Median (IQR)	0.10 (0.10, 0.33)	0.10 (0.10, 0.33)	0.10 (0.10, 0.33)	0.10 (0.10, 0.33)	0.10 (0.07, 0.33)	0.45 (0.10, 0.63)	3.83 (3.83, 3.83)
Range	-28.43, 122.07	-8.37, 97.10	0.00, 6.63	-28.43, 122.07	-0.03, 3.60	0.10, 2.53	3.83, 3.83
Treatment
Plaque	705 (76%)	147 (65%)	28 (64%)	497 (82%)	26 (76%)	6 (75%)	1 (100%)
Plaque and TTT	149 (16%)	71 (31%)	10 (23%)	60 (9.9%)	6 (18%)	2 (25%)	0 (0%)
Plaque and PDT	9 (1.0%)	5 (2.2%)	2 (4.5%)	2 (0.3%)	0 (0%)	0 (0%)	0 (0%)
Enucleation	56 (6.1%)	4 (1.8%)	4 (9.1%)	46 (7.6%)	2 (5.9%)	0 (0%)	0 (0%)
PLSU	2 (0.2%)	0 (0%)	0 (0%)	2 (0.3%)	0 (0%)	0 (0%)	0 (0%)
Stereotactic Radiatio,	1 (0.1%)	0 (0%)	0 (0%)	1 (0.2%)	0 (0%)	0 (0%)	0 (0%)
No Follow up after FNAB	1 (0.1%)	0 (0%)	0 (0%)	1 (0.2%)	0 (0%)	0 (0%)	0 (0%)
Observation	0 (0%)	0 (0%)	0 (0%)	0 (0%)	0 (0%)	0 (0%)	0 (0%)
Fine Needle Aspiration Biopsy (FNAB)								>0.9
No	2 (0.2%)	0 (0%)	0 (0%)	2 (0.3%)	0 (0%)	0 (0%)	0 (0%)
Yes	922 (100%)	227 (100%)	44 (100%)	608 (100%)	34 (100%)	8 (100%)	1 (100%)
Route of FNAB
Not done	4 (0.4%)	0 (0%)	1 (2.3%)	3 (0.5%)	0 (0%)	0 (0%)	0 (0%)
Trans-scleral	244 (27%)	49 (22%)	17 (39%)	164 (27%)	8 (24%)	5 (62%)	1 (100%)
Pars Plana	598 (65%)	171 (76%)	22 (50%)	379 (63%)	23 (70%)	3 (38%)	0 (0%)
Clear Cornea	22 (2.4%)	1 (0.4%)	1 (2.3%)	20 (3.3%)	0 (0%)	0 (0%)	0 (0%)
Enucleation	49 (5.3%)	4 (1.8%)	3 (6.8%)	40 (6.6%)	2 (6.1%)	0 (0%)	0 (0%)
Cells Visible
No	185 (21%)	64 (31%)	10 (24%)	104 (18%)	6 (18%)	1 (12%)	0 (NA%)
Yes	682 (79%)	142 (69%)	31 (76%)	475 (82%)	27 (82%)	7 (88%)	0 (NA%)
¹ n (%)
² Kruskal-Wallis rank sum test; Fisher's exact test

Table 4

Outcomes by FST

**Table 4. Outcomes by Fitzpatrick Scale**
Outcomes	Overall, N = 924¹	Fitzpatrick (FST) Scale						p-value²
Outcomes	Overall, N = 924¹	Missing N = 227 (25%)¹	FST Type I N = 44 (4.8%)¹	FST Type II N = 610 (66%)¹	FST Type III N = 34 (3.7%)¹	FST Type IV N = 8 (0.9%)¹	FST Type V N = 1 (0.1%)¹	p-value²
Follow-up period (mons)								0.008
Mean (SD)	42 (35)	37 (34)	36 (29)	45 (36)	38 (38)	42 (41)	4 (NA)
Median (IQR)	32 (17, 62)	27 (9, 57)	33 (17, 42)	34 (20, 67)	25 (12, 62)	44 (16, 46)	4 (4, 4)
Range	0, 210	0, 210	1, 111	0, 198	0, 154	0, 124	4, 4
Unknown	43	2	3	36	1	1	0
Local Recurrence								0.033
No	854 (97%)	219 (98%)	40 (98%)	558 (97%)	30 (94%)	6 (86%)	1 (100%)
Yes	26 (3.0%)	5 (2.2%)	1 (2.4%)	17 (3.0%)	2 (6.2%)	1 (14%)	0 (0%)
Vision Loss
No	400 (45%)	123 (55%)	20 (48%)	237 (41%)	15 (45%)	4 (57%)	1 (100%)
Yes	483 (55%)	102 (45%)	22 (52%)	338 (59%)	18 (55%)	3 (43%)	0 (0%)
Death
No	825 (94%)	211 (93%)	34 (83%)	541 (94%)	31 (97%)	7 (100%)	1 (100%)
Yes	57 (6.5%)	15 (6.6%)	7 (17%)	34 (5.9%)	1 (3.1%)	0 (0%)	0 (0%)
Death from malignant melanoma								0.2
No	43 (77%)	11 (65%)	6 (86%)	25 (81%)	1 (100%)	0 (NA%)	0 (NA%)
Yes	13 (23%)	6 (35%)	1 (14%)	6 (19%)	0 (0%)	0 (NA%)	0 (NA%)
¹ n (%)
² Kruskal-Wallis rank sum test; Fisher's exact test

Tables 5 (A, B and C)

A) Univariate Analysis

**Table 5A. Metastasis by Fitzpatrick Scale**
Outcomes	Overall, N = 924¹	Fitzpatrick (FST) Scale						p-value²
Outcomes	Overall, N = 924¹	Missing N = 227 (25%)¹	FST Type I N = 44 (4.8%)¹	FST Type II N = 610 (66%)¹	FST Type III N = 34 (3.7%)¹	FST Type IV N = 8 (0.9%)¹	FST Type V N = 1 (0.1%)¹	p-value²
Metastasis
No	726 (82%)	190 (85%)	27 (66%)	472 (82%)	29 (91%)	7 (100%)	1 (100%)
Yes	154 (18%)	34 (15%)	14 (34%)	103 (18%)	3 (9.4%)	0 (0%)	0 (0%)
Location of Metastasis								>0.9
Liver metastasis	120 (78%)	28 (82%)	10 (71%)	80 (78%)	2 (67%)	0 (NA%)	0 (NA%)
Lung metastasis	4 (2.6%)	1 (2.9%)	0 (0%)	3 (2.9%)	0 (0%)	0 (NA%)	0 (NA%)
Liver and lung metastasis	26 (17%)	3 (8.8%)	4 (29%)	19 (19%)	0 (0%)	0 (NA%)	0 (NA%)
Other metastasis	3 (2.0%)	2 (5.9%)	0 (0%)	0 (0%)	1 (33%)	0 (NA%)	0 (NA%)
¹ n (%)
² Fisher's exact test

B) Univariate Binomial Regression

Interpretation: as the Fitzpatrick scale increases (i.e. as skin color becomes darker), the odds of metastasis decrease!

**Table 5B. Univariate Regression of Metastasis based on Fitzpatrick Scale**
Metastasis	N	Event N¹	OR²	95% CI²	p-value	q-value³
Fitzpatrick (FST) Scale	648				0.020	0.020
FST Type I		14	—	—
FST Type II		103	0.42	0.22, 0.85
FST Type III		3	0.20	0.04, 0.69
¹ There were no instances of metastasis for FST types IV and V
² OR = Odds Ratio, CI = Confidence Interval
³ False discovery rate correction for multiple testing

C) Multivariate Regression

Adjustments will be made for: age, tumor largest diameter, and thickness

I will try both Poisson and Binomial distributions for the glm link function (not sure which one is more accurate). A binomial logistic regression was used in the Genetic Analysis of Uveal Melanoma paper from 2019.

Explanation of when to use Poisson versus Binomial from stackexchange:

"If your outcome is discrete, or more precisely, you are dealing with counts (how many times something happen in given time interval),then the most common choice of the distribution to start with is Poisson distribution. The problem with Poisson distribution is that it is rather inflexible in the fact that it assumes that mean is equal to variance, if this assumption is not met, you may consider using quasi-Poisson family, or negative binomial distribution (see also Definition of dispersion parameter for quasipoisson family).

If your outcome is binary (zeros and ones), proportions of “successes” and “failures” (values between 0 and 1), or their counts, you can use Binomial distribution, i.e. the logistic regression model. If there is more then two categories, you would use multinomial distribution in multinomial regression."

C.1) Poisson

Interpretation: FST becomes less of a protective factor for metastasis as adjustments for age, largest basal diameter, and thickness DLS are made.

**Table 5C.1. Staged, Multivariate Regression of Metastasis (Poisson)**
Characteristic	Metastasis ~ FST			Metastasis ~ FST + Age			Metastasis ~ FST + Age + Largest Basal Diameter			Metastasis ~ FST + Age + Largest Basal Diameter + Thickness DLS
Characteristic	IRR¹	95% CI¹	p-value²	IRR¹	95% CI¹	p-value²	IRR¹	95% CI¹	p-value²	IRR¹	95% CI¹	p-value²
Fitzpatrick (FST) Scale	0.52	0.33, 0.87	0.010	0.55	0.34, 0.93	0.020	0.69	0.43, 1.13	0.13	0.66	0.40, 1.12	0.11
Age				1.02	1.00, 1.03	0.035	1.01	1.00, 1.02	0.2	1.02	1.00, 1.03	0.042
Largest Basal Diameter (mm)							1.18	1.13, 1.24	<0.001	1.24	1.17, 1.32	<0.001
Thickness DLS (mm)										0.97	0.89, 1.04	0.4
¹ IRR = Incidence Rate Ratio, CI = Confidence Interval
² Statistically significant P-values (<0.05) emboldened

C.2) Binomial

Interpretation: Similar trends to the Poisson regressions. FST become less of a protective factor for metastasis when adjustments for age, largest basal diameter, and thickness DLS are made.

**Table 5C.2. Staged, Multivariate Regression of Metastasis (Binomial)**
Characteristic	Metastasis ~ FST			Metastasis ~ FST + Age			Metastasis ~ FST + Age + Largest Basal Diameter			Metastasis ~ FST + Age + Largest Basal Diameter + Thickness DLS
Characteristic	OR¹	95% CI¹	p-value²	OR¹	95% CI¹	p-value²	OR¹	95% CI¹	p-value²	OR¹	95% CI¹	p-value²
Fitzpatrick (FST) Scale	0.43	0.24, 0.78	0.005	0.46	0.25, 0.84	0.011	0.54	0.29, 1.00	0.052	0.53	0.27, 1.03	0.063
Age				1.02	1.00, 1.03	0.020	1.01	1.00, 1.03	0.070	1.02	1.00, 1.04	0.017
Largest Basal Diameter (mm)							1.26	1.19, 1.33	<0.001	1.34	1.24, 1.44	<0.001
Thickness DLS (mm)										0.97	0.87, 1.08	0.6
¹ OR = Odds Ratio, CI = Confidence Interval
² Statistically significant P-values (<0.05) emboldened

Survival Analysis

Table 6

**Table 6. Adjusted Hazard Ratios of Metastasis and Death for FST and Relevant Covariates**
Characteristic	Metastasis			Death
Characteristic	HR¹	95% CI¹	p-value	HR¹	95% CI¹	p-value
FST	1.07	0.58, 1.99	0.8	1.02	0.30, 3.43	>0.9
Age	1.03	1.01, 1.05	0.001	1.04	1.01, 1.08	0.022
Largest Basal Diameter	1.01	0.94, 1.07	0.9	0.94	0.82, 1.08	0.4
Thickness DLS (mm)	1.22	1.09, 1.37	<0.001	1.32	1.03, 1.70	0.030
¹ HR = Hazard Ratio, CI = Confidence Interval

Risk of and Survival Probability of Metastasis

Risk of and Survival Probability of Death

Log-Rank Tests and Cox Models for Metastasis

#Log-rank test: difference between the survival curves for metastasis
survival::survdiff(Surv(time_till_mets_new, mets_new) ~ fst, data=data_survival)

## Call:
## survival::survdiff(formula = Surv(time_till_mets_new, mets_new) ~ 
##     fst, data = data_survival)
## 
## n=112, 576 observations deleted due to missingness.
## 
##        N Observed Expected (O-E)^2/E (O-E)^2/V
## fst=1 13       13     15.2    0.3172    0.3741
## fst=2 97       97     95.1    0.0378    0.2564
## fst=3  2        2      1.7    0.0524    0.0536
## 
##  Chisq= 0.4  on 2 degrees of freedom, p= 0.8

# Traditional adjustment for confounders is to use a Cox model and estimate hazard ratio
# Unadjusted Cox model
survival::coxph(Surv(time_till_mets_new, mets_new) ~ fst, data = data_survival)

## Call:
## survival::coxph(formula = Surv(time_till_mets_new, mets_new) ~ 
##     fst, data = data_survival)
## 
##       coef exp(coef) se(coef)     z     p
## fst 0.1722    1.1880   0.2681 0.642 0.521
## 
## Likelihood ratio test=0.42  on 1 df, p=0.515
## n= 112, number of events= 112 
##    (576 observations deleted due to missingness)

# Adjusted Cox model
survival::coxph(Surv(time_till_mets_new, mets_new) ~ fst + age + largest_basal_diameter + thickness_dls, data = data_survival)

## Call:
## survival::coxph(formula = Surv(time_till_mets_new, mets_new) ~ 
##     fst + age + largest_basal_diameter + thickness_dls, data = data_survival)
## 
##                            coef exp(coef) se(coef)     z        p
## fst                    0.070029  1.072540 0.315622 0.222 0.824409
## age                    0.028778  1.029197 0.009049 3.180 0.001470
## largest_basal_diameter 0.005612  1.005628 0.033678 0.167 0.867644
## thickness_dls          0.197509  1.218364 0.058280 3.389 0.000702
## 
## Likelihood ratio test=21.51  on 4 df, p=0.0002505
## n= 97, number of events= 97 
##    (591 observations deleted due to missingness)

Log-Rank Tests and Cox Models for Death

#Log-rank test: difference between the survival curves
survival::survdiff(Surv(time_till_death, death_new) ~ fst, data=data_survival)

## Call:
## survival::survdiff(formula = Surv(time_till_death, death_new) ~ 
##     fst, data = data_survival)
## 
## n=38, 650 observations deleted due to missingness.
## 
##        N Observed Expected (O-E)^2/E (O-E)^2/V
## fst=1  5        5   3.7489     0.417      0.49
## fst=2 32       31  33.1699     0.142      1.45
## fst=3  1        1   0.0811    10.408     10.70
## 
##  Chisq= 11.3  on 2 degrees of freedom, p= 0.004

# Traditional adjustment for confounders is to use a Cox model and estimate hazard ratio
# Unadjusted Cox model
survival::coxph(Surv(time_till_death, death_new) ~ fst, data = data_survival)

## Call:
## survival::coxph(formula = Surv(time_till_death, death_new) ~ 
##     fst, data = data_survival)
## 
##        coef exp(coef) se(coef)      z     p
## fst -0.1057    0.8997   0.5317 -0.199 0.842
## 
## Likelihood ratio test=0.04  on 1 df, p=0.8442
## n= 38, number of events= 37 
##    (650 observations deleted due to missingness)

# Adjusted Cox model
survival::coxph(Surv(time_till_death, death_new) ~ fst + age + largest_basal_diameter + thickness_dls, data = data_survival)

## Call:
## survival::coxph(formula = Surv(time_till_death, death_new) ~ 
##     fst + age + largest_basal_diameter + thickness_dls, data = data_survival)
## 
##                            coef exp(coef) se(coef)      z      p
## fst                     0.02131   1.02154  0.61731  0.035 0.9725
## age                     0.03981   1.04061  0.01741  2.287 0.0222
## largest_basal_diameter -0.06099   0.94083  0.07145 -0.854 0.3933
## thickness_dls           0.27933   1.32225  0.12835  2.176 0.0295
## 
## Likelihood ratio test=6.12  on 4 df, p=0.19
## n= 33, number of events= 33 
##    (655 observations deleted due to missingness)

Preliminary Fitzpatrick Results_1001

Samuel Goldstein

6/9/2022